A Monograph of Nymphaea Subgenus Hydrocallis (Nymphaeaceae)

A Monograph of Nymphaea Subgenus Hydrocallis (Nymphaeaceae)
Author(s): John H. Wiersema

Source: Systematic Botany Monographs, Vol. 16, A Monograph of Nymphaea Subgenus
Hydrocallis (Nymphaeaceae) (Apr. 6, 1987), pp. 1-112
Published by: American Society of Plant Taxonomists
Stable URL: http://www.jstor.org/stable/25027681
Accessed: 26-04-2018 15:08 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
American Society of Plant Taxonomists is collaborating with JSTOR to digitize, preserve

and extend access to Systematic Botany Monographs
This content downloaded from 163.178.101.95 on Thu, 26 Apr 2018 15:08:55 UTC
All use subject to http://about.jstor.org/terms
A MONOGRAPH OF NYMPHAEA SUBGENUS HYDROCALLIS
(NYMPHAEACEAE)
John H. Wiersema
Department of Botany
University of Maryland
College Park, Maryland 20742
Abstract. Nymphaea subgenus Hydrocallis is described to include 14 species. Evidence for this
classification was obtained from numerical taxonomy, scanning electron microscopy of seeds and
pollen, chromosome numbers, floral biology, artificial hybridization, flavonoid chemistry, and general
morphology. Representative species from other subgenera of Nymphaea were included in many of
these investigations for comparative purposes. Seed investigations were exceedingly valuable in defin
ing certain species and in clarifying their phylogenetic relationships. Pollen studies, on the contrary,
were of little value in distinguishing individual species but provided evidence for the relationship of
subg. Hydrocallis to other subgenera. The addition of information on chromosome number and
flavonoid profile for 12 taxa, supplemented by observations on floral biology and crossability, has
permitted a general phylogenetic interpretation of subg. Hydrocallis. Two major evolutionary lines are
apparent. One includes N. amazonum, N. prolif?ra, N. lasiophylla, and N. lingulata. Hybridization
events within this line involving subg. Lotos may have contributed to the formation of N. rudgeana,
but more evidence is needed to substantiate this interpretation. The other lineage includes N. novo
granatensis, N. gardneriana, N. conardii, and N. jamesoniana, with a secondary branch involving N.
glandulifera, N. potamophila, N. oxypetala, and perhaps N. belophylla. Nymphaea tenerinervia exhib
its affinities to both major lines and appears to represent an evolutionary link between them. A
complete taxonomic treatment of subg. Hydrocallis, including the description of the new subspecies N.
amazonum subsp. pedersenii, is presented.
INTRODUCTION
The nearly cosmopolitan genus Nymphaea Linnaeus is composed of aquatic

herbs with perennial rhizomes and mostly floating leaves. The flowers are showy
and borne solitarily, containing numerous (often intergrading) petals and sta
mens, and many carpels. Members of Nymphaea subg. Hydrocallis are further
characterized by completely fused carpels; highly developed, mostly cl?vate, car
pellary appendages; nocturnal flowering habit; and usually tetrameric arrange
ment of outer floral parts. Subgenus Hydrocallis includes 14 species native to the
American tropics and subtropics, with many of the species restricted to South
America.
Although regional floristic treatments have included a few species, no com
prehensive treatment of Nymphaea subg. Hydrocallis has been published since
those of Caspary (1878) and Conard (1905). Both of these treatments suffered
from a paucity of available plant material, both living and preserved. Only a few
collections were available at that time for most of the taxa treated, and living
material of only two species was obtainable. The complete range of variation
within the different species groups was therefore not embraced by these taxono
mists. This may have tended to accentuate supposed species differences and
rendered species descriptions and keys mostly incomplete and unworkable. In
2 SYSTEMATIC BOTANY MONOGRAPHS VOLUME 16
addition to the lack of sufficient distributional and general morphological infor

mation on many species of subg. Hydrocallis, only very limited information was
available on their reproductive biology and pollen and seed morphology, and
nothing was known regarding chromosome number or phytochemistry.
A herbarium study was thus undertaken to gather all available morphological
information, which was then analyzed by numerical taxonomic methods to pro
vide a preliminary assessment of taxonomic groupings within subg. Hydrocallis.
The generated groupings formed a basis for determining necessary field coverage
and focusing the attention of other more concentrated studies. An expanded
numerical taxonomic study, incorporating data from field populations, was then
completed and is detailed elsewhere (Wiersema 1984b). Additional studies of
seed and pollen morphology, chromosome number, floral biology, and flavonoid
chemistry, all presented here, provided evidence from several lines of inquiry for
the phylogenetic and taxonomic conclusions which follow.
TAXONOMIC HISTORY
The genus Nymphaea was established by Linneaus (1753) but originally in

cluded the present genera Nuphar Smith and Nelumbo Adanson as well. Subse
quent confusion arose with the retention of the name Nymphaea by Salisbury
(1806) for the yellow water lilies [Nuphar] and the establishment of a new name,
Castalia, for the white water lilies [Nymphaea]. This action preceded a similar
segregation of these genera by Smith (1809), who applied Nuphar to the yellow
water lilies and retained Nymphaea in its current status. After over a century of
controversy regarding the proper name for the two genera (Greene 1887a, 1887b,
1888; Britten 1888; Lawson 1888; Conard 1916; Mackenzie 1927), the name Nym
phaea was conserved, based on the European white water lily N. alba Linnaeus.
The first generic treatment to include members of the currently recognized
subgenus Hydrocallis was that of de Candolle (1821), who placed Nymphaea
rudgeana G. Meyer and N. blanda G. Meyer in separate sections, sections Lotos
de Candolle and Castalia (Salisbury) de Candolle respectively. Walpers (1842)
subsequently assigned N. amazonum Martius & Zuccarini and N. lasiophylla
Martius & Zuccarini to section Castalia, but did not treat N. rudgeana or N.
blanda. Planch?n (1852, 1853) was the first to recognize the similarity of the
cl?vate carpellary appendages of these taxa and established Nymphaea section
Hydrocallis. His section included all of the above species except N. rudgeana,
which he tentatively placed in synonomy under N. ampia (Salisbury) de Candolle,
and four new species. The most comprehensive treatment of the subgenus was
prepared by Caspary (1878) for Martius's Flora brasiliensis. Later ^Conard (1905),
in his worldwide monograph of Nymphaea, divided 34 specie^.into two groups,
Apocarpiae and Syncarpiae [corresponding to Caspary's (1865) sections Lyto
pleura and Symphytopleura], based on the extent of fusion of the carpel walls.
Conard recognized five subgenera, with three of these, including subg. Hydrocal
lis, expressing the syncarpous condition.
Conard, whose treatment of subgenus Hydrocallis was derived principally
from Caspary (1878), recognized ten species in this group. He tentatively added
Nymphaea gibertii (Morong) Conard to the nine species indicated by Caspary: N.
1987 NYMPHAEA SUBGENUS HYDROCALLIS 3
amazonum, N. blanda, N. gardneriana Planch?n, N. jamesoniana Planch?n, N.

lasiophylla, N. oxyp?tala Planch?n, N. rudgeana, N. stenaspidota Caspary, and N.
tenerinervia Caspary. These same species were recognized by Henkel et al. (1907)
in a treatment of the entire Nymphaeaceae. Since that time, and prior to the
initiation of my own studies, two additional species were described in the subge
nus: N. wittiana Ule (Ule 1915) and N belophylla Trickett (Trickett 1971).
MATERIALS AND METHODS
Seed Morphology. Seeds were obtained in most cases from herbarium speci
mens. Much of this material was the result of my own field collections, although a
small percentage of such seeds was obtained from borrowed herbarium speci
mens. In other instances seeds were obtained from plants in outdoor cultivation,
and in two such instances the seeds were the result of artificial hybridizations.
Available material from 12 species of subg. Hydrocallis was initially scrutinized
with the aid of a dissecting microscope. At that time, measurements of seed
length and width were taken, seed color was noted, and representative collections
were selected for more intensive light and scanning electron microscopy. Seeds of
four additional species of Nymphaea, representing the subgenera Nymphaea, Lo
tos, and Brachyceras, were also included for comparative purposes. Seed material
from 86 populations of these 16 species was initially examined; the 49 collections
receiving more intensive study are listed in Wiersema (1984b).
Three procedures were employed in examining the seed coat morphology of
these seeds. 1) In an SEM study of surface topography, seeds without arils were
required. These were selected from specimens; if such seeds were unavailable, the
aril was mechanically removed after soaking of seeds in Aerosol OT solution.
Seeds were then mounted on aluminum stubs with double adhesive transparent
tape. The samples were coated to 250 ? with a gold/palladium alloy in a Technics
Hummer V Sputter Coater, observed at 20 kV with an ETEC Autoscan SEM,
and photographed with Polaroid Type 55 P/N film. 2) For further examination of
surface cells, the outer testa was detached with forceps from softened seeds. This
testa was mounted in Hoyer's solution on slides, and examined and photographed
with a camera-equipped compound microscope. With the aid of a micrometer, ten
cells from each slide were measured for length, width, and the extent of intercala
tion at midcell. Because cell shape varies considerably over the seed surface, only
cells located medially in regular longitudinal rows were measured to provide a
more standardized estimate of cell shape for each collection of seeds. 3) To aid in
interpreting seed coat anatomy, seeds of all 16 species studied were fixed in 70%
ethanol-formalin-acetic acid, dehydrated in a tertiary butyl alcohol series, embed
ded in Paraplast (melting point, 56-57?C), sectioned at 8-15 jim on a rotary
microtome, and stained with safranin and fast green.
Pollen Morphology. All pollen for scanning electron microscopic analysis was
obtained from herbarium specimens, most of these from my own recent field
collections. The 30 populations furnishing pollen for SEM studies are listed in
Wiersema (1984b). Represented among these were 12 species of subg. Hydrocallis
as well as certain species of subgenera Lotos, Brachyceras, and Nymphaea. Pollen
preparation followed the procedure of Lynch and Webster (1975).
Chromosome Numbers. All counts were obtained from root tips, mostly using
an aceto-orcein staining procedure following pretreatment with 8-hydroxyquino
line. A total of 25 populations were examined for their chromsome numbers
(Appendix 1). Voucher specimens are deposited at UNA.
Floral Biology and Artificial Hybridization. Although field observations were
made at the time of collection in Mexico, Ecuador, Venezuela, Argentina, Brazil,
and Florida, the majority of observations were garnered from plants cultivated
from field-collected tubers. Plants were cultivated outdoors in Tuscaloosa, Ala
bama (33?N latitude) during May to October of 1982-1984; the majority of data
was gathered during 1983. A total of 28 populations of subg. Hydrocallis were
cultivated though not all progressed to flowering. They were grown in galvanized
washtubs (No. 3 size) in accordance with the recommendations of Pring (1941,
1949).
Several procedures were employed with flowers at anthesis. As no potential
pollinators were ever observed in any of the flowers, the latter were not bagged
but were merely tagged the day of the treatment. Some flowers were allowed to
develop normally in the absence of cross-pollination to test for autogamous or
agamospermous seed production. If seed production resulted from this treatment,
other flowers of the same plant were emasculated, allowed to develop without
pollination, and eventually tested for seed set. Other emasculated flowers were
artificially pollinated with pollen of another available flower (almost always
another species). This established that the effects of emasculation were not the
result of injury to flowers and enabled an assessment of the hybridization poten
tial between various species. To increase the number of available crosses, anthers
of some flowers were removed and allowed to dehisce in sealed plastic containers,
thus allowing these flowers both to serve as a source of pollen and to be polli
nated. Pollen treated in this manner exhibited no apparent loss of viability.
Subsequent to these procedures fruits were allowed to mature, and the seeds
or aborted ovules were collected at the time of dehiscence or decay of the flower
or fruit. Seeds or ovules from a given treatment were spread on p?trie dishes,
oven-dried, and later assayed for the percentages fully developed, partially devel
oped, and aborted. Pollen fertility was also examined for many species with the
use of aceto-carmine jelly according to standard procedures (Radford et al. 1974).
Data on crossing experiments were meant to be interpreted loosely in view of
the lack of replicate crosses and the possible effects of environmental factors, as
each developing fruit experienced a somewhat different set of environmental
conditions. Therefore, rigorous controls on the collection and assessment of seeds
were not enforced. Limitations in the number of potential crosses which became
available precluded a more thorough investigation. Nevertheless, a total of 53
artificial crosses was attempted among members of subg. Hydrocallis, an addi
tional 8 crosses were carried out between individuals of subg. Hydrocallis and N.
lotus of subg. Lotos, and one cross involved N. amazonum and N. odor ata of
subg. Nymphaea.
Flavonoid Chemistry. Leaf material for flavonoid analysis came from: 1) air
dried leaves from field collections, 2) air-dried leaves from mostly outdoor culti
vations, and 3) fresh leaves from outdoor cultivation. The majority of material
was of the first type. Leaves were crushed or fragmented by hand and subjected
to three 24 hour extractions in an aqueous methanol solution which was succ?s
sively decreased to ca. 50% methanol. The first extraction was carried out under
slight heating; the latter two with agitation. The combined methanol-water eluates
were evaporated under vacuum on a rotary evaporator to an aqueous solution.
Following several extractions of this solution with chloroform, flavonoids were
partitioned into ethyl acetate. The ethyl acetate was then evaporated to dryness
under vacuum, and the flavonoids redissolved in absolute methanol for use in
two-dimensional paper chromatography. Chromatographie separations were
achieved using tertiary butanol:glacial acetic acidiwater (3:1:1 by volume) as one
solvent and 15% aqueous acetic acid as the second solvent on Whatman 3MM
paper, and followed the procedures of Mabry et al. (1970). Further separations of
compounds unresolvable on paper were carried out on 0.1 mm Cellulose MN300
and 0.1 mm Polyamide-TLC6 precoated plates. Best results on polyamide TLC
plates were observed using a solvent system of 1,2-dichloroethane:methanol:
butanone:water (50:25:21:1 by volume).
Purified compounds were eluted from Chromatographie paper or plates with
methanol or, if necessary, compounds were purified using Sephadex LH-20 or
Polyamide SC6 column chromatography. Similarity of compounds from different
populations was detected by cochromatography on TLC precoated plates in a
minimum of three solvent systems (at least one of these on polyamide). Examina
tion of these plates was aided by fuming with ammonia and spraying with Natur
stoffreagenz A.
For identification of flavonoid compounds, the six standard UV spectra were
run in methanol (Mabry et al. 1970). Compounds were then hydrolyzed in boiling
6% HC1 for one hour. The aglycone was recovered in ethyl acetate for UV
spectral analysis and cochromatography with known standards. The aqueous por
tion was subjected to circular chromatography on 0.1 mm cellulose 400 TLC
plates with sugar standards using ethyl acetate:pyridine:water (6:3:2 by volume).
The sugars were visualized with aniline pthalate spray. Rf values of isolated com
pounds were determined on paper in various solvent systems and compared with
published values. For further characterization of certain compounds, proton
NMR spectra were obtained in deuterated DMSO.
A total of 54 populations representing 17 species and several assumed hybrids
were examined for their flavonoid profiles. These included three members of
Nymphaea subg. Nymphaea, two of subg. Brachyceras, and one of subg. Lotos,
with the remainder belonging to subg. Hydrocallis.
To examine the relationships among the various species of subg. Hydrocallis
and to explore their affinities to other members of Nymphaea, a statistical analysis
was applied to the chemical data by using the NT-SYS statistical package (Rohlf
et al. 1977). Two separate data matrices were subjected to cluster analyses. In
both matrices Jaccard coefficients were used to represent the degree of associa
tion between each pair of OTUs. This coefficient is determined by the:
number of shared compounds present

total number of compounds present
for any two OTUs. The first matrix was computed from data of the 56 compounds
isolated for the 54 populations or OTUs examined. As this analysis was incapable
of detecting similarities among different compounds, such as different glycosides

of the same aglycone, a second data set was analyzed. This incorporated the
presence-absence data for 11 aglycones and 9 glycosidic linkages from each OTU.
GENERAL MORPHOLOGY AND LIFE HISTORY
Habit. Species of Nymphaea are all perennial aquatic herbs. They are rooted
in what is always a rich organic substrate at the bottom of ponds or sluggish
streams, with the leaves and flowers elevated to the water surface on elongate
petioles and peduncles. A single mature plant may occupy an area from Vi m in
diameter in smaller forms to 3 m or more in larger forms. Flowers are floating or
may be emergent as much as 40 cm on stout peduncles. In Nymphaea subg.
Hydrocallis they are mostly floating, although a few species have slightly emer
gent flowers. Fruits of Nymphaea develop under water, being retracted there by a
bending or coiling of the peduncle; in subg. Hydrocallis an elongated S-shaped
bending is observed in most species.
Plants of Nymphaea withstand seasonal periods of cold or drought either as
seeds or as tuberous rhizomes. Although in subg. Hydrocallis tubers have not
demonstrated an ability to survive freezing temperatures, they are somewhat
resistant to drought. In fact, following cessation of growth at least a brief period
of drying out appears to be necessary for resprouting of tubers. Seeds of these
species may germinate directly following their release or may require a period of
drought.
Stems. The stems of some species of Nymphaea are elongate and horizontal,
but those of subg. Hydrocallis usually consist of an erect caudex a few centimeters
in length, although it may be up to several centimeters long in some species.
Inland species may develop a smooth and globose tuber or one that is more ovoid
and densely woolly. In many species these tubers are developed on lateral shoots,
either at the nodes of elongate stolons or, as in two species, within abortive
tuberiferous flowers. Certain species produce stolons only at the time of germina
tion of tubers; in others stolons are formed throughout the vegetative period and
are an effective means of vegetative propagation.
In species which do not form additional tubers, a tuberous rhizome serves as
the resting state. Several sprouts may develop from this dormant structure at the
onset of the next growing season. The longevity of individual plants which must
perennate in this fashion may be considerably reduced, as such a rhizome in
cultivation seems to remain viable only a few years. All species which fit this
pattern rely heavily on autogamous seed production for their survival and disper
sion. In N. rudgeana, which seems to inhabit more permanent aquatic habitats, a
true resting tuber may not be formed, with the plant maintaining constant growth.
This species also relies extensively on autogamy.
Roots. Although the petiolar and peduncular bases readily decay and absciss
from tubers at the onset of dormancy, certain roots are somewhat more persistent
and may perform a useful function. Roots attached to tubers of several species of
subg. Hydrocallis have been observed to be highly contracted, this contraction
probably serving to retract the tuber well into the substrate as the dry season
progresses. Conard (1905) indicates that contractile roots of other subgenera of
Nymphaea display anatomical differentiation from typical nutritive roots. The

former he observed only in plants which produce resting tubers.
Leaves. In subg. Hydrocallis leaves are mostly orbicular to elliptic in outline
and generally floating, although they are occasionally emergent due to exposure
from drought or under crowded conditions. In N. oxyp?tala, however, the leaves
are almost exclusively submersed, with very reduced floating leaves produced
only infrequently. Submerged-leaved forms of N. rudgeana have also been ob
served. This type of growth form seems to be somewhat restricted to lotie envi
ronments, particularly in N. rudgeana. In lentic situations, submersed juvenile
leaves are produced by a number of species, both from germinating seeds and
tubers. These are often sagittate in outline and thus resemble the submersed,
sagittate adult leaves of N. oxypetala. Both N. potamophila and N. belophylla
produce sagittate floating leaves, but the existence of submersed leaves in these
taxa has not been verified.
Leaf blades of all taxa have entire margins, except in N. rudgeana where they
are irregularly dentate. The blade is also much thicker or coriaceous in N. rudge
ana than in other members of subg. Hydrocallis. Pubescence is absent from the
undersurface of leaf blades in subg. Hydrocallis, although generally present in the
related subg. Lotos (de Candolle) Conard. In subg. Hydrocallis leaves are less
peltate than those of subg. Lotos, the lobes being fused only slightly or not at all
in most species. Some populations of N. gardneriana and N. lasiophylla are excep
tions to this pattern, however. The majority of species, excepting N. glandulifera,
most N. conardii and N. jamesoniana, and some N. gardneriana, exhibit accessory
leaf pigmentation, usually purplish or reddish. This is especially evident on imma
ture leaves, particularly on their exposed abaxial surface. This coloring is confined
to scattered splotches or spots in mature leaves but appears to be more continu
ous in juvenile leaves. Although the specific chromatic compounds responsible
are not known, those species with such coloration also exhibit the most diverse
flavonoid profile.
An important feature of Nymphaea leaves is the presence of variously shaped
sclereids or idioblast cells within the mesophyll. They are discussed by Conard
(1905) and have attracted the attention of other researchers as well (Gaudet 1960;
Malaviya 1962; Rao & Banerjee 1979). Their taxonomic value in subg. Hydrocal
lis was first exploited by Caspary (1878), who termed them "pachycysts." Several
distinct forms are evident in leaf blades in subg. Hydrocallis, except those of N.
oxypetala where they have not been observed. The most common type were
referred to by Conard as stellate idioblasts or by others as astrosclereids, these
occurring in the spongy mesophyll. They are of least importance taxonomically
and are not evident externally. In dried herbarium specimens of leaves the other
forms are often much more apparent. Most noticeable are the trichosclereids
(referred to by some authors as polyramous sclereids), which project above the
palisade mesophyll and produce tiny prominences on the upper surface of dried
leaves. These are found in the laminas of all members of subg. Hydrocallis except
N. oxypetala, although variation in their density exists among taxa. In N. lingulata
they usually produce characteristic starshaped expressions on the leaf surface. The
most diagnostic sclereid shape is that evidenced by Conard's bipolar idioblasts,
which I have termed acicular sclereids because of their needlelike morphology.
These occur in N. jamesoniana, where they are in greatest abundance, in N.
conardii, N. gardneriana, N. novogranatensis, N. glandulifera, N. belophylla, N.

lingulata, and many populations of N. amazonum subsp. pedersenii. In the latter
two they are more filiform in shape and are far less apparent externally than in
other taxa, where they often produce short linear expressions on the upper sur
face of dried leaves.
Acicular sclereids are absent from leaf blades of N. prolif?ra, N. lasiophylla,
N. amazonum subsp. amazonum, N. rudgeana, N. potamophila, and N. oxyp?
tala. In N. tenerinervia these sclereids always develop Y-shaped branches, which
characterize this species.
The leaf venation also exhibits some variation within subg. Hydrocallis. In
most species this venation resembles that of other species of Nymphaea, with a
basically palmate arrangement of most major veins and a pinnate arrangement
associated with the midvein. In N. conardii, N. gardneriana, and especially N.
jamesoniana, cross veins are evident between the major veins near the center of
the leaf, producing a characteristic spider-webbed appearance. The number of
major veins differs among the taxa as well, being fewest in the sagittate-leaved
species and greatest in N. rudgeana or N. lasiophylla. The number of leaf veins is
roughly correlated with the leaf size within a given species.
Petioles of most species of subg. Hydrocallis are glabrous, although pubes
cence has been observed, especially in young leaves, in N. tenerinervia, N. glan
dulifera, N. gardneriana, and N. amazonum. The usefulness of leaf pubescence as
a taxonomic character is doubtful in all but N. amazonum, where a distinct and
persistent ring of hairs is formed at the petiolar apex. In other species pubescence
may not persist in adult leaves or may be absent in some populations. Some
variation is apparent among species in the patterning of air canals within the leaf
stalks. In N. oxy p?tala, N. jamesoniana, N. amazonum, and N. rudgeana, two
large air canals and several smaller ones are present. Two or four large canals are
likely in N. gardneriana and N. novogranatensis, with four the usual number in N.
conardii; smaller canals are also present in these species. In N. lingulata, 2-6
major canals are present but smaller canals are absent; a dense ring of tissue
surrounds the inner canals. The number of large air canals increases to 4 or 8 in
N. tenerinervia, 8 in N. prolif?ra, and 8-12 in N. lasiophylla. Nymphaea lasio
phylla resembles N. lingulata in lacking lesser canals.
Flowers. Flowers of subg. Hydrocallis resemble those of other Nymphaea
subgenera in many ways; the unique departures are the highly developed and
cl?vate nature of the carpellary appendages and the tetrameric disposition of the
perianth. These two features vary in the degree to which they are manifest in
various species. In subg. Hydrocallis the carpels are fused over most of their
lateral walls, which accounts for the inclusion of this subgenus with subgenera
Lotos and Nymphaea in the syncarpous group of the genus [see Moseley (1961)
for a discussion of syncarpy vs. apocarpy in Nymphaea]. The flowers open noctur
nally like those of subg. Lotos, although in this latter subgenus they remain open
through the morning as well.
The sepals of some members of subg. Hydrocallis are marked abaxially with
tiny black or purplish spots or streaks. Such markings are evident only in a related
group of species including N. amazonum, N. prolif?ra, N. lingulata, and N. teneri
nervia. Nymphaea lasiophylla, which otherwise would be grouped here, lacks such
markings. In many ways the extent of coloration of the sepals parallels that of the
leaves, as both are pure green in N. jamesoniana, N. conardii, and N. glandulif

era. In some dried specimens of these latter species, of N. gardneriana, and of the
sagittate-leaved species reddish or brownish lines of uncertain origin are evident
on the sepals.
Petals of all species of subg. Hydrocallis are fairly uniform in color, ranging
from a creamy white to a very faint yellow. In N. rudgeana a pinkish color may
develop in older flowers. The outer petals are typically more rounded apically
with the inner more acute or acuminate. In N. oxypetala all petals, together with
the sepals, are long-acuminate. Species vary in the extent to which the petals form
ordered whorls of four. All petals are arranged thus in most species of subg.
Hydrocallis, but in N. prolif?ra, N. lasiophylla, N. lingulata, and N. oxypetala this
arrangement is only apparent in the outer petals, the inner appearing more spi
ralled in their attachment. The petals are normally gradually narrowed from the
outer to the inner, but in N. novogranatensis and N. tenerinervia the three outer
whorls are distinctly broader than the inner petals. The petals of N. lasiophylla
and N. lingulata display an ornate or lacework type of venation that is most
striking in the dried condition.
The outer stamens are petaloid in all species of subg. Hydrocallis except N.
lasiophylla and N. lingulata, where the transition from petals to stamens is more
abrupt. Species vary in the degree to which the stamens are appendaged by the
prolongation of the connective, but they are far less appendaged than those of
subg. Br achy ceras (Caspary) Conard. The anthers are positioned medially on the
stamens; this characteristic is especially evident in the outer stamens and differs
from the marginal placement of anthers in subg. Nymphaea. In several members
of subg. Hydrocallis the filaments, especially the inner, are suffused with blue or
purple. This is very evident in N. oxypetala, N. prolif?ra, and N. lingulata and less
so in N. tenerinervia and N. novogranatensis.
The carpels are completely fused laterally, with single partitions existing be
tween adjacent carpels in the ovary. The upper surface of each carpel forms a ray of
stigmatic tissue which terminates abaxially in a free carpellary appendage. These
appendages vary from lingulate-tapering in N. lasiophylla and N. lingulata; cylin
dric-clavate in N. novogranatensis and N. tenerinervia; linear-clavate in N. glandu
lifera, N. gardneriana, and N. conardii; to strongly cl?vate in N. rudgeana and N.
amazonum. They range from 3 mm long in some flowers of N. jamesoniana to
almost 3 cm in N. oxypetala. They are strongly upwardly incurved or coiled in many
species, particularly in older flowers, but in N. oxypetala are abruptly bent at the
distal one-third. In many species these appendages are deeply pigmented.
The stigmatic papillae of several members of subg. Hydrocallis are unique
within Nymphaea in that the individual cells separate from one another to form
a powdery mass which covers the stigmatic surface (Fig. la, b). This condition
exists in N. tenerinervia, N. novogranatensis, N. amazonum, N. prolif?ra, and N.
rudgeana.
Species vary in their relative numbers of certain flower parts, although the
number varies considerably within a given taxon. The fewest number of petals is
found in N. lingulata, the most occurring in N. prolif?ra. The greatest number of
stamens, over 300, are present in some flowers of N. amazonum subsp. pederse
mi. The occurrence of acicular sclereids (Fig. Id) in various flower parts is, on the
contrary, quite consistent within a given species, particularly with regard to the
FIG. 1. Unusual anatomical features of certain flowers of Nymphaea subg. Hydrocallis. a. Termi
nal papilla cell of powdery stigma in N. rudgeana (Wiersema et al 2309). b. Subterminal papilla cell of
N. rudgeana (Wiersema et al. 2309). c. Spherical sclereid from anther of N. oxypetala (Fern?ndez
1299). Note embedded crystals of calcium oxalate. d. Acicular sclereid from anther of N. potamophila
(Madison et al. 6155). Bar = 10 um.
petals and stamens. Acicular sclereids may be found within the sepals, petals,
stamens, and carpellary appendages, or even projecting from the stigmatic ray
tissues in a few species. They are thick and needlelike in many species, but in
others they are thinner or more filiform. In the stamens of N. oxypetala unusual
spherical sclereids (Fig. lc) are apparently released (in dried anthers at least)
from the staminal interior along three sutures lateral to and between the anther
sacs. They were studied by Chifflot (1902) but their functional significance is
unknown.
In N. prolif?ra and N. lasiophylla, abortive tuberous flowers are produced.

Similar structures have been reported for other syncarpous species of Nymphaea
but appeared to be an unusual occurrence (Mohan Ram & Nayyar 1974; Mitra &
Subramanyam 1982). They are produced in abundance in N. prolif?ra and N.
lasiophylla and remain entirely submersed until they are released. An outer layer
of sepaloid bracts surrounds the inner tuber, which is densely woolly on its ex
posed surface. The tubers give rise to leaves in whorls alternating with whorls of
similar tuberous flowers. Consequently, three or four orders of branching of these
abortive flowers may result. Roots develop directly from the tuberous tissue.
Tubers readily absciss along a layer of tissue at the apex of the peduncle, some
already having progressed to the point of being juvenile plants. The peduncles of
abortive flowers resemble those of normal flowers in these two species in having
the same number and arrangement of air canals. Peduncular air canals usually
consist of an inner ring of 6 large canals surrounded by 12 smaller ones; however,
in some species all 18 may be approximately equal in size. In N. oxypetala only 4
large canals are present.
Fruits. In subg. Hydrocallis fruits, as in other species of Nymphaea, are gener
ally globose in shape, ranging from a few centimeters in diameter in most species
to the 8 cm in diameter observed in an Argentine population of N. jamesoniana.
Their development in all species of subg. Hydrocallis is dependent upon success
ful pollination, but in several species they always develop autogamously. In N.
jamesoniana and N. conardii the ovaries appear to enlarge even before self-polli
nation has occurred. In all species the sepals are persistent, with the petals and
stamens mostly decayed at the time of fruit dehiscence. The fruiting peduncles in
subg. Hydrocallis bend only slightly to retract the developing fruits beneath the
water surface, in contrast to the highly recoiled peduncles in subg. Nymphaea.
SEED MORPHOLOGY
According to Conard (1905), the seeds of water lilies display marked charac
teristics which can be of systematic value. They were the object of a study of the
Nymphaeaceae (sensu lato) by Weberbauer (1894), whose findings regarding
Nymphaea are discussed by Conard. The seed coat in Nymphaea has generally
been regarded as bitegmic (Chifflot 1902; Conard 1905; Khanna 1967). The outer
most cell layer is composed of lignified cells (sclereids), whose interdigitating
radial cell walls form a puzzlelike network over the surface of the seed. These
sclereids are, for the most part, arranged in longitudinal rows. They may be
almost isodiametric or squarish, but more frequently are elongated in a direction
perpendicular to the rows, except along the raphe where the elongation is in the
direction of the rows. The raphe may form a ridge of varying prominence along
one side of the seed. Other longitudinal ridges may be formed by the intercalating
ends of adjacent rows of sclereids being alternately raised up or level. The rows,
and the ridges when present, are interrupted as they converge toward both the
chalazal and micropylar ends of the seed. The outer tangential surface of the
sclereids may be variously supplied with papular outgrowths as well. These hair
like structures are, according to Conard, outgrowths of the radial cell walls of the
sclereids.
Another feature of Nymphaea seeds is the aril, a bellshaped outgrowth of the

funiculus, which variously encloses the seed and, through its mucilage and
trapped air bubbles, serves to provide flotation for seeds subsequent to their
release from the ovary.
Variations in the location and size of the hairlike papillae and in the surface
topography formed the basis for Weberbauer's (1894) distinction of four types of
seed testae in Nymphaea. These types, with representative taxa as indicated by
Weberbauer and supplemented by Conard, are as follows (adapted from Conard,
1905): 1) Surface bare, smooth, and shining: includes chiefly temperate species of
subg. Nymphaea. 2) Surface with longitudinal ridges along alternate cell junc
tures; no hairs present: includes certain members of subg. Lotos. 3) Surface
lacking ridges, but with hairs in longitudinal rows along alternate cell junctures:
includes the remaining taxa of subg. Lotos, most members of subg. Brachy ceras,
subg. Anecphya (Caspary) Conard, and two species of subg. Hydrocallis, N.
amazonum and 'W. blanda" (the latter actually N. rudgeana, according to Con
ard). 4) Ridges absent; hairs short, present along both longitudinal and transverse
cell junctures: includes two members of subg. Brachyceras.
Seeds of the subtropical N. mexicana Zuccarini of subg. Nymphaea, the larg
est seeds of all Nymphaea, are peculiar in several additional characteristics as
well. They are uniformly covered with long hairs originating from all sides of the
digitate surface sclereids, with no ridges present (Capperino & Schneider 1985).
According to Conard's interpretation, they combine characteristics of Weber
bauer's third and fourth types.
Collinson (1980) conducted a scanning electron microscopic study of seeds of
both fossil and extant Nymphaeaceae (sensu lato), including those of several
species of Nymphaea. Seeds of nymphaeaceous genera were differentiated on the
basis of testal features and relative positioning of the micropyle and hilum. In the
species of Nymphaea examined, the micropyle consisted of a shallow circular
depression and the hilum an adjacent but separate rounded scar. Both structures
together form part of an apical cap at one end of the seed.
Scanning electron microscopy has been utilized by other workers in studying
seeds of individual genera of the Nymphaeaceae (e.g., Barclaya Wallich
[=Hydrostemma Wallich], Schneider 1978; Ondinea Hartog, Schneider & Ford
1978). Seeds of Nymphaea subg. Hydrocallis have not been examined previously
with SEM, since herbarium collections containing mature seeds were rare or not
available for most species.
Results and Discussion
Seed Size and Color. Seeds of Nymphaea subg. Hydrocallis range from nearly
orbicular in N. jamesoniana and N. oxypetala to highly ellipsoid in some popula
tions of N. amazonum. With the exception of the relatively large seeds of N.
rudgeana, seeds of subg. Hydrocallis are smaller than those of other species of
Nymphaea. The smallest are those of N. jamesoniana, which barely exceed 0.5
mm in diameter. For some species, such as N. tenerinervia and N. glandulifera,
seed size and shape are reasonably uniform throughout the distributional range.
In others, such as N. conardii and especially N. amazonum and N. ampia of subg.
Brachyceras, both size and shape vary from one region to another. Seed material
of other members of subg. Hydrocallis is too scanty to enable an assessment of

geographic variation.
Although seed color, as determined from dried material, does seem to vary
from one species to another, it is of questionable taxonomic value. Within a given
species seed color may vary from one part of its range to another and from one
stage of seed maturation to another. Also, colors observed on dried material may
fluctuate with the particular drying process utilized and may in no way reflect
those actually present in nature.
Surface Morphology. Seeds of Nymphaea subg. Hydrocallis exhibited consid
erable variability in surface morphology among the various species. Three species
specific patterns of variation are related to the papillae or hairs on the seed
surface and three more are characteristics of the testal sclereids. These variations
are: 1) in length of hairs, ranging from 10-50 \im in N. jamesoniana to over 300
[xm in N. rudgeana; 2) in arrangement of hairs, with three types: a) in longitudinal
rows along alternating cell junctures, b) in longitudinal rows along each cell
juncture, and c) along both longitudinal and transverse cell junctures; 3) in spatial
distribution of hairs, either singly or in clumps; 4) in shape of surface sclereids,
from nearly square through short-rectangular to long-rectangular; 5) in the extent
of interdigitation among surface sclereids; and 6) in the fine surface topography of
sclereids, from smooth or slightly ruminate to granulate. Several of these features
are evident in Figure 2a, b.
On all seeds of subg. Hydrocallis examined, a distinct micropylar ridge was
observed surrounding the micropyle. A portion of this ridge was noticeably
broadened as an extension of the raphe, which terminated there as the hilum.
Cells of this ridge were smaller, more isodiametric, and less intercalated than
those over the remainder of the seed surface. With this arrangement the micro
pyle and hilum were not observed as distinctly separate structures, as reported for
other species of Nymphaea (Collinson 1980). In addition, no apical cap could be
readily discerned on any of these seeds.
Seed Coat Anatomy. Light microscopic examination of whole mounts and
sections of the seed coat aided in interpreting surface features and provided
evidence for other variation among species. Differences in fine surface structure
of seed coats were readily observable from whole mount preparations. Cells with
a granulate surface appeared areolate in the surface focal plane due to the optical
effects of these surface irregularities (Figs. 4b, e, h, 5b, e, h). Those with smooth
or ruminate surfaces did not display this characteristic areolate pattern when
similarly viewed (Figs. 3b, c, h, 6b, e, h, 8b, e, h, 9b). Numerous pit canals were
evident in the tangential walls of surface sclereids. In seed testal whole mounts
under the light microscope these pit canals appear as opaque granules, varying
somewhat in their density and configuration among the various taxa (e.g., Figs.
3b, 6b). In seed coat sections the pit canals of most species are clearly transparent
(e.g., Fig. 4c); however, in certain species dark inclusions are evident in some
canals (e.g., Fig. 6c, f).
In a longitudinal or transverse section of a typical mature seed of Nymphaea
subg. Hydrocallis, the surface sclereids display extensive thickening on their radial
and outer tangential walls, which with the thin inner walls enclose a narrow
lumen. An additional tier of compressed thin-walled cells is perceptible beneath
the outer sclereid coat, with the pink stain imparted to both by safranin indicating
FIG. 2. Light and scanning electron micrographs of seeds of Nymphaea subg. Hydrocallis. a.
SEM of whole seed, illustrating short, solitary hairs and highly intercalated, rectangular surface
sclereids with granulate fine surface topography, b. SEM of whole seed, illustrating medium-length,
clumped hairs with less intercalated, more isodiametric sclereids having a smooth topography, c. Near
median longitudinal section of seed of N. novogranatensis (Wiersema & Gonz?lez 2234) showing
overall seed anatomy; o = outer integument, i = inner integument, p = perisperm, a = aril, mr =
micropylar ridge, op = operculum, pi = plumule, c = cotyledon, e = endosperm. Bar = 50 |xm.
the presence of lignin. The outer integument is often detached from the inner
upon drying of the seed by breakdown of these thin-walled cells and the inner cell
walls of the sclereid coat. The inner integument, which stains a deep red with
safranin, appears as an amorphous band, although Chifflot (1902) indicated it was
composed of two layers of cells. Lying just within the inner integument, a narrow
transparent layer surrounds the starch-filled perisperm. Perisperm occupies ca.

80-90% of the inner seed volume, with a small band of endosperm surrounding
the tiny embryo. The overall seed anatomy can be seen in Figure 2c; details of the
testal anatomy are best viewed in the testal sections of individual species (e.g.,
Fig. 4c, f).
Although the anatomical observations accumulated are not sufficient to per
mit an estimate of intraspecific variation, some exceptions to the general features
described above are apparent. In N. rudgeana (Fig. 3c) and to some extent in N.
conardii (Fig. 5f ) the outer tangential wall of the sclereid coat is comparatively
thinner than in other members of subg. Hydrocallis, with the cell lumen conse
quently larger. In N. lasiophylla (Fig. 3f) and N. lingulata (Fig. 3i) the inner
tangential walls of this layer are more lignified. Such is also the case in some of
the N. conardii seeds examined, which suggests that a certain degree of variation
in lignification patterns might be expected in the other species as well. Although
there is some variation in the thickness of the inner integument among other
members of subg. Hydrocallis, this layer appears to be more developed in the
larger-seeded N. rudgeana (Fig. 3c), in N. lotus Linnaeus of subg. Lotos (Fig. 9c),
and N. ampia of subg. Brachyceras (Fig. 8i). Comparably sized seeds of N.
odor ata Ait?n of subg. Nymphaea and N. elegans Hooker of subg. Brachyceras
are not similarly endowed (Fig. 8c, f). Their inner integuments are ca. one-half or
less as thick as the outer integuments, in contrast to the similar thickness of the
two layers in N. rudgeana and N. lotus. Seeds of N. odor ata of subg. Nymphaea
appear to have more than two cell layers constituting the outer integument (Fig.
8c), in agreement with the observations of Chifflot (1902) on other members of
this subgenus.
Seed Classification. Based on the seed characters noted, species of Nymphaea
subg. Hydrocallis tend to cluster into four major groups. Group I includes only N.
rudgeana, whose seeds are ellipsoid, with longitudinal rows of very long solitary
hairs along alternate cell junctures and long-rectangular, smooth-surfaced scle
reids. Within the subgenus these seeds are distinguished primarily by their larger
size, longer papillae, and variant integumentary anatomy (Fig. 3a-c).
Group II contains species with seeds that are ellipsoid, with longitudinal rows
of short-to-medium-length solitary hairs along alternate cell junctures and long
rectangular sclereids. The latter are very regularly arranged and have a ruminate
fine surface topography and more lignified inner tangential walls. Included in this
group are two species of eastern Brazil, N. lasiophylla (Fig. 3d-f ) and N. lingulata
(Fig. 3g-i), which differ in the former having seeds with medium length hairs and
the latter seeds with short hairs. The apparent similarity of these two species on
grounds of seed morphology is not particularly surprising, when one considers
their uniquely similar floral morphology; they are the only species in subg. Hydro
callis with tapering (rather than cl?vate) carpellary appendages.
Group III contains species whose seeds are suborbicular to short-ellipsoid,
with usually short-to-medium-length solitary hairs which are variously arranged,
and rectangular to nearly square surface sclereids. Seeds of this group are unique
within the genus Nymphaea in having a granulate fine surface topography. Seven
species constitute this group, their seeds differing somewhat in size and shape.
The group has been divided into subgroups based upon the arrangement of sur
face hairs. Subgroup Ilia, containing N. amazonum (Fig. 4a-c), N. tenerinervia
FIG. 3. Group I and II seeds of Nymphaea subg. Hydrocallis. a-c. Group I, N. rudgeana. d-f.
Group II, N. lasiophylla. g-i. Group II, N. lingulata. a, d, g. SEM of whole seeds, showing arrange
ment of papillae along alternate cell junctures (a. Wiersema et al. 2299. d. Wiersema et al. 2316. g.
Wiersema et al. 2308.). b, e, h. Light micrographs of seed coat, lacking granulate surface, with fewer
pit canals (arrow) in b and numerous pit canals in e and h. c, f, i. Light micrographs of seed coat
section. Note thin outer tangential wall and thick inner integument in c and thicker inner tangential
walls (arrow) of surface sclereids in f and i; o = outer integument, i = inner integument. Horizontal
bar = 0.1 mm, vertical bar = 20 \im.
16
FIG. 4. Group Ilia seeds o? Nymphaea subg. Hydrocallis. a-c. N. amazonum subsp. amazonum.
d-f. N. tenerinervia. g-i. N. gardneriana. a, d, g. SEM of whole seeds, showing arrangement of
papillae along alternate cell junctures (a. Wiersema et al. 2304. d. Wiersema et al. 2307. g. Wiersema
2247a x 2214). b, e, h. Light micrographs of seed coat, displaying granulate surface, c, f, i. Light
micrographs of seed coat section showing details of testal anatomy; o = outer integument, i = inner
integument, p = perisperm, a = aril, t = transparent layer surrounding perisperm, arrow = pit canal.
Horizontal bar = 0.1 mm, vertical bar = 20 urn.
17
(Fig. 4d-f ), N. gardneriana (Fig. 4g-i), and N. prolif?ra, is distinguished in hav

ing seeds with longitudinal rows of hairs along alternate cell junctures and short
to long-rectangular surface sclereids. Seeds of these species differ somewhat in
size and papilla length.
The seed morphology of N. gardneriana was determined by viewing seeds from
artificial hybridizations involving N. gardneriana (female parent) and N. conardii
or N. novogranatensis (male parent), because N. gardneriana seldom produces
seeds in nature. The seed coat in hybridized Nymphaea seeds would be expected to
display primarily maternal attributes, because ca. 90% of the mature seed origi
nates from integumentary and nucellar tissues of maternal genotype, tissues that
are well-formed prior to fertilization. Observations of presumably non-hybrid seeds
of N. gardneriana from three Venezuelan and Paraguayan herbarium collections
which display a similar morphology support such an interpretation. Although seeds
of N. prolif?ra have never been observed in field populations or on herbarium
material, those recently obtained from an artificial hybridization of this species with
N. amazonum subsp. pedersenii indicate that its seeds are characteristic of this
group as well.
Subgroup IHb, including N. novogranatensis (Fig. 5a-c) and N. conardii (Fig.
5d-f), is distinct in having longitudinal rows of seed hairs along each cell junc
ture. The surface sclereids in this subgroup vary from short-rectangular in N.
novogranatensis to more nearly square or isodiametric in N. conardii. In compari
son with other Group III species, they appear to be somewhat less thickened on
their outer tangential walls, which in dried seeds of N. novogranatensis may
collapse to form a ridged surface topography. Seeds of N. conardii are somewhat
transitional to those of the next subgroup, with hairs occasionally present along
transverse cell junctures.
Subgroup IIIc, containing only N. jamesoniana, has the smallest and most
nearly orbicular seeds with the shortest surface hairs among members of subg.
Hydrocallis. These hairs are located along both transverse and longitudinal cell
junctures (Fig. 5g-i). Certain seeds show a reduction in numbers of transversely
arranged hairs over parts of the surface and are therefore intermediate with those
of the previous subgroup.
Group IV contains species whose seeds have short-rectangular or nearly
square surface sclereids with smooth topography and have longitudinal rows of
hairs along alternate cell junctures, with medium-to-long hairs occurring mostly in
clumps. Two species, N. oxypetala (Fig. 6a-c) and N. glandulifera (Fig. 6d-f ), are
included here, their seeds differing somewhat in shape. Based on overall mor
phology, this would seem to be a rather odd pairing?N. glandulifera appearing
similar in floral and vegetative anatomy to certain species of Group III and N.
oxypetala seeming not-too-closely related to any other species of the Hydrocallis
group. An association of N. glandulifera and N. oxypetala is supported by N.
potamophila of northern Brazil, which displays a morphology somewhat interme
diate between them. Seeds of N. potamophila (Fig. 6g-i), which has sagittate
leaves reminiscent of N. oxypetala and flowers similar to those of N. glandulifera,
exhibit the smooth surface topography and clumping of hairs characteristic of this
group. Surface sclereids of these seeds are, in general, more irregularly shaped
and more deeply intercalated than those of the other two species.
An unusual feature was observed on seeds from two populations otherwise
FIG. 5. Group Illb and IIIc seeds of Nymphaea subg. Hydrocallis. a-c. Group Illb, N. novo
granatensis. d-f. Group Illb, N. conardii. g-i. Group IIIc, N. jamesoniana. a, d, g. SEM of whole
seeds (a. Wiersema & Gonz?lez 2234. d. Wiersema & Gonz?lez 2214. g. Wiersema et al. 2277). b, e, h.
Light micrographs of seed coat, displaying granulate surface, c, f, i. Light micrographs of seed coat
section. Note thinner outer tangenital wall of testa and surface papillae (arrow) in f. Horizontal bar =
0.1 mm, vertical bar = 20 ^m.
19
t*>i ^ US'* JF
FIG. 6. Group IV seeds of Nymphaea subg. Hydrocallis. a-c. N. oxypetala. d-f. N. glandulifera.
g-i. N. potamophila. a, d, g. SEM of whole seeds, showing clumped distribution of hairs (a. Wier
sema & Gonz?lez 2232. d. Liesner 2892. g. Fr?es 22686). b, e, h. Light micrographs of seed coat, with
numerous pit canals (arrow) and lacking granulate surface, c, f, i. Light micrographs of seed coat
section, showing pit canals with dark inclusions (arrows). Horizontal bar = 0.1 mm, vertical bar = 20
urn.
20
k* <sr k\ c>j k\ k> ? k\ *-i
FIG. 7. Hypothetical evolutionary scheme

the relationships of Weberbauer's seed types
referable to N. glandulifera from

had developed not along the transv
sclereids but rather occurred solita
wall. In the Venezuelan material b
marginal positioning of papillae nor
supporting the retention of these t
If Weberbauer's (1894) seed-class
ployed, the seeds of all members of
the exception of Subgroups Illb a
seeds of N. jamesoniana (Subgroup
of N. conardii and N. novogranat
scribed by any of Weberbauer's cat
Phylogeny of Hydrocallis Seeds. F
scheme for Nymphaea as presented
Conard, Weberbauer (1894), and my
berbauer's seed classification, indica
within the genus. From this constr
seed must be the primitive seed ty
major subdivisions of Nymphaea an
FIG. 8. Seeds of other subgenera of Nymphaea. a-c. N. odorata of subg. Nymphaea. d-f. N.
elegans of subg. Brachyceras, g-i. N. ampia of subg. Brachyceras, a, d, g. SEM of whole seeds (a.
Wiersema 342. d. Pringle 1955. g. Wiersema 2204a). b, e, h. Light micrographs of seed coat with
numerous pit canals and lacking granulate surface, c, f, i. Light micrographs of seed coat section. Note
thicker inner integument in i and additional cell layers present in outer integument of c. Horizontal
bar = 0.1 mm, vertical bar = 20 urn.
22
An examination of available seeds from other subgenera of Nymphaea pro

vides a further characterization of Weberbauer's seed classes. Seeds of N. odo
rata of subg. Nymphaea (Fig. 8a-c) are highly cuticularized, devoid of surface
papillae, and have somewhat irregularly shaped, short-rectangular sclereids. In
dicative of Weberbauer's Type 1 seeds, they are larger than any of subg. Hy
drocallis, excepting those of N. rudgeana, and do not appear to be closely
related to Hydrocallis seeds. Seeds of N. elegans of subg. Brachy ceras (Fig.
8d-f), characterized by Conard as Type 4, have very irregularly shaped sur
face sclereids with a seemingly random distribution of numerous short papillae.
They bear a superficial resemblance to seeds of N. jamesoniana of subg. Hy
drocallis, but differ in their increased size, in the anatomy of the outer testal
cells, and in the smooth surface topography of these cells. Since these two
species are in separate subgenera which contain members possessing the prob
able ancestral seed type, the apparent similarity in their seeds can only be
attributed to evolutionary convergence.
Seeds of N. ampia, also of subg. Brachyceras, and those of N. lotus of subg.
Lotos both exhibit Weberbauer's Type 3 morphology, the probable ancestral seed
type. Those of N. ampia (Fig. 8g-i) display the very regular arrangement of
long-rectangular sclereids reminiscent of Group II, from which they are distin
guishable by their larger size and longer surface hairs. The seeds of N. lotus (Fig.
9a-c), with a size comparable to N. ampia seeds, have a less regular arrangement
of long-rectangular sclereids. Those obtained from a naturalized Florida popula
tion differed slightly from the native African counterparts in having shorter and
more randomly arranged hairs.
Nymphaea lotus seeds in sectional view are very similar to those of N. rudge
ana (Group I). The integumentary layers correspond both in thickness and lignifi
cation patterns between the two species and comparatively fewer pit canals are
evident in their testae than in most of the other species examined. Certain other
morphological characteristics of N. rudgeana, which differ from other members of
subg. Hydrocallis, also suggest a link with subg. Lotos. These include dentate
leaves, stamens with broad anther sacs, and often emergent flowering habit.
However, differences in pubescence, sepal venation, and flowering responses sup
port the current classification of the two species. In view of the seed evidence
presented and these other morphological considerations, N. rudgeana might rep
resent either a species separately derived from an ancestor to both subgenera, a
member of either subgenus bearing a superficial resemblance to the other due to
convergent evolution, or perhaps a taxon of polyphyletic origin via some past
hybridization event between the two subgenera. Additional evidence relevant to
these interpretations is presented in the discussions of flavonoid chemistry and
chromosome number.
A general comparison of Nymphaea subg. Hydrocallis seeds with the other
subgenera observed implicates either Group I or Group II as the most primitive
seed type within subg. Hydrocallis. Both groups display characteristics such as
sclereid shape, papillae arrangement, spatial distribution of papillae, and fine
surface topography in common with members of these other subgenera. With N.
rudgeana seeds of Group I already discussed, seeds of Group II seem to most
closely approximate the ancestral seed type in Nymphaea. Seeds of Group Ilia
resemble this hypothetically ancestral type, but differ in their more irregular
FIG. 9. Seed of Nymphaea lotus of Nymphaea subg. Lotos, a. SEM of whole seed (Wiersema
1992). b. Light micrograph of seed coat. Note virtual absence of pit canals, c. Light micrograph of
seed coat section. Note thickness of inner integument and relatively thin outer tangential wall of
surface sclereids. Horizontal bar = 0.1 mm, vertical bar = 20 urn.
arrangement of surface sclereids and granulate fine surface topography. This lat
ter feature, although it seems to be restricted within Nymphaea to subg. Hydro
callis, bears a strong resemblance to the cuticular papillae reported by Schneider
and Ford (1978) on seeds of the closely related genus Ondinea and to the granu
late surface observed by Collinson (1980) on Brasenia Schreber seeds. If one
assumes Conard's phylogenetic scheme for Nymphaea to be reasonably accurate,
a further examination of seeds of subgenera Brachy ceras, Lotos, and Anecphya
would be helpful in defining the primitive vs. derived status of this character
within Nymphaea.
The trend toward smaller seeds with shorter and more randomly distributed
hairs in Groups Illb and IIIc, however, would have to be considered advanced
within the subgenus. Other probable derived characters are the clumped distribu
tion of hairs and the more isodiametric surface sclereid shape evidenced in Group
IV seeds.
A final note should be made regarding the absence of two other species of
Nymphaea subg. Hydrocallis from this phase of investigation. These are N. belo
phylla and N. prolif?ra, both excluded due to a lack of available seed material at
the time of the initial study. In the former case this could be attributed to the
scarcity of collections of this species; however, in the latter case it was due to a
total preponderance of asexual reproduction, thereby circumventing normal seed
production, in all populations encountered thus far. On the basis of other mor
phological considerations, N. belophylla would be expected to display seeds akin
to those of Group IV. It is now possible to characterize seeds of N. prolif?ra,
however; those produced from a recent artificial hybridization display the charac
teristics of Group Ilia, as predicted by Wiersema (1984b).
POLLEN MORPHOLOGY
The pollen of Nymphaea has been described as both monosulcate or monopo

rate. The single aperture may be interpreted as an encircling band or ring (mono
sulcate) or as an expanded pore (monoporate). The first view is held by Walker
(1974) and the latter is maintained by Jones and Clarke (1981). In actuality both
views may be correct, as the encircling sulcus is believed to have arisen from the
expansion of a single aperture, followed by the development of a lid or operculum
from a thickening of the aperture membrane. In Nymphaea the operculum almost
completely fills the aperture, leaving the ringed sulcus at its margin.
Walker (1974) proposes that the pollen of Nymphaea represents a primitive
pollen type within the ranalian complex. He examined pollen of N. candida Presl
and N. mexicana of subg. Nymphaea and N. lotus of subg. Lotos. All three
displayed a bandlike sulcus, the plane of which was believed to be perpendicular
to the axis of the pollen grain from the center of its tetrad. The fact that the sulcus
is indeed perpendicular to this axis has been verified by Jones and Clarke (1981)
in tetrads of N. alba pollen. Walker (1974) classifies Nymphaea pollen as repre
senting an evolutionary progression from more primitive anazonasulculate types,
with the ringlike furrow closest to the distal pole of the grain (relative to its tetrad
position), to zonizonasulculate forms, with the ring positioned at the equator of
the grain. Zonasulculate pollen in general is believed to be derived from a single
distal sulcus (anasulcate); therefore, those pollen types exhibiting this aperture
ring closest to the distal pole should be the most primitive. Nymphaea candida is
thus considered by Walker to possess the primitive pollen type in Nymphaea and
N. lotus a more advanced type.
In a related paper, Walker and Doyle (1975) present SEM pollen micrographs
of N. viol?cea Lehmann (which Conard treats as a variety of N. gigantea Hooker)
of subg. Anecphya, which displays zonizonasulculate pollen, and "N. amazonum"
of subg. Hydrocallis. The latter more likely depicts pollen of N. odorata of subg.
Nymphaea, as it displays the ornamentation typical of that species and is clearly
not representative of Hydrocallis pollen.
Several other workers have examined pollen of Nymphaea. Ueno and Kitagu
chi (1961) studied the wall structure of pollen of several Nymphaeaceae, including
N. tetragona Georgi of subg. Nymphaea. A verrucate ektexine (the outer wall)?
as was apparent to the workers cited above in N. alba and N. candida?was
evident in pollen of this taxon. The exine ornamentation of N. tetragona was also
observed by Meyer (1964), who further examined pollen of N. dentata Schu
macher [=N. lotus] of subg. Lotos and N. stellata Willdenow of subg. Brachy
ceras. These latter species clearly possess the zonizonasulculate type of pollen
grain and generally lack the ornamentation present on pollen of subg. Nymphaea.
According to Caspary (1888), their respective subgenera both have smooth-sur
faced pollen, as does subg. Hydrocallis. Wodehouse (1935) examined pollen of N.
amazonum of subg. Hydrocallis and indicated it to be smooth as well. Others who
have studied Nymphaea pollen include Kuprianova (1976), who examined several
species of subg. Nymphaea, and Chaturvedi (1974), who examined pollen of a
horticultural form of N. stellata.
The pollen grains of subg. Hydrocallis possess the distinctive bandlike aper
ture characteristic of other Nymphaea pollen. According to Walker's (1974) classi
fication of zonasulculate pollen, they vary from anazonasulculate to zonizonasul
culate depending on the amount of expansion or compression of the operculum.
In subg. Hydrocallis the operculum bears a doughnut-shaped ring around its
perimeter. This ring seems to be produced from the constriction of the inner
portion of the operculum, with the dilated outer region delimited by a shallow
furrow. A puckered surface may develop over this constricted portion; otherwise,
the pollen grain is completely smooth. In overall shape it is reminiscent of a tiny
hamburger bun.
Only slight differences are detectable among the various species of subg.
Hydrocallis with regard to pollen morphology (Figs. lOa-i, lla-c). Differences
exist mainly in pollen size and in the extent of demarcation of the doughnut
shaped ring. This ring is fairly conspicuous in all species but is only slightly
evident in N. jamesoniana (Fig. lie). Pollen size is greatest in N. lingulata, N.
rudgeana, and N. lasiophylla, in decreasing order, with the other species observed
being fairly uniform in size. Confident determinations of most species of subg.
Hydrocallis are not possible, and little illumination of phylogenetic relationships
within the subgenus is provided from pollen analysis.
The greatest utility of pollen analysis has been in assessing the relationship of
subg. Hydrocallis to other subgenera of Nymphaea. As already indicated, the
pollen of subg. Nymphaea is highly ornamented and anazonasulculate. Pollen of
N. odorata (Fig. lid) and N. mexicana (Fig. lie) exemplify this morphology,
which appears very different from that of subg. Hydrocallis. In pollen of N.
mexicana, however, the operculum is patterned in a manner similar to that found
in subg. Hydrocallis. The inner portion is comparatively smooth, but along its
margin a dilated verrucate band is observed which may be homologous to the
doughnutlike ring evident in subg. Hydrocallis. This ring is clearly evident from
an examination of pollen of N. lotus of subg. Lotos (Fig. llf). Pollen of this
species also displays a smooth exine like that of subg. Hydrocallis. Pollen of subg.
Brachyceras, as evidenced in N. ampia (Fig. llg), N. elegans (Fig. llh), and N.
capensis Thunberg (Fig. Hi) shows no evidence of this ring whatsoever. The
relatively uniform pollen morphology exhibited by these three species and other
members of subg. Brachyceras that have been reported is characterized by a
rather broad equatorial furrow (zonizonasulculate), virtually identical hemispheric
portions, and a slightly roughened or puckered surface. Pollen of subg. Anecphya
was not available for examination but from published accounts (Walker & Doyle
1975) appears identical to pollen of subg. Brachyceras.
Pollen of subg. Lotos appears most similar to that of subg. Hydrocallis. The
close similarity observed between the two subgenera suggests a common origin for
this pollen type. Additional evidence for such a viewpoint will be discussed under
phylogeny.
CHROMOSOME NUMBERS
Published chromosome counts on other species of Nymphaea have indicated a
base number of x = 14 for the genus. The majority of counts were obtained by
FIG. 10. Scanning electron micrographs of pollen of nine species of Nymphaea subg. Hydrocallis.
a. N. novogranatensis (Wiersema & Gonz?lez 2234). b. N. glandulifera (Standley 73128). c. N. oxype
tala (Fern?ndez 1299). d. N. tenerinervia (Wiersema et al. 2307). e. N. lasiophylla (Wiersema et al.
2317). f. N. lingulata (Wiersema et al. 2308). g. N. rudgeana (Wiersema & Gonz?lez 2205). h. N.
amazonum subsp. amazonum (Wiersema 1400). i. N. prolif?ra (Asplund 16024). Bar = 5 \im.
Langlet and S?derberg (1927) and Gupta (1978, 1980); at the time of Gupta's
(1980) report counts were available for 22 of the ca. 40 species. Among the
apocarpous groups of Nymphaea, counts of 2n = 28, 56, and 84 have been
reported within subg. Brachyceras and 2n = 224 for the monotypic subg. Anec
phya. Within the syncarpous line counts of 2n = 56, 84, and 112 (and a number of
aneuploid counts: 64, 96, 105, 120, 160) have been reported for subg. Nymphaea,
and 2n = 28, 56, 84, and 112 for subg. Lotos. This evidence suggests that all other
FIG. 11. Scanning electron micrographs of pollen of subgenera Hydrocallis, Nymphaea, Lotos,
and Brachyceras of Nymphaea. a-c. Pollen of subg. Hydrocallis. a. N. conardii (Wiersema & Gonz?lez
2214). b. N. gardneriana (Wiersema & Gonz?lez 2210). c. N. jamesoniana (Wiersema 2031). d-e.
Pollen of subg. Nymphaea. d. N. odorata (Wiersema 784). e. N. mexicana (Wiersema 281). f. Pollen of
N. lotus of subg. Lotos (Wiersema 1442). g-i. Pollen of subg. Brachyceras. g. N. ampia (Wiersema &
Gonz?lez 2204a). h. N. elegans (Wiersema 1981). i. N. capensis (Robinson et al. 44). Bar = 5 um.
subgenera appear to share this base number. No species of subg. Hydrocallis had
been previously counted.
Successful mitotic counts were obtained from 24 populations, representing 11
of the 14 recognized species of subg. Hydrocallis and 2 probable hybrids. One
putative hybrid population (Wiersema et al. 2296) failed to yield an acceptable
count due to insufficient material. Those counts obtained are indicated in Appen
dix 1. Consistency of counts was observed for all populations within each species.
On the basis of the observed counts some obvious relationships within the
subgenus become apparent. From information already discussed the base number
within Hydrocallis would be expected to be x = 14. Diploid counts of 2n = 28 for
N. jamesoniana, N. conardii, N. gardneriana, and N. novogranatensis, which are
documented in Figure 12a-d, support such an assumption. Seeming to contradict
this evidence is the existence of 2n = 18 in N. amazonum, N. prolif?ra, N.
lasiophylla, and N. lingulata (Fig. 12e-h).
The disparity in the two chromosome lines is believed to be the result of a
series of aneuploid alterations, thereby reducing the haploid number from 14 to 9.
Similar reductions in chromosome number have been observed in other genera of
both plants (Stebbins 1971) and animals (White 1973). Such alterations, which
have been termed "centric fusions" (White 1973) or "Robertsonian fusions"
(Jones 1976), can be achieved via unequal reciprocal translocations, resulting in
the formation of a larger metacentric chromosome from two smaller acrocentric
ones, with subsequent fixation of this translocation in the homozygous condition.
Five stepwise fusions of this type would be required to account for the aneuploid
shift observed between the two chromosome lines of subg. Hydrocallis. The dif
ference in relative size of the chromosomes between the two groups supports such
an interpretation. In the 2n = 28 line, all chromosomes are of similar size, as
apparent in Figure 12a-d. The 2n = 18 group, however, exhibits 10 large and 8
smaller chromosomes (e.g., Fig. 12g, h), the ratio expected from five haploid
fusions. The existence of an intermediate count of 2n = 20 for N. tenerinervia
(Fig. 12i), a species which demonstrates morphological and phytochemical ties to
both chromosome lines, also agrees with this interpretation. In N. tenerinervia 8
large and 12 smaller chromosomes are observed, in conformity with the expecta
tion that four fusions would have been required in this species.
An alternative explanation would reverse the proposed direction of chromo
some evolution, deriving the 2n = 28 line from the more primitive 2n = 18 line
through a series of aneuploid increases. This would suggest a primitive status for
subg. Hydrocallis within Nymphaea, as all other subgenera appear to be based on
what would then be a more advanced 2n = 28 number. Such an interpretation is
not supported by the preponderance of morphological and anatomical evidence in
the genus.
Two other species of subg. Hydrocallis yielded distinctive chromosome num
bers. Nymphaea rudgeana is believed to have 2n = 42. Although numerous
chromosome squashes have been prepared of this species, the observation of
actively mitotic cells has proven more difficult than in other species. As only a few
such cells in the proper configuration have been observed, some reservation ac
companies this reported count.
The suggestion put forth in the discussion of seed morphology that N. rudgeana
is somehow related to subg. Lotos is not contradicted by the 2n = 42 count.
Members of subg. Lotos have thus far yielded counts of 2n = 28, 56, 84, and 112. A
rare hybridization event involving an n = 28 Lotos individual and an n = 14
Hydrocallis parent would result in an individual with 42 chromosomes. Additional
karyological research, particularly with meiotic preparations, or enzyme electro
phoretic study is needed to further explore this possibility. Although numerous
hybrids, both natural and artificial, have been documented in Nymphaea, none are
believed to exist from crosses involving two subgenera (Conard 1905; Wood 1959).
FIG. 12. Chromosomes of nine species and one subspecies of Nymphaea subg. Hydrocallis. a-d.
2n = 28. Note similar sizes of chromosomes, a. N. jamesoniana (Wiersema etal. 2277). b. N. onardii
(Wiersema & Gonzalez 2214). c. N. gardneriana (Wiersema 1403). d. N. novogranatensis (Wiersema &
Gonz?lez 2234). e-h. 2n = 18. Note 10 larger and 8 smaller chromosomes, especially in g and h. e. N.
amazonum subsp. pedersenii (Wiersema et al. 2270). f. N. prolif?ra (Wiersema et al. 2273). g. N.
lasiophylla (Wiersema et al. 2293). h. N. lingulata (Wiersema et al. 2308). i. 2n = 20, N. tenerinervia
(Wiersema et al. 2307). j. 2n = 84, N. oxyp?tala (Wiersema & Gonz?lez 2232). Bar = 5 um.
The other species of subg. Hydrocallis examined, N. oxyp?tala, appears to be

a hexaploid derivative of n = 14 with 84 chromosomes (Fig. 12j). With the
possible exception of N. rudgeana, N. oxypetala was the only polyploid species
observed. The extreme divergence in chromosome number manifested by this
species is paralleled by its morphological distinctiveness. On the basis of seed
morphology and general morphology, N. oxypetala appears related to N. pota
mophila, N. belophylla, and N. glandulifera; however, none of these three species
has been examined karyologically.
Chromosome counts were also obtained from two Argentine populations
which are suspected hybrids involving N. prolif?ra and N. amazonum subsp.
pedersenii. Both populations yielded triploid counts of 3n = 27 (Appendix 1).
These populations propagate via the same mode of asexual reproduction as N.
prolif?ra but exhibit a morphology somewhat intermediate between the two
hypothetical parents. Several other similar but distinguishable asexual forms ex
ist in this region, which may also prove to be triploid. Indeed, one such form
has been so determined by Camilo Luis Quarin (pers. comm.) at CTES (Appen
dix 1). On the basis of their distinct morphologies and greater resemblance to N.
prolif?ra, it is believed that these triploids may have originated from several
independent hybridization events with N. prolif?ra serving as the diploid
partner.
FLORAL BIOLOGY AND ARTIFICIAL HYBRIDIZATION
A wealth of information has been gathered on the floral or pollination biology

of Nymphaea and the Nymphaeaceae. Early observations and experiments are
surveyed by Meeuse and Schneider (1980). Noteworthy among these studies are
the series of papers by Schmucker (1932,1933, 1935), who studied several tropical
species from the Eastern Hemisphere. The pollination syndrome of Nymphaea, as
studied chiefly in the diurnally flowering subgenera Anecphya, Brachyceras, and
Nymphaea, was summarized by Schneider (1979). Flowers are generally protogy
nous, although some species have been reported as homogamous. First-day
flowers generally open slightly less than on later days, the floral appendages
spreading to permit access to the fluid-filled stigma. The tips of these appendages
form a narrow, often slippery, vertical passage to the stigma. Small insect visitors
may become trapped in the flowers and drowned in the stigmatic fluid. This fluid
reportedly contains a powerful surfactant which may be secondarily responsible
for these drownings but is primarily important in washing pollen from the bodies
of insect visitors (Meeuse & Schneider 1980). Insect pollinators are attracted by
the color and odor of flowers and include mostly Hymenoptera (especially Halicti
dae), although Cole?ptera or D?ptera may also be present.
Anther dehiscence occurs in second-day protogynous flowers, with the stig
matic fluid drying up at or by this time. Insects are generally free to forage for
pollen due to the dryer nature of the stigma and the incurving of the inner
stamens and carpellary appendages. Pollen-covered insects which subsequently
visit first-day flowers effect cross-pollination. Observations on N. mexicana (Cap
perino & Schneider 1985), N. odorata (Schneider & Chaney 1981), N. elegans
(Schneider 1982a), and N. gigantea (Schneider 1982b) generally support this type
of pollination syndrome. The flowers of these species open for two, three, three,
and four successive days, respectively.
Much less information exists on the floral biology of the night-flowering
subgenera Lotos and Hydrocallis. Other than brief observations on certain mem
bers of subg. Lotos (Conard 1905; Mauve 1967; Mitra & Subramanyam 1982;
Wiersema 1982) most of the data relate to subg. Hydrocallis. Field observations
of floral biology have been made on N. rudgeana, N. glandulifera, and N.
amazonum (Cramer et al. 1975; Prance & Anderson 1976; Prance 1980). All
observations indicate a cantharophilous pollination syndrome involving scarab
beetles of the genus Cyclocephala Latreille, including C. verticalis Burmeister,
C. castanea Olivier, and C. mollis Endr?di. The beetles have been reported to
feed on the highly developed carpellary appendages in subg. Hydrocallis (Prance
1980; Meeuse & Schneider 1980). Schneider (1982a) suggests an evolutionary
relationship between the beetle-pollination syndrome and the presence of these
enlarged clubshaped appendages. The lack of observation on the presence of
stigmatic fluid in subg. Hydrocallis has further led Meeuse and Schneider (1980)
to speculate on the possible absence of this fluid in these beetle-pollinated
flowers. Some question also remains whether beetles are indeed trapped inside
flowers during the day; some observers (Cramer et al. 1975) report they are and
others (Prance & Anderson 1976; Prance 1980) insist they can readily escape but
may passively remain.
Numerous crosses have been conducted within Nymphaea. These have been
made largely for horticultural purposes and have not involved members of subg.
Hydrocallis (Henkel et al. 1907; G. H. Pring 1934a, 1934b; G. W. Pring 1960).
Countless horticultural and several natural hybrids are known to exist within
Nymphaea (Wood 1959).
Floral Biology
Successful flowering in cultivation was achieved with eight species (N. amazo
num, N. conardii, N. gardneriana, N. jamesoniana, N. lingulata, N. novogranaten
sis, N. prolif?ra, N. tenerinervia); Walter Pagels (pers. comm.) supplied additional
data on a ninth, N. glandulifera. Flowers usually open for two consecutive nights
(sometimes three in N. jamesoniana). All exhibit protogamy except N. lingulata,
in which the anthers were observed to dehisce the first night. A pool of stigmatic
fluid is present in all first-day flowers, as shown in Figure 13a.
Differences in floral behavior among the taxa are apparent, particularly in
times of opening and closure of first-day flowers and closure of second-day
flowers. All second-day flowers observed commence opening at dusk, regardless
of species, except those of N. lingulata which open 1-2 hours later. On the basis
of similarities in floral responses, certain species groupings are apparent. One
includes N. glandulifera, N. conardii, N. gardneriana, and N. jamesoniana.
Flowers of this group open and close in synchrony. All these species, except N.
gardneriana, exhibit abundant autogamous seed production. Two other species,
N. novogranatensis and N. tenerinervia, are also similar in their opening and
closing times for flowers both nights. Autogamous seed production was observed
in some flowers of N. novogranatensis but in none of N. tenerinervia. Flowers of
N. lingulata, N. prolif?ra, and N. amazonum are not similar to other species in
FIG. 13. Night-blooming flowers of Nymphaea subg. Hydrocallis in cultivation, a. Fi

flower of N. amazonum subsp. amazonum x N. lasiophylla (Wiersema et al. 2296). Note
stigmatic fluid in center of flower, b. First-day flower of N. novogranatensis (Wiersema &
2234). c. Second-day flower of N. novogranatensis (Wiersema & Gonz?lez 2234). d. First-day f
N. lingulata (Wiersema et al. 2308). Bar = 2 cm.
their flowering times. Those of N. amazonum appear to be the most unusu

this regard. In addition to their unusual flowering schedule, anther dehisc
second-day flowers is delayed until well past midnight, in contrast to other
whose anthers dehisce when the flowers open. Autogamous seed production
observed in N. lingulata and N. amazonum subsp. amazonum but not
amazonum subsp. pedersenii or N. prolif?ra.
Unlike the four species of the first grouping, all other species possess fl
which remain open at least 1-2 hours after midnight, with some not closing
dawn. In this regard N. novogranatensis, N. tenerinervia, N. amazonum, N
lata, and N. prolif?ra display some relation to one another. Published o
FIRST-DAY FLOWERS SECOND-DAY FLOWERS
?- JAMESONIANA, CONARDII, GARDNERIANA, GLANDULIFERA

NOVOGRANATENSIS, TENERINERVIA
? AMAZONUM
PROLIFERA
? LINGULATA
. VARIATION DUE TO SEASONAL OR OTHER FACTORS
CDT = CENTRAL DAYLIGHT TIME
FIG. 14. Flower opening times for first- and second-day flowers of nine species of Nymphaea
subg. Hydrocallis.
tions on N. rudgeana (Caspary 1878; Conard 1905) support its inclusion in this
latter group as well.
With regard to those species of subg. Hydrocallis observed, a floral clock
(Fig. 14) can be constructed similar to that produced by Conard (1905) for Nym
phaea. Although some degree of temporal separation is apparent between flowers
of the two groups just mentioned, only in N. amazonum is complete temporal
isolation approached due to the delayed pollen release in second-day flowers.
However, first-day flowers of N. amazonum could still be pollinated with pollen
of other species whose second-day flowers remain open until dawn. The potential
for hybridization between most species thus is not greatly affected by flowering
time.
Several items relevant to pollination are also apparent. One concerns differ
ences in floral odors between various species. These differences might relate to
the attraction of different species of pollinators, an interesting speculation put
forth by Meeuse and Schneider (1980) which demands further investigation.
Flowers of the pre-midnight flowering group generally produce a stronger and
more pleasant odor. Considerable odor is emitted from an entire population, as
the latter can readily be detected from a car passing by at night. Another differ
ence among species relates to the degree of opening of flowers. As a rule, first
day flowers open internally only enough to allow a small passage, ca. 1 cm or so
in diameter, to the fluid-filled stigma (Fig. 13b, d). In flowers of Nymphaea
conardii and N. gardneriana, however, no such passage was perceived. In second
day flowers the inner perianth and stamens of N. novogranatensis and N. teneri
nervia were frequently observed in a closed position at the time first-day flowers
were open. Presumably an insect visitor in these latter species would be required
to force its way into the floral interior in its foraging, thus ensuring a liberal
dusting with pollen. In both species the inner petals are greatly reduced and
would offer little resistance to such an entry (Fig. 13c).
Insect visitors to flowers of subg. Hydrocallis have been observed and cap
tured on only two occasions in the field. Cydocephala verticalis was collected
from the type locality of N. conardii in Venezuela and C. aff. p?trida Burmeister
(males were needed for positive identification according to B. C. Ratcliffe, pers.
comm.) from a population of N. lasiophylla in northeastern Brazil. This latter
Cydocephala, although requiring further identification, differs from those previ
ously reported as pollinators of subg. Hydrocallis. Further evidence of beetles
feeding on carpellary appendages has been detected in N. oxypetala flowers of
Venezuela, although no beetles were observed at that time.
Adaptations for autogamy are observable in certain species. In some, such as
N. conardii and N. jamesoniana, the stigmatic fluid persists in second-day flowers
and the stigma remains receptive to the pollen now released. Although the stigma
is likewise covered with fluid in second-day N. gardneriana flowers, it is no longer
receptive, as evidenced from the failure of all crosses attempted on that day. Two
other species appear to be approaching cleistogamy. In N. jamesoniana, flowers
frequently fail to open the first night and are therefore effectively homogamous.
Those of many populations, as in Ecuador and Florida, open only slightly, the
outer parts spreading only to a vertical position. Autogamous seed production is
abundant. In N. lingulata flowers are also homogamous. They, too, open only
slightly, as the thick, coriaceous sepals prevent the petals and stamens from
further reflexing (Fig. 13d). Autogamous seed production is also evident. No
seeds were produced from any emasculated flowers of subg. Hydrocallis in the
absence of cross-pollination, which indicates autogamy and not agamospermy is
responsible for seed production in unemasculated flowers under these conditions.
Crossing Studies
The pollen fertility of the eight species examined was presented in Wiersema
(1984b). With the exception of a Brazilian population which exhibited less than
5% viability, all other populations have pollen fertility exceeding 75%. This Bra
zilian population is clearly of hybrid origin, probably involving N. lasiophylla and
N. amazonum subsp. amazonum.
An assortment of crosses was conducted among the various species; see table
1. Unfortunately most crosses were conducted among members of the 2n = 28
chromosome line?N. conardii, N. gardneriana, N. jamesoniana, and N. novo
granatensis?as more flowers of this group were available in cultivation. All of
these species exhibited a moderate degree of interfertility. Crosses between this
group and the 2n = 18 line were very limited in number, but only N. novograna
Table 1. Percentage seed set from artificial crosses among several species of Nymphaea subg. Hydrocal
Nymphaea.
Pollen parent
N. N. N. N. N. N. N
Seed parent amazonum conardii gar
Ia
N. amazonum
2 1 2 1
N. conardii 10-20 60 20-50 40
3 1 3
N. gardneriana 20-40 80 20-50
N. jamesoniana
N. lingulata
1 1
N. novogranatensis 30 <10 40-80
1
N. prolif?ra <5
N. rudgeana
1
N. tenerinervia 20
N. lotus
aNumber of crosses. bEstimated percentage of normally developed seeds.
tensis, the most profusely flowering species in cultivation, has indicated any de
gree of interfertility with this other line. Interestingly, some degree of seed set
was observed in crosses involving N. novogranatensis and most other available
species of subg. Hydrocallis, which supports a central phylogenetic position for
this species within the subgenus. The lowest interfertilities recorded for N. novo
granatensis involve crosses with N. lingulata and N. jamesoniana, two species
which may occupy more peripheral positions in the phylogenetic tree. No seed set
was observed from any crosses of species of subg. Hydrocallis with those of other
subgenera. Most involved N. lotus of subg. Lotos, although none of these crosses
included N. rudgeana, which is presumed to bear some relation to this other
subgenus.
FLAVONOID CHEMISTRY
Mostly fragmentary information exists on the flavonoid composition of nym

phaeaceous plants. In their survey of flavonoid compounds in the angiosperms,
Gornall et al. (1979) report seven different classes of flavonoids as having been
reported for the Nymphaeaceae. Among these are delphinidin, cyanidin or pelar
gonidin, quercetin or kaempferol, myricetin, luteolin or apigenin, C-glycoflavones,
and proanthocyanidins. Some of this information was accumulated by Bate-Smith
(1963), who included Nuphar lutea (Linnaeus) Smith and Nymphaea capensis in his
report on phenolic constituents of various plants. He reported myricetin, delphini
din, ellagic acid, caffeic acid, p-coumaric acid, sinapic acid, and ferulic acid to be
present in Nymphaea capensis leaves and ellagic acid and perhaps luteolin in Nu
phar lutea leaves. The presence of ellagic acid and luteolin in Nuphar lutea was later
confirmed and the other four phenolic acids were also detected (Bate-Smith 1968).
Other information stems from a survey of anthocyanins in Nymphaea candida, in
which galactosides of both delphinidin and cyanidin were reported (Bendz & Jons
son 1971). In addition, luteolin 8-C-?-D glucopyranoside (orientin, a C-glycosyl
flavone) was detected in the leaves of Nymphaea alba (Taku et al. 1970) and a
glycoside of pelargonidin was reported from the flower of Victoria regia Lindley
[=V. amaz?nica (Poeppig) Sowerby] (Miele et al. 1969, cited by Harborne et al.
1975).
In another more inclusive study of the flavonoids of Nymphaeaceae, Wohlfart
and Gademann (1974) examined both the leaves and shoots of Victoria amaz?n
ica, Euryale ferox Salisbury, and five species of Nymphaea: N. alba, N. capensis,
N. caerulea Savigny, N. lotus, and N. rubra Roxburgh. Several flavonols were
identified for the Nymphaea species, including gossypitrin, rhamnetin, myricetin,
kaempferol-4-methyl ether, quercetin-3',4',5,7-tetramethyl ether, quercetin,
and glycosides of quercetin. The presence of several additional unidentified com
pounds was also noted, including one designated as "Flavone A." A multitude of
compounds was reported from Victoria, which included gossypitrin, gossypetin,
gossypetin-6-methyl ether, myricetin, rhamnetin, tricin, kaempferol-4-methyl
ether, quercetin tetramethyl ether, and several flavones including Flavone A. For
Euryale, gossypitrin, kaempferol-4-methyl ether, robinetin, Flavone A, and sev
eral additional flavones were documented.
A considerable number and diversity of flavonoid compounds were en

countered among the species examined. Some 56 compounds were isolated and
compared among the various populations studied, with 53 of these proving to be
flavonols or flavones. Their distribution among the 54 Nymphaea populations
studied is indicated in Appendix 2. Although C-glycosyl flavones of apigenin and
luteolin were included among these, the vast majority (44) were O-glycosides of
quercetin, myricetin, or kaempferol. Nine glycosides of quercetin, 8 of isorham
netin (3'-methylquercetin), 3 of rhamnetin (7-methylquercetin), 2 of 3-methyl
quercetin, 1 of rhamnazin (3',7-dimethylquercetin), and one believed to be of
3,3'-dimethylquercetin were found along with 7 glycosides of myricetin, 4 of 3
methylmyricetin, and 5 of kaempferol. Certain of these glycosides, in particular
those of isorhamnetin and kaempferol, may actually represent acylated deriva
tives of other glycosides. Although hydrolysis of these provided identical agly
cones and sugars, the glycosides were consistently separated chromatographically
on polyamide TLC. Identical UV spectra were obtained from such compounds.
From their spectral differences and the results of hydrolysis, at least four addi
tional glycosides appear to be acylated with aromatic acids (see Harborne et al.
1975). Among these are glycosides of kaempferol, quercetin, and myricetin.
Structures of all identified compounds are presented in Figure 38 of Appendix 2.
Of the flavonols in combination with monosaccharides, 13 involved galactose
(many of these probably acylated), 10 glucose, 7 rhamnose (a few of these per
haps acylated), 5 xylose, and 1 arabinose. Disaccharides of glucose (4) and ruti
nose (3) were also found in combination with flavonols. Although the majority of
glycosides were 3-O-glycosides, three 7-O-glycosides, one 4'-0-glycoside, a
3',5'-0-diglycoside, and several 3'-0-glycosides were identified. Glucose was
found at all of the above positions on the flavonol ring system; however, xylose
was present only at the 3' position, with the other sugars restricted to the 3
position.
Intraspecific variation. Phenograms from the two cluster analyses, comparing
a) the distribution of all compounds and b) the distribution of the various agly
cones and glycosidic combinations among the 54 populations examined, are pre
sented in Figures 15 and 16, respectively. From an inspection of the two pheno
grams it is evident that for some species the flavonoid profile was quite consistent
among the various populations examined. Within Nymphaea subg. Hydrocallis
this is particularly true of N. jamesoniana, N. prolif?ra, N. rudgeana, N. novo
granatensis, and N. tenerinervia, for which three or more populations were exa
mined, and appears to be the case in N. lingulata and N. lasiophylla, although
only two populations were examined in the latter cases.
Interpopulational differences were manifest in other species of subg. Hydro
callis, however, in particular N. amazonum, N. gardneriana, and N. conardii, but
could not be assessed for N. oxypetala, for which only one population was sam
pled. In four of the six N. amazonum populations examined the flavonoid profile
was fairly congruous. The two remaining populations differ in their possession of
3'-and 4'-glucosylated quercetips, two compounds which also occur in N. rudge
ana. These two populations, from southern Venezuela and Argentina, share cer
tain morphological features which distinguish them from other representatives of
VOUCHER COUNTRY
0.25
_J_
0.55
>
0.85 No.a TAXON OF ORIGIN
AMP LA Mexico
GARDNERIANA Argentina
JAMESONIANA Mexico
JAMESONIANA Ecuador
JAMESONIANA Florida, USA
JAMESONIANA Argentina
JAMESONIANA Brazil
GARDNERIANA Venezuela
MEXICANA Alabama, USA
AMP LA Florida, USA
AMP LA Ecuador
OXYPETALA Venezuela
ELEGANS Florida, USA
ODORATA Alabama, USA
CANDIDA West Germany
LINGULATA Brazil
LINGULATA Brazil
LOTUS Florida, USA
AMAZONUM Argentina
RUDGEANA Venezuela
RUDGEANA Brazil
RUDGEANA Brazil
AMAZ. X LAS10. Brazil
AMAZONUM Ecuador
TENERINERVIA Brazil
AMAZONUM Venezuela
AMAZONUM Ecuador
AMAZONUM Brazil
TENERINERVIA Brazil
TENERINERVIA Brazil
TENERINERVIA Brazil
PROLIFERA Ecuador
PROLIFERA Argentina
PROLIFERA Argentina
PROL. X AMAZ. ? Argentina
PROLIFERA Argentina
NOVOGRANATENSIS Venezuela
AMAZONUM Venezuela
LASIOPHYLLA Brazil
LASIOPHYLLA Brazil
CONARDII Venezuela
CONARDII Nicaragua
CONARDII Nicaragua
CONARDII Brazil
CONARDII Venezuela
Collection numbers are ofH.
J.Wiersema, except Nos. 8588 and 8612 which are
of R. R. Haynes.
FIG. 15. Phenogram derived from presence/absence data for 56 compounds isolated from
ions of Nymphaea.
VOUCHER COUNTRY
0.25 0.55 0.85 No.* TAXON OF ORIGIN
1345 AMPIA Mexico
-2021 AMPLA Ecuador
1452 AMPLA Florida, USA
274 ODORATA Alabama, USA
1981 ELEGANS Florida, USA
?2196 CANDIDA West Germany
'1403 GARDNERIANA Venezuela
1428 GARDNERIANA Venezuela
2247 GARDNERIANA Argentina
-1343 JAMESONIANA Mexico
-2030 JAMESONIANA Ecuador
?2032 JAMESONIANA Florida, USA
2265 JAMESONIANA Argentina
2290 JAMESONIANA Brazil
2232 OXYPETALA Venezuela
281 MEXICANA Alabama, USA
i-]1411 NOVOGRANATENSIS Venezuela
2234 NOVOGRANATENSIS Venezuela
2028 PROLIFERA Ecuador
2241 PROLIFERA Argentina
2238 PROL. X AMAZ. ? Argentina
?2240 PROL. X AMAZ. ? Argentina
Ecuador
2013 AMAZONUM Ecuador
E2027 AMAZONUM
2291 AMAZONUM Brazil
Brazil
2314 TENERINERVIA Brazil
E2307 TENERINERVIA
2311 TENERINERVIA
? 2310 TENERINERVIA
Brazil
Brazil
-1436 AMAZONUM Venezuela
? 2296 AMAZ. X LASIO. Brazil
-2270 AMAZONUM Argentina
-2221 AMAZONUM Venezuela
? 2205 RUDGEANA Venezuela
2285 RUDGEANA Brazil
2299 RUDGEANA Brazil
1442 LOTUS Florida, USA
2214 CONARDII Venezuela
8588 CONARDII Nicaragua
8612 CONARDII Nicaragua
2230 CONARDII Venezuela
2289 CONARDII Brazil
2293 LASIOPHYLLA Brazil
2316 LASIOPHYLLA Brazil
2283 LINGULATA Brazil
2308 LINGULATA Brazil
8588 Nos.
Collection numbers are of J. H. Wiersema, except and 8612 which are
of R. R. Haynes.
FIG. 16. Phenogram derived from presence/absence data for 11 aglycones and 9 glyc
linkages of 54 populations of Nymphaea.
N. amazonum. In particular, they produce flowers with a greater number of

stamens and leaves which are more elliptic, have a sparser arrangement of tricho
sclereids, and an abundance of red pigmentation. The Argentine population,
moreover, produces several flavonoid compounds not detected in the South
Venezuelan plants. Among these are the 3-rutinosides of quercetin and myrice
tin, both of which are present in N. prolif?ra of Argentina and several Argentine
triploids believed to be derived from N. amazonum and N. prolif?ra. The pres
ence of these additional compounds in the Argentine N. amazonum, in light of
differences in morphology and floral biology, supports the establishment of sub
specific rank for this group of plants.
Despite the variability in flavonoid compounds evident among the five popu
lations of N. gardneriana examined, some obvious similarities are apparent. All
produce quercetin 3-glucoside and 3-rhamnoside, the most common flavonoids
among the species sampled. In addition, plants of three populations produce
rhamnetin 3-rhamnoside, a compound not detected in any other Nymphaea. A
fourth population, although lacking this compound, produces two other rhamne
tin glycosides, the 3-glucoside and 3-rutinoside, which were not found elsewhere
in the genus. Although rhamnetin compounds were not detected in one N. gard
neriana population, when they are present they appear to be diagnostic for N.
gardneriana. A related compound, 3',7-dimethylquercetin (rhamnazin) 3-rham
noside, isolated from a single N. gardneriana population, was not detected in any
other Nymphaea population studied.
Five populations referable to N. conardii from Brazil, Venezuela, and Nicara
gua demonstrated a high degree of variability in flavonoid compounds produced.
Only one compound, vitexin, was common to plants from all three regions, al
though it was not detected in one of the Venezuelan populations. Both Nicara
guan populations were unique among the Nymphaea studied in producing kaemp
ferol 3-rhamnoside and an acylated derivative of this compound. A significant
quantity of quercetin 3-rhamnoside (quercetrin) was also produced by one of
these populations but was not detected from the second. Of the two Venezuelan
populations surveyed, one produced four different 3-galactosides of kaempferol,
isolatable due to differing Rfs on polyamide TLC, and the other produced four
galactosides of isorhamnetin. These differences are probably due to the effects of
acylation, which appear to be similar for the two populations. The presence of
only galactosides in these populations, whose plants cannot be distinguished mor
phologically, supports the assignment of both to the same species.
The Brazilian population was characterized by isorhamnetin 3-glycosides,
only one of these, a galactoside, in common with the isorhamnetin glycosides of
the Venezuelan population. Two isorhamnetin 3-rhamnosides, separable on the
basis of Rfs, and a 3-diglucoside were also isolated from the Brazilian plants. In
summary, the mutual presence of isorhamnetin derivatives links the Brazilian
population to one of the Venezuelan populations, the latter being linked to the
other Venezuelan population by their production of seemingly similarly acylated
galactosides. Plants of this second Venezuelan population are further linked to
the Nicaraguan plants in their coproduction of kaempferol.
While the pattern of variation just discussed and the absence of kaempferol
and isorhamnetin compounds from the other species of Nymphaea demonstrate
an association between the five populations of N. conardii, it could as well be
used to support the establishment of subspecific categories. This would seem to be

most appropriate for the Central American populations of N. conardii north of
Panama where subtle morphological distinctions are apparent, such as fewer
acicular sclereids in the stigmatic rays or the presence of reddish pigmentation in
the leaves.
Interspecific variation. The phenograms of Figures 15 and 16 also illustrate the
relationship among members of Nymphaea subg. Hydrocallis and their degree of
similarity to the other species of Nymphaea examined. Several characteristic
species groupings are evident in both phenograms. Among these is the close
relationship between N. jamesoniana and N. gardneriana. The basis for this simi
larity is their coproduction of quercetin 3-glycosides, the only compounds iso
lated from N. jamesoniana. Although these compounds are found in several other
species of Nymphaea, so are several additional compounds. Among N. gardneri
ana fewer other compounds are detectable, most of these glycosides of 7-methyl
quercetin (rhamnetin).
Another species which failed to group with these two but which yielded a
similar depauperate flavonoid profile was N. conardii. Its apparent dissimilarity
was due to an absence of quercetin glycosides, these having been replaced by
glycosides of 4'-methylquercetin (isorhamnetin) or kaempferol. However, the
lack of several major compounds produced by other species of Nymphaea sug
gests an association among these three species. The compounds generally absent
from this group include C-glycosyl-luteolins, three apparently nonflavonoid com
pounds, myricetin derivatives, and compounds with B-ring glycosylation. Small
quantities of orientin (a C-glycosyl-luteolin) and 3'-glucosides of both quercetin
and myricetin have been isolated from certain populations of N. gardneriana,
which suggests a less distant relationship between this species and other members
of subg. Hydrocallis.
A second group of species of subg. Hydrocallis which consistently clustered
together includes N. novogranatensis, N. prolif?ra, N. tenerinervia, and N. amazo
num. Similarities among members of this group include the absence of methyla
tion, little or no production of flavonol rhamnosides, and the presence of orientin
and the three previously mentioned nonflavonoid compounds. (Although two of
the latter three compounds show some evidence of being xanthones, they have
not been fully characterized and thus the three are designated A, B, and C.) With
the exception of N. novogranatensis, all produce one or more glycosides of my
ricetin as well. Although none of these features was, in itself, unique in Nym
phaea, their cumulative effect served to isolate this group, particularly from the
members of the previous group.
A third Hydrocallis grouping composed of two species is also distinguishable.
In terms of individual compounds produced the relationship between the two, N.
lingulata and N. lasiophylla, is almost totally obscured. Some of this apparent
dissimilarity is due to the presence of several unique compounds in N. lasiophylla.
Among these are what appear to be 3,5-dimethylated compounds. Although
these compounds proved to be rather unstable and thus difficult to identify with
certainty, their blue fluorescence and UV spectral data would suggest such a
structure. Nymphaea lasiophylla also produces another unique aglycone that,
from spectral data and the results of hydrolysis, has been tentatively identified as
3,3'-dimethylquercetin. In addition, several compounds which appear to be acy
lated with aromatic acids are also present in the exceedingly complex flavonoid
profile of N. lasiophylla.
The similarity between N. lasiophylla and N. lingulata is apparent from an
analysis of their aglycones and glycosidic combinations. Both are characterized by
the predominance of 3-methylated forms of quercetin and myricetin. Between
the two species, only in N. lasiophylla have nonmethylated forms of these agly
cones been detected. As a result of this 3-methylation, glycosylation in both
species is restricted to 7- and 3'-glucosides or 3'-xylosides. Glycosylation at the
7 position is unique to these two species among the representatives of Nymphaea
examined.
Although 3-methylation is lacking among other species of subg. Hydrocallis,
it is evident in several other Nymphaea species, including N. elegans and N. ampia
of subg. Brachyceras, and N. candida and N. odorata of subg. Nymphaea. All
four species and N. lotus of subg. Lotos characteristically produce 3'-xylosides as
well. This latter feature is not restricted to N. lasiophylla and N. lingulata within
subg. Hydrocallis, since it is also present in N. tenerinervia and in some popula
tions of N. novogranatensis (in low levels) and N. amazonum. Congruity of sev
eral 3-methylated compounds underlies the clustering of N. lingulata and mem
bers of subg. Brachyceras and subg. Nymphaea in Figure 15. The association of
N. lingulata with these taxa suggests a relationship that will be elaborated on in
the discussion of phylogeny.
The flavonoids of two additional species of subg. Hydrocallis remain to be
discussed. One of these is N. rudgeana, whose flavonoid profile is dominated by
C-glycosylflavones and quercetrin glycosides, including the 3'- and 4'-glucoside.
Both of these latter compounds are present in certain populations of N. amazo
num as already mentioned; the 4'-glucoside was not observed elsewhere in the
genus. With the exception of quercetrin, the most concentrated flavonoid in N.
rudgeana, each compound detected in N. rudgeana is also present in the Argen
tine N. amazonum subsp. pedersenii, indicating a possible affinity between these
two taxa. Flavonoid evidence does not, however, clarify the relationship between
N. rudgeana and N. lotus of subg. Lotos that was postulated from morphological
and seed evidence. Both similarities and differences exist in the flavonoid profiles
of the two species. Further phytochemical analysis of other members of subg.
Lotos could aid in resolving this dilemma.
The remaining species of subg. Hydrocallis examined, N. oxypetala, has a
flavonoid profile which combines elements of some of the other groups. The
absence of C-glycosylflavones and the presence of quercetrin align N. oxypetala
with species of the first group. However, the presence of Compounds A, B, and C
in N. oxypetala together with myricetin derivatives suggests a linkage with the
second group. Unfortunately three additional species of subg. Hydrocallis, N.
glandulifera, N. belophylla, and N. potamophila, which on morphological grounds
appear to be most closely allied to N. oxypetala, were not available for chemical
examination.
Hybridization. A number of putative hybrid populations were also examined.
All of these populations reproduced exclusively by asexual means and were as
sumed to be incapable of sexual reproduction. Chromosome counts obtained from
two such populations indicated them to be triploid. Both of these and three
additional populations, all from Argentina, cluster with N. prolif?ra, one of the
putative parents based on morphological considerations. This cluster is chemically

unique in the production of myricetin 3-arabinoside, the sole arabinoside found
in Nymphaea. Only one of these five populations displayed the B-ring glycosyla
tion characteristic of the Argentine N. amazonum subsp. pedersenii, the other
putative parent. Curiously, this population also appeared most similar in overall
morphology to N. amazonum subsp. pedersenii.
A sixth hybrid population, from northeastern Brazil, was most similar to N.
amazonum in its aglycone and glycosylation pattern. Several compounds of this
hybrid were not detected elsewhere, including what appeared to be acylated
forms of quercetin 3-galactoside. Interestingly, both quercetin 3'-mono- and
diglucosides were believed present. Although 3'-monoglucosides are present in
several other species of subg. Hydrocallis, only one species, N. lasiophylla, pro
duces what appear to be 3'-diglucosides as well. Although in N. lasiophylla these
are myricetin diglucosides, this similar pattern provides some evidence for N.
lasiophylla as the other parental type.
Novel or Unusual Compounds. Several flavonoid compounds present in the
taxa examined are noteworthy with respect to their uncommon occurrence in
other plant groups. Five of these compounds are related to one another in that all
are 3'-xylosides. According to information in Harborne and Williams (1975,
1982), the only flavonol 3'-xyloside reported prior to 1981 was quercetin
3'-xyloside, from a species of Euphorbia Linnaeus (Rizk et al. 1976). Some of
these 3'-xylosides are produced in large quantities in several species of Nym
phaea. One such compound is quercetin 3'-xyloside.
Myricetin 3'-xyloside was a major flavonoid compound in several species of
Nymphaea, including N. lotus of subg. Lotos and N. candida and N. odor ata of
subg. Nymphaea. This compound, however, was not reported by Wohlfart and
Gademann (1974) from their flavonoid study of N. lotus and N. alba, the latter
species a close relative of N. odor ata and N. candida. Instead, they reported
gossypitrin to be present in all of the species of Nymphaea they examined. Both
gossypitrin and myricetin 3'-xyloside appear as yellow compounds under UV
light and have similar Rf values in the common Chromatographie solvents. As the
structure of myricetin 3'-xyloside is clearly indicated from UV spectral data, the
results of acid hydrolysis, and NMR spectroscopy, this prior report of gossypitrin
is believed to be erroneous, since no evidence for this compound was observed.
The three other 3'-xylosides were associated with 3-methylation. One, 3
methylquercetin 3'-xyloside, was present in N. ampia, N. elegans, N. candida,
and N. odorata, and in N. lingulata of subg. Hydrocallis. Another, 3-methyl
myricetin 3'-xyloside, was detected in all of these species as well as in N.
lasiophylla. The third and most unusual compound of this type was 3-methyl
myricetin 3'-xyloside, 5'-glucoside, which was present only in N. lingulata.
This compound is glycosylated at two positions on the B-ring, an extremely
unusual structural feature.
Three additional compounds that had not been reported prior to 1981 also are
structurally related. The three are 3'-diglucosides of quercetin, myricetin, and 3
methylmyricetin. Analogous 3'-monoglucosides have been reported in Aesculus
Linnaeus (Wagner 1964), Cannabis Linnaeus (Rao & Seshadri 1941), and Oe
nothera Linnaeus (Howard & Mabry 1970). The class of compounds in question is
found only in N. lasiophylla and a suspected hybrid involving N. lasiophylla, the

quercetin derivative present only in the putative hybrid.
Two sets of compounds present in N. lasiophylla also appear to be unusual.
One is believed to be a 7-glucoside of 3,3'-dimethylquercetin, a glycoside not
previously reported. The other group of compounds appear to be, from their blue
fluorescence, spectral data, and Rfs, 3,5-dimethylated compounds. At least three
such compounds appear to be present, perhaps of quercetin, myricetin and 3'
methylquercetin, three aglycones known or suspected to occur in 3-methylated
form in this species.
Of interest are the numerous acylated glycosides suspected in several species.
Most of these appear to be galactosides, although a few involving rhamnose and
glucose have been detected. An additional methanol peak observed with UV
spectroscopy indicates that several compounds may be acylated with aromatic
acids, as the absorption peak of the acids is superimposed on the normal flavo
noid spectral bands (Harborne & Williams 1975). Acylated glycosides appear to
be present in several species of subg. Hydrocallis but were not observed in species
surveyed from other subgenera of Nymphaea.
PHYLOGENY
Position of subg. Hydrocallis within Nymphaea. The phylogeny developed by

Conard (1905) for Nymphaea (presented in the discussion of seed morphology)
placed subg. Hydrocallis within the syncarpous evolutionary line, which also in
cludes the subgenera Nymphaea and Lotos. Subgenus Hydrocallis and subg. Nym
phaea were believed to be derived from a common ancestral line from which
subg. Lotos had diverged at some earlier time. Members of subg. Hydrocallis
might thus be expected to more closely resemble subg. Nymphaea than subg.
Lotos. The pollen morphology of the three subgenera contradicts this assumption.
Pollen of subg. Nymphaea is considered to represent a more primitive type in the
genus, that of subgenera Hydrocallis and Lotos a more advanced type (see discus
sion of pollen morphology). The close similarity between pollen of subgenera
Hydrocallis and Lotos suggests a common origin for this derived pollen type. At
least three other seemingly advanced features within Nymphaea are shared by
subgenera Lotos and Hydrocallis but not by subg. Nymphaea. Such features are
nocturnal flowering, highly developed carpellary appendages, and medial posi
tioning of anthers on the stamens. The derived nature of these characters is ap
parent from comparisons with the apocarpous subgenera Anecphya and Brachy
ceras. Subgenera Nymphaea, Anecphya, and Brachyceras are diurnally flowering,
have less-developed carpellary appendages, and mostly marginal positioning of
anthers on the stamens. The states of these subgenera with regard to several
major characters are presented in Table 2.
A few features are shared by subgenera Nymphaea and Hydrocallis but not by
subg. Lotos, such as entire leaves, glabrous leaf blades, and lack of evident sepal
venation; however, at least the latter two probably represent ancestral states for
these characters. Thus a common ancestry excluding subg. Lotos would not be
required to explain their presence. In the case of the derived character states
Table 2. Character states of subgenera of Nymph
Anecphya Brachyceras Hydrocallis
Carpel fusion incomplete incomplete complete
Flowering time diurnal diurnal nocturnal
Carpellary absent slightly developed, well-developed,

appendages triangular-tapered usually cl?vate
Position of anthers
on outer stamens unknown marginal medial
Leaf margins dentate entire to dentate entire (dentate

in N. rudgeana)
Pubescence of absent absent absent

On
mature blades
Sepal venation inconspicuous inconspicuous inconspicuous
Rhizome erect erect erect
Pollen surface puckered slightly puckered smooth
Doughnutlike absent absent present

ring of operculum
Seed hairs present present present
Stolons absent absent often present
Peduncular 6 (5-7) 6 (5-7) 6 (5) or 18 (4

major air canals in N. oxypetala)
shared by subgenera Lotos and Hydrocallis, the hypothesis of a shared ancestry

excluding subg. Nymphaea would seem to represent the most parsimonious expla
nation, as four instances of convergent evolution would be dictated according to
Conard's model.
Relationships among species of subg. Hydrocallis. Chromosome counts have
indicated at least two major chromosomal lines within subg. Hydrocallis, corre
sponding to 2n = 28 and 2n = 18. The 2n = 28 line is considered ancestral,
because it is common to all other subgenera of Nymphaea as well. A number of
other character states can be postulated as being ancestral within subg. Hydrocal
lis from comparisons with other subgenera. With regard to phytochemistry, the
presence of myricetin compounds, B-ring glycosylation of flavonoids, orientin,
and the nonflavonoid compounds A, B, and C all appear to represent primitive
conditions. Seeds which exhibit rectangular surface sclereids, solitary hairs along
alternate cell junctures, and greater size would also be considered primitive. With
pollen, the presence of a well-defined doughnut-shaped ring on the operculum
appears to be ancestral. Flowers which remain open all night are believed to have
retained the primitive condition. Other characteristics which would be considered
more primitive are: a radiate leaf venation centrally, stolon formation, shorter
carpellary appendages, the presence of two or four major air canals in the peti
oles, six major canals in the peduncles, and the absence of acicular sclereids from
all flower parts. Both a powdery stigma and a granulate seed surface, unique to
subg. Hydrocallis, might be considered derived within Nymphaea; however, these
character states occur in several species of subg. Hydrocallis, including those
which exhibit mostly primitive features, which suggests an ancestral status within
the subgenus.
Of the four species for which chromosome counts indicate 2n = 28, N. novo
granatensis appears to approximate most closely the hypothetical Hydrocallis
ancestor, as defined in the preceding paragraph, from which it differs mainly in
lacking stolons and myricetin compounds. The other three species, N. gardneri
ana, N. conardii, and N. jamesoniana, exhibit a weblike leaf venation pattern, a
depauperate flavonoid profile, a progression toward smaller seeds, elongate car
pellary appendages, and a premidnight flowering schedule, all of which are be
lieved to be derived within this line. Of the three, N. gardneriana appears to have
retained a more primitive flavonoid chemistry and seed morphology than the
other two species. On the basis of seed and pollen morphology and floral biology,
N. jamesoniana might be considered the most specialized member of the group,
although N. conardii, in its production of kaempferol and isorhamnetin, must also
be viewed as advanced.
A second evolutionary line which includes N. glandulifera, N. potamophila,
N. oxypetala, and perhaps N. belophylla is thought to have diverged from the 2n
= 28 group. Although not having been karyotyped, N. glandulifera is obviously
related to this chromosome line on the basis of floral biology and overall mor
phology; however, in its seed morphology it has clearly diverged from this group.
Certain advanced seed features, the absence of acicular sclereids in the stigmatic
rays, and the lack of a weblike leaf venation pattern suggest an association be
tween N. glandulifera and the sagittate-leaved N. potamophila and N. oxypetala,
the latter two representing more specialized taxa. Nymphaea oxypetala, with its
submersed habit, unusual floral morphology, production of spherical staminal
sclereids, and polyploid nature, is perhaps the most specialized of all members of
subg. Hydrocallis. The position of N. belophylla remains uncertain at this time;
the possibility that N. belophylla represents a recent hybrid between two of these
species cannot be eliminated.
Although N. novogranatensis appears to approximate the ancestral form
within the 2n = 28 line, it also most closely resembles the 2n = 18 group. Its
closest relative is N. tenerinervia, both in morphology and floral biology. Nym
phaea tenerinervia, 2n = 20, is closely allied phytochemically to N. amazonum of
the 2n = 18 line and may represent a link between the two major groups. All
three of these species exhibit most of the primitive features discussed above and
are believed to be associated with the emergence of this additional chromosome
line. The other three members of the 2n = 18 line, N. prolif?ra, N. lasiophylla,
and N. lingulata, are similar in lacking the tetrameric arrangement of flower parts
present in other species of subg. Hydrocallis. Nymphaea prolif?ra and N. lasio
phylla are further related as the only members of subg. Hydrocallis which produce
tuberiferous flowers, but this feature is absent from N. lingulata. Among these
three species, N. prolif?ra is most similar to N. amazonum with regard to overall
morphology and phytochemistry. Nymphaea lasiophylla and N. lingulata, the only
members of subg. Hydrocallis with tapering carpellary appendages, are obviously
closely related, their relationship further substantiated by distinctive seed mor
phologies and flavonoid profiles.
Several features which might be presumed to be ancestral due to their pres
ence in other subgenera are absent from most species of subg. Hydrocallis but
present in N. lingulata and N. lasiophylla. In particular, the less ordered arrange
ment of the perianth, absence of petaloid stamens, tapered carpellary append
ages, certain seed characteristics, and presence of 3-methylated flavonoids in
these two species fit this pattern. However, the probable pathway for the emer
gence of the 2n = 18 chromosome line, as already discussed, would seem to have
involved an ancestor similar to N. novogranatensis, N. tenerinervia, and N.
amazonum, which do not display these characteristics. This supposed pathway
would dictate that N. lingulata and N. lasiophylla be viewed as more advanced
within the 2n = 18 line. Assuming this phylogenetic interpretation to be correct,
the features in question must have been acquired secondarily by N. lingulata and
N. lasiophylla. The possibility that this acquisition may be the result of genetic
interaction with other subgenera remains to be investigated.
Intersubgeneric hybridization with subg. Lotos may have been involved in the
formation of N. rudgeana; however, enough evidence has not yet been gathered
to confirm this hypothesis. The relationship of N. rudgeana to subg. Lotos has
been discussed in several sections above. Within subg. Hydrocallis, N. rudgeana is
most closely allied phytochemically to N. amazonum subsp. pedersenii.
TAXONOMY
Nymphaea Linnaeus, Sp. pi. 510. 1753, nom. cons. Leuconymphaea Ludwig ex
Kuntze, Revis, gen. pi. 1: 11. 1891.?Type: Nymphaea alba Linnaeus
(typ. cons.).
Castalia Salisbury, Ann. Bot. (K?nig & Sims) 2: 71. 1806.?Lectotype, here
designated: Castalia ampia Salisbury [=Nymphaea ampia (Salisbury) de

Candolle].
Perennial, aquatic herbs. Rhizome erect or horizontal, ovoid to cylindric,

often sprouting freely, rooted in substrate of lakes, ponds, etc., elongate stolons
present or absent. Leaf blades floating, occasionally emergent or submersed,
elliptic-sagittate to orbicular, entire to spinose-dentate, acute-tapered to truncate
at apex, cordate to sagittate at base, broadly or not at all peltate at base of sinus,
with few to numerous trichosclereids embedded in mesophyll, occasionally also
with acicular sclereids, glabrous or pubescent; venation mostly radiate but with a
prominent midvein. Petioles usually elongating to water surface, glabrous or pu
bescent, with two to several large air canals and often many smaller ones. Pedun
cles mostly stouter than petioles, especially when emergent, glabrous or pubes
cent, with several larger air canals and many smaller ones, strongly recoiled to
only slightly bent in fruit. Flowers solitary, floating or emergent, opening diur
nally or nocturnally, strongly to faintly odorous, perfect, actinomorphic, the parts
partially fused to and departing at various levels from the receptacular tissue
surrounding the carpels. Sepals 4 (rarely 3 or 5), free, hypogynous, often contain
ing sclereids, the veins prominent or inconspicuous. Petals 7-40, free, white,
blue, red, or yellow, in various shades, inserted in several series, hypogynous to
perigynous, occasionally containing sclereids, the outer alternate with sepals;
transition from petals to stamens gradual or abrupt. Stamens numerous (20-700),
multiseriate, perigynous to epigynous, often containing sclereids, outer filaments
broader and often petaloid, inner filaments narrower or filiform, connectives with
or without terminal appendages, anthers introrse, bilocular. Carpels many (5-47),
forming a ring embedded in cupshaped receptacular and appendicular tissue and
fused with it both abaxially and adaxially, free or coherent laterally, upper surface
of each carpel forming a ray of stigmatic tissue, this ray usually terminating
abaxially in a free, upwardly incurved carpellary appendage, this appendage
(when present) triangular, tapered, linear, or cl?vate; ovules numerous, anatro
pous, with laminar placentation; axis of flower often projecting centrally as a
short process. Fruit an irregularly dehiscent, berrylike capsule, ripening under
water; seeds with a floating membranous aril, a small straight embryo, little
endosperm, and abundant perisperm.
Nymphaea subg. Hydrocallis (Planch?n) Conard, Publ. Carnegie Inst. Wash. 4:

200. 1905. Nymphaea sect. Hydrocallis Planch?n, FI. Serres Jard. Eur. 8:
120. 1852. Nymphaea subsect. Hydrocallis (Planch?n) Caspary in Mar
tius, Fl. bras. 4(2): 160. 1878.?Lectotype, here designated: Nymphaea
amazonum Martius & Zuccarini.
Nymphaea tribe Chamaelotos Lehmann pro parte, Hamburger Garten- Blu
menzeitung 9: 203. 1853.
Rhizome erect, tuberous, ovoid to cylindric, usually with contractile roots

which pull tuber into substrate during dry season, producing stolons either only at
germination or throughout vegetative period. Leaf blades elliptic-sagittate to or
bicular, strongly to scarcely peltate at base of sinus, containing trichosclereids in
mesophyll, also often with acicular sclereids, green above and below, often suf
fused, mottled, or spotted with reddish or purplish pigment, particularly below,

glabrous. Petioles glabrous to glabrate, or pubescent or only apically so, with
usually 2 or 4 (-8-12) major air canals. Peduncles glabrous or glabrate or rarely
pubescent at apex, with (4-) 6 or 18 major air canals, retracting or bending only
slightly in fruit. Flowers floating or emergent, opening nocturnally, often strongly
odorous, in N. lasiophylla and N. prolif?ra often aborting and becoming tuberifer
ous, sepals and outer petals (sometimes inner petals and outer stamens as well) in
distinct whorls of four. Sepals green, sometimes suffused with red or brown, often
flecked with black or purple, rounded to long-acuminate at apex, with or without
acicular sclereids, inconspicuously or faintly nerved. Petals 8-35, white to light
yellow, the outer often greenish abaxially, rounded to long-acuminate at apex,
with or without acicular sclereids; transition to stamens mostly gradual, rarely
abrupt. Stamens 24-324, creamy white to light yellow, the inner sometimes suf
fused with blue or purple, usually containing coarse or fine acicular sclereids,
appendages short or absent. Carpels 11-47, completely coherent laterally, ap
pendages cream-colored or suffused with red or purple, to 30 mm long, often
strongly incurved, with linear-clavate, cl?vate, or Ungulate-tapering tips, stigmatic
rays with or without projecting acicular sclereids and papillae cells separating into
powdery mass or remaining attached; axial process absent or blunt to acuminate.
Seeds globose to elliptic, with numerous short to long papillae, the surface
smooth or finely ruminate or granulate.
Key to the Species of Nymphaea Subgenus Hydrocallis
1. Carpellary appendages Ungulate-tapering; transition from petals to stamens abrupt, the outer
stamens only slightly or not at all petaloid.
2. Petals 20-26; tuberiferous flowers usually present; acicular sclereids absent from leaf
mesophyll. 12. N. lasiophylla.
2. Petals 8-14; tuberiferous flowers absent; acicular sclereids common in leaf mesophyll.
13. N. lingulata.
1. Carpellary appendages linear to strongly cl?vate; transition from petals to stamens gradual,
the outer stamens distinctly petaloid.
3. Leaf blades irregularly dentate; seeds 1.2-2.4 mm long. 14. N. rudgeana.
3. Leaf blades entire; seeds rarely exceeding 1.3 mm long.
4. Mature leaves mainly submersed (if floating then reduced in size), lacking sclereids in
mesophyll; sepals and outer petals acuminate. 7. N. oxypetala.
4. Mature leaves floating, sclereids common in mesophyll; sepals and outer petals
rounded to acute.
5. Leaf blades sagittate, over 1.8 times as long as wide.
6. Carpellary appendages 10-17 mm long; acicular sclereids present in leaf meso
phyll and in stigmatic rays. 5. N. belophylla.
6. Carpellary appendages ca. 7-9 mm long; acicular sclereids absent from leaf
mesophyll and from stigmatic rays. 6. N. potamophila.
5. Leaf blades suborbicular to elliptic, less than 1.6 times as long as wide.
7. Leaf veins with slightly to strongly evident cross veins centrally, forming weblike
venation pattern; stigmatic rays often containing few to numerous acicular scle
reids.
8. Carpellary appendages 3-9 mm long, less than Vs times length of longest
sepal; seeds 1.1-1.2 times as long as broad. 1. N. jamesoniana.
8. Carpellary appendages mostly 9 mm or more long, the longest more than Vs
times length of longest sepal; seeds 1.2 or more times as long as broad.
9. Plants producing stolons only during germination of tubers; flowers au

togamous, with fruits normally maturing; leaves uniformly green below,
rarely reddened. 2. N. conardii.
9. Plants stoloniferous throughout growing season; flowers no
the fruits seldom maturing; leaves usually mottled with rust
ment below. 3. N. gardneria
1. Leaf veins lacking evident cross veins centrally, the venation ra
rays not containing acicular sclereids.
10. Leaves uniformly green above and below; sepals uniformly g
petals with numerous acicular sclereids. 4. N.
10. Leaves spotted or suffused with red or purple, especially on
sepals usually suffused with brown or purple abaxially or w
black flecks; petals lacking acicular sclereids.
11. Carpellary appendages cylindric-clavate, apiculate at apex
abruptly narrowed from 3 outer whorls.
12. Leaves mostly with mucronulate apices; acicular scle
mesophyll not forked; sepals uniformly colored abaxially
8. N. novogranatensis.
12. Leaves with obtuse or rounded apices; acicular scle
mesophyll forked; sepals usually flecked with purple abax
9. N. tenerinervia.
11. Carpellary appendages strongly cl?vate, blunt at apex
gradually narrowed from outer petals.
13. Tuberiferous flowers usually present; leaves lacking rin
cence at apex of petiole; only sepals and outer petals i
whorls of four. 11. N. prolif?ra.
13. Tuberiferous flowers never produced; leaves usually with rin
pubescence at apex of petiole; sepals, petals, and outer stamen
in distinct whorls of four. 10. N. amazonum.
1. Nymphaea jamesoniana Planch?n, FI. Serres Jard. Eur. 8: 120. 185

nymphaea jamesoniana (Planch?n) Kuntze, Revis, gen. pi. 1: 11
Castalia jamesoniana (Planch?n) Britton & P. Wilson, Bot. Por
5(2): 305. 1924.?Type: Ecuador. Guayas: Guayaquil, Mar 1846,
son 546 (lectotype, here designated: G!; isolectotypes: B! BM! FI-W
Material at K contains a leaf and flower of N. jamesoniana and on
N. ampia (Salisbury) de Candolle.
Nymphaea sagittariaefolia Lehmann, Hamburger Garten- Blumenzeit
216. 1853.?Type: "in America centrali," Lehmann f. s.n. (not l
Castalia gibertii Morong, Ann. New York Acad. Sei. 7: 48. 1893. N
gibertii (Morong) Conard, Publ. Carnegie Inst. Wash. 4: 21
Type: Paraguay. Pilcomayo River, above the falls, 6 May 188
Morong 1028 (lectotype, here designated: NY!; isolectotypes: M
US!).
Rhizome to 7 cm long, to 3.5 cm in diameter, ovoid to cylindric, lacking
stolons. Leaf blades to 23.5 cm long, to 18.5 cm wide, 1.18-1.55 times as long as
wide, delicate to subcoriaceous, elliptic, entire, rounded at apex, peltate 1-9 mm
from base of sinus, the sinus 0.34-0.49 times length of blade, the lobes rounded
to obtuse- or acute-angled; upper surface green, sometimes with dark flecks,
when dried with rows of few to numerous tiny prominences from underlying
trichosclereids and with short elevated lines from abundant underlying acicular
sclereids; undersurface green, sometimes with darker flecks, when dried usually
marked with slender, branched, rusty brown lines; venation conspicuously web
like centrally, dichotomous and faintly reticulate toward margin, the principal
veins 11-17. Petioles to 5 mm in diameter, glabrous, with 2 major and 4-6 minor
air canals. Peduncles to 7.5 mm in diameter, glabrous, with an inner ring of 6-7
major air canals surrounded by 12-14 minor ones. Flowers floating, the sepals,
petals, and outer stamens in distinct whorls of four. Sepals 2.5-6.5 cm long, 1.1
3.5 cm wide, uniformly green, ovate, acute to somewhat rounded at apex, when
dry marked with slender rusty brown lines, containing numerous acicular scle
reids, inconspicuously nerved. Petals 12, 16, or 20, containing acicular sclereids,
the outer greenish at least abaxially, mostly equal to or somewhat shorter and
narrower than sepals, with apex acute to rounded, the inner 1.8-4.5 cm long,
0.7-2 cm wide, creamy white, acute or rarely obtuse; transition to stamens
gradual. Stamens 35-85, creamy white, containing acicular sclereids, the outer
most to 3.9 cm long, with filaments to 15 mm wide, the inner with filaments to 2.5
mm wide and connectives prolonged to 0.3 (-0.5) mm beyond anthers, the inner
anthers 0.36-0.65 times as long as stamen; pollen diameter 26-32 |im. Carpels
19-33, the appendages 3-9 mm long, cream-colored with somewhat reddish or
purplish tips, the tips slightly cl?vate, 0.6-1.5 mm in diameter, the stigmatic rays
with none or few, occasionally numerous, projecting acicular sclereids and papil
lae cells remaining attached; axial process mostly 2-3 mm long, broadly acute to
acuminate. Fruits usually present, to 8 cm in diameter; seeds 0.5-0.8 mm long,
0.45-0.7 mm in diameter, 1.1-1.2 times as long as broad, brownish or whitish
when dry, with numerous short hairs 10-50 \im long, the surface finely granulate.
Tuberiferous flowers absent. Chromosome number: 2n = 28. Figs. 5, 11, 12, 17.
Phenology. Collected in flower and fruit throughout the year in tropical areas
and during warmer seasons in subtropical areas.
Distribution. (Fig. 18). Tropical and subtropical America, including Florida,
Mexico, Central America, Colombia, Ecuador, Brazil, Paraguay, and Argentina;
in freshwater ditches, ponds, or slow-moving streams, often with Characeae.
Specimens Examined. U.S.A. Florida: Citrus Co., ford leading to Sheep Island, "The Cove,"
Floral City, 1 Oct 1940, St. John s.n. (FLAS); DeSoto Co., 2.5 mi NW of Arcadia, 4 Nov 1967,
Conard s.n. (FLAS, USF); Lee Co., ca. 7 mi E Ft. Myers, Les 114 (USF), Les 140 (USF); Ft. Myers,
Les 159 (USF); Levy Co., Lebanon Station, Cooley et al. 7151 (FLAS, USF); Sarasota Co., Fla 72,
0.4 mi E Myakka R., Wiersema 2032 (UNA, US, USF).?Mexico. Chiapas: Mpio Arriaga, 6 km N
Arriaga, Breedlove 26918 (ENCB); Mpio Acala, road to La Angostura, 12-14 km S Tuxtla, Breedlove
37370 (MO), Mpio Tonal?, NW Puerto Arista, Breedlove & Thome 20917 (ENCB, MO, NY). Ta
basco: N-20, 4 km E Poblado el Naranjito, Men?ndez 255 (BM, K, MO). Tamaulipas: Mex-180, 14
km N of R?o Carrizal, Haynes 5441 (UNA), Wiersema 1341 (F, UNA, US). Veracruz: Mpio Paso de
Ovejas, Mata Mateo, Ventura 7681 (ENCB).
El Salvador. La Uni?n: Lago del Olomega, 8 km W La Uni?n, Fassett 28615 (F, WIS). San
Miguel: N end of Laguna Jocotal, San Salvador, Fassett 28565 (F, WIS).?Honduras. Yoro: 3 km N
of Las Lomas, Lentz 1187 (EAP, UNA).?Nicaragua. Matagalpa: 8 km E San Isidro, Haynes 8572
(MO, UNA). Rivas: Las Piedras, Haynes 8430 (MO, UNA). Prov. unkown: vicinity of Managua, lago
de Granada, Gamier 298 (US).
Argentina. Chaco: Depto La Leonesa, Ruta 11 between Resistencia and Formosa, Wiersema &
Vanni 2265 (CTES, UNA, CORD). Corrientes: Depto Mburucuy?, Estancia "Santa Teresa," Peder
sen 7741 (C); Depto San Miguel, Ruta 5, 9 km SE Ca? Cat?, Wiersema et al. 2277 (CTES, UNA).
Formosa: Depto Piran?, 6 km NW Piran?, Morel 531 (LIL); Depto Patino, riacho Montelindo, Arenas
2060 (BACP, UNA). Salta: Depto Oran, Aguaray, Schreiter 11035 (LIL). Santa Fe: Depto Capital,
FIG. 17. Nymphaea jamesoniana (Wiersema et al. 2290). a. Undersurface of leaf. b.

flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross section of peduncle, f
surface of leaf.
Laguna Setubal, Rinc?n Norte, Mart?nez-Achenbach 97, 159 (LIL). Tucum?n: Depto Mon
Acharal, Venturi 1701 (BAB, GH, LIL, US).?Brazil. Alagoas: BR-316 at km 172, 8 km E P
dos Indios, Wiersema et al. 2290 (INPA, IPA, MO, NY, UNA). Bahia: Rio Salitre, between J
and Sobradinho, Wiersema et al. 2300 (IPA, UNA). Cear?: BR-116 at km 65, 65 km S Fo
Wiersema et al. 2315 (IPA, UNA). Maranh?o: BR-316 at km 598, 23 km W Rio Parnaiba, W
et al. 2306 (IPA, UNA). Mato Grosso: Mun. Pocon?, Rodavia Transpantaneira, da Silv
(UFMT). Mato Grosso do Sul: Estrada Transpantaneira, Acampamento do Exercito, Heringe
868 (UB). Pernambuco: Mun. Boa Vista, BR-428 at km 76, Riacho das Gar?as, Wiersema et
(IPA, UNA). State unknown: S?o Marcos, L?tzelburg 20502 (M).?Colombia. Atl?ntico: re
Barranquilla, Elias 1387 (F, US).?Ecuador. El Oro: between M?chala and Santa Rosa, 6 Ju
Guldmann s.n. (K). Guayas: hwy 25 Guayaquil-Babahoyo, 17 km S Babahoyo, Wiersema et
90?W 60?W
FIG. 18. Distribution of Nymphaea jamesonian
(AAU, UNA, USF); hwy 25 Guayaquil-Babahoyo, 16

(UNA).?Paraguay. Central: Asunci?n, June 1858, G
de Limpio a Emboscada, Schinini 22006 (CTES); Pr
1978, Arenas s.n. (BACP, UNA).
Observations of Argentine material {Wier

cate that flowers of TV. jamesoniana open f
Flowers begin opening at dusk and closing a
opening three nights have their stigmas cove
nights and dry the third night. Stamens of three-day flowers are exposed and
begin dehiscing the second night. Two-day flowers fail to open the first night of
this three-day sequence. In the Argentine plants the flowers open wide, the sepals
and petals becoming broadly reflexed; however, in plants cultivated from Ecuador
(Wiersema et al. 2026) and from Florida (Wiersema 2031) the flowers barely open,
the sepals and petals spreading only to a vertical position. All flowers emit a
strong aromatic fragrance. Crossing studies indicate cross-pollination is possible
both the first and second night for three-day flowers. Self-pollination is clearly
possible on the second day.
Reproduction in N. jamesoniana is entirely dependent on seed production.
No form of vegetative propagation was observed in field or cultivated plants. Seed
production appears to be largely autogamous, with fruits and seeds always devel
oping, even in the absence of potential pollinators.
Nymphaea jamesoniana has the smallest seeds with the shortest surface papil
lae in subg. Hydrocallis. Based on considerations of chromosome number, phyto
chemistry, overall morphology, and seed morphology, N. jamesoniana displays its
strongest affinity to N. conardii. Its shorter carpellary appendages readily sepa
rate it from this species.
Unusually robust forms have been observed both in Florida and Argentina.
Plants in Argentina and Paraguay have slightly shorter carpellary appendages and
more coriaceous leaves than those in other areas and have been treated as N.
gibertii. Seed morphology and phytochemistry of these plants is identical to that
observed in typical N. jamesoniana. Plants in northeastern Brazil produce larger
seeds than other N. jamesoniana populations and have slightly longer and more
cl?vate capellary appendages. Seeds of Mexican plants were also unusual in their
shorter, almost nonexistent, papillae.
The specimen from the Delessert Herbarium at G was chosen as the lectotype
of N. jamesoniana, as this specimen was cited in Planchon's expanded 1853 treat
ment along with material of the Herbarium Webbianum which I did not examine.
In describing his N. sagittariaefolia, Lehmann (1853) cited only: "in America
centrali collegit filius Guilielmus Lehmann." According to Caspary (1878) no such
specimen existed in Lehmann's herbarium, only the collection of Jameson which
was not mentioned in Lehmann's manuscript. Lehmann (1854) later indicated his
species to be synonymous with N. jamesoniana Planch?n; the collections of Jame
son from Lehmann's herbarium at B and at K bear the name N. sagittariaefolia in
Lehmann's handwriting.
2. Nymphaea conardii Wiersema, Brittonia 36: 213. 1984.?Type: Venezuela.

Barinas: Distrito Sosa, pond along stream off road between Puerto Nu
trias and Ciudad de Nutrias, ca. 1 km N of junction with road to Bruzual,
29 Aug 1981, Wiersema & Gonz?lez 2214 (holotype: MO!; isotypes: F!
MY! NY! UNA! VEN!).
Rhizome to 4.5 cm long, to 3.5 cm in diameter, ovoid, producing stolons only

during germination. Leaf blades to 18 cm long, to 14 cm wide, 1.05-1.5 times as
long as wide, delicate, ovate to broadly elliptic, entire, rounded at apex, peltate
0.5-3.5 mm from base of sinus, the sinus 0.37-0.48 times length of blade, the
lobes acute to rounded at apex with slight protuberance; upper surface green,
when dried with numerous tiny prominences from underlying trichosclereids inter
spersed with short elevated lines often visible from underlying acicular sclereids;
undersurface green, when dried often marked with slender, branched rusty brown
lines; venation slightly to evidently weblike centrally, dichotomous and faintly
reticulate toward margin, the principal veins 9-15. Petioles to 4 mm in diameter,
glabrous, with 2-4 major air canals and 4-6 minor ones. Peduncles to 6.5 mm in
diameter, glabrous, with an inner ring of 6 major air canals surrounded by 12
minor ones. Flowers floating, with sepals and petals in distinct whorls of four.
Sepals 3-6 (-7.2) cm long, 1-3 cm wide, uniformly green, oblong-ovate, acute or
somewhat rounded at apex, often marked with slender rusty brown lines when
dry, containing acicular sclereids, inconspicuously nerved. Petals usually 12, 16,
or 20, containing numerous acicular sclereids, the outer greenish abaxially,
shorter than sepals, with apex acute to somewhat rounded, the inner 2.1-4.4 cm
long, 0.6-2 cm wide, creamy white, acute to acuminate; transition to stamens
gradual. Stamens 24-80, creamy white, containing acicular sclereids, the outer
most to 3.3 cm long, with filaments to 15 mm wide, the inner with filaments to 1.5
mm wide and connectives prolonged to 0.3 mm beyond anthers, the inner anthers
0.4-0.7 times as long as stamen; pollen diameter 25-40 \im. Carpels 16-30, the
appendages 7-16 mm long, cream-colored, with cl?vate tips to 1.2 (-1.5) mm in
diameter which usually darken upon drying, the stigmatic rays with projecting
acicular sclereids and papillae cells remaining attached; axial process 1-1.5 mm
long and blunt. Maturing fruits usually present; seeds 0.65-0.9 mm long, 0.5-0.7
mm in diameter, 1.2-1.4 times as long as broad, with longitudinal rows of short
hairs 25-70 \im long, the surface finely granulate. Tuberiferous flowers absent.
Chromosome number: 2n = 28. Figs. 5, 11, 12, 19.
Phenology. Collected in flower and fruit from June (Brazil) or July (else
where) to December.
Distribution. (Fig. 20). Southern Mexico through Central America to north
ern South America, including Colombia, Venezuela, and northern Brazil; also
found in Cuba, Puerto Rico, and Hispaniola; mostly in temporary ditches or
ponds in inland habitats.
Representative Specimens. Cuba. Santa Clara: Yag?ey Chico, in swampy forest at Laguna del
Yanco, Ekman 17007 (G, S).?Dominican Republic Duarte: Pimental, Ekman 13277 (S).?Haiti.
Dept. du Nord: Bayeux, between Bord de la Mer and Morne ? Madame, Ekman 2679 (S).?Puerto
Rico. A?asco, near "Hatillo," Sintenis 5688 (BM, G, GH, GOET, K, M, P, S).
Mexico. Chiapas: Mpio Jiguipilas, 16 km ESE of Tierra and Libertad along road to Villa Flores,
Breedlove 27347 (MO).
Belize. Punta Gorda, Schipp 1026 (BM, G, F, K, MO, NY, S).?El Salvador. Santa Ana: 13
km N of Texistepeque, 28 km N of Santa Ana, Fassett 28279 (F, WIS).?Honduras. Comayagua: 5
km W of Comayagua, Molina & Molina 22704 (BM, F, G, S).?Nicaragua. Chontales: 2 Vi km SW of
Quebrada Niscala along road Acoyapa-San Carlos, Haynes 8588 (UNA). Le?n: Las Penitas, Haynes
8612 (UNA). Managua: Sierra de Managua, 1930-1940, Gamier s.n. (F). Rio San Juan: 0.5 km S R?o
Oyate along road to San Miguelito, Nee & Miller 27620 (MO, UNA).?Panama. Canal Zone: Barro
Colorado Island, Ebinger 553 (F). Dari?n: pasture belonging to Pablo Oth?n, Duke & Bristan 283
(MO). Panam?: between Panam? and Chepo, Dodge et al. 16712 (G, MO, P). San Blas: Ailigand?
area, Hammel & D'Arcy 5007 (MO).
Brazil. Alagoas: Mun. Marimbondo, km 218 of BR-316 in Rio Tapuia, Wiersema et al. 2289
(IPA, UNA). Para: Bel?m-Brasilia, Rio Marsh, Brantjes 707801 (SP).?Colombia. Guajira: R?oha
cha, Haught 4431 (COL, US). Santander: Puerto Wilches and vicinity, Killip & Smith 14821 (F, GH,
NY, US).?Venezuela. Apure: hwy 19 at km 44, W of Guasdualito, Wiersema & Gonz?lez 2230
FIG. 19. Nymphaea conardii (Wiersema & Gonz?lez 2214). a. Undersurface of leaf. b. Opened
flower, c. Unopened flower bud. d. Cross section of petiole, d. Cross section of peduncle.
(MY, UNA, US, VEN). Bol?var: on road to San Felix, between Upata and Altagracia, 1-10 km NW
Upata, Steyermark 57693 (F, VEN). Cojedes: Sistema de Riego Las Majaguas, Presas 1-2-3, Sector
Agua Blanca, Trujillo 11181 (MY). Portuguesa: Laguna El Sam?n, near Acarigua, Solazar 27 (US,
VEN). Tachira: between Piscuri and Ayari, Steyermark & Rabe 96622 (NY, US, VEN). Zulia:
carretera Machiques-Col?n, 15 km S Maracaibo, Steyermark & Fern?ndez 99560 (VEN).
Observations on cultivated material from the type locality (Wiersema &

Gonz?lez 2214) and from Nicaragua (Haynes 8588) indicate that flowers of N.
conardii open for two successive nights. They begin opening at dusk both nights
and commence closing around midnight, being fully closed by 1 a.m. The stigma is
wet with stigmatic fluid both nights. The flowers emit a strong but pleasant aroma.
The stamens are not exposed until they begin to dehisce in second-day flowers. The
stigma is clearly receptive to pollination the second night as seeds are produced
autogamously. Numerous individuals of a scarab beetle, Cyclocephala verticalis
Burmeister, were observed in flowers of N. conardii at the type locality.
Reproduction in N. conardii is entirely dependent on seed production; no
form of vegetative propagation has been observed either in field or cultivated
plants. Much of this seed production appears to be the result of autogamy, with
fruits and seeds always developing even in the absence of potential pollinators.
80? .70?___6J)? 50?
FIG. 20. Distribution of Nymphaea conardii and N. glandulifera.
Nymphaea conardii is most closely allied to N. gardneriana (no.. 3) and N.

jamesoniana (no. 1), based on evidence from seed morphology, phytochemistry,
chromosome number, and overall morphology. It also displays some taxonomic
affinity to N. glandulifera (no. 4), although comparative data on phytochemistry
and chromosome number are lacking for this species. Features distinguishing
these other species from N. conardii are discussed under the other species.
Populations of N. conardii in Mexico and northern Central America display
several differences from other populations in South America and the Greater
Antilles. Leaves from these populations are usually suffused or mottled with red
pigment, and fewer acicular sclereids are present in the floral stigmatic rays.
These plants also produce mostly glycosides of kaempferol; glycosides of isorham
netin are found in other populations of TV. conardii in South America. The pre
sence of kaempferol in plants from the Venezuelan type locality supports their
inclusion in N. conardii, as does the identical seed morphology and floral biology
of plants from the two regions. Although these populations may prove to be
deserving of infraspecific recognition, no formal recognition is proposed here.
3. Nymphaea gardneriana Planch?n, FI. Serres Jard. Eur. 8: 120. 1852. Nymphaea
passiflora Lehmann, Hamburger Garten- Blumenzeitung 9: 205, 217.
1853, nom. superfl. Leuconymphaea gardneriana (Planch?n) Kuntze, Re
vis, gen. pi. 1: 11. 1891.?Type: "Brasilia, in prov. Piauhy," Jul-Sep
1839, Gardner 2476 (lectotype, here designated: G!; isolectotypes: B!

BM! FI-W, G! K! P! US!). The specimen at B from the herb. Lehmann
also contains material of Gardner 3568.
Nymphaea stenaspidota Caspary in Martius, Fl. bras. 4(2): 175. 1878.?Type:
Brazil. Goi?s: "in lacubus parvis ad Concei?ao," 1840, Gardner 3568
(lectotype, here designated: BM!; isolectotypes: K! B!). A second sheet
at K labelled Gardner 3568 bears one leaf of this collection and one
flower of N. ampia (Salisbury) de Candolle. The material at B is mounted
with Gardner 2476.
Nymphaea wittiana Ule, Notizbl. K?nigl. Bot. Gart. Berlin 6: 294. 1915.?
Type: Brazil. "Amazonas" [Roraima]: in standing bodies of water at
Surumu by the Serra do Mel in the region of the upper Rio Branco, Aug
1909, Ule 8120 (lectotype, here designated: B!; isolectotypes: B, not lo
cated, as photo F! GH!). Ule 8120 at K and G(2) are N. tenerinervia
Caspary.
Rhizome ovoid, stolon-forming throughout vegetative period. Leaf blades to

21.5 cm long, to 15.5 cm wide, 1-1.5 times as long as wide, subcoriaceous, elliptic
to suborbicular, entire, rounded at apex, peltate 0-8 mm from base ?f sinus, the
sinus 0.33-0.52 times length of blade, the lobes acute to rounded at apex; upper
surface uniformly green, when dried with radiating rows of tiny prominences from
underlying trichosclereids, with acicular sclereids positioned between the rows
also often visible; undersurface green, often suffused or mottled with rusty brown
hue, when dried often marked with slender branched, rusty brown lines; venation
distinctly weblike centrally, dichotomous and faintly reticulate toward margins,
the principal veins 9-17. Petioles to 4 mm in diameter, glabrous or glabrate with
long septate hairs, with 2 or 4 major and 6-10 minor air canals. Peduncles to 6
mm in diameter, glabrous, with an inner ring of 5-6 major air canals surrounded
by 10-12 minor ones. Flowers floating, the sepals and petals in distinct whorls of
four. Sepals 4-8 cm long, 1.2-3.3 cm wide, green, oblong-ovate, acute or rarely
somewhat rounded at apex, when dry marked with slender rusty brown lines,
containing abundant acicular sclereids, inconspicuously nerved. Petals 16-28, con
taining abundant acicular sclereids, the outer greenish abaxially, shorter but nar
rower to slightly wider than sepals, with apex obtuse- to acute-angled, the inner
2.6-5.9 cm long, 0.7-1.9 cm wide, creamy white, acute to acuminate; transition
to stamens gradual. Stamens 44-125, creamy white, containing numerous acicular
sclereids, the outermost to 4.6 cm long, with filaments to 13 mm wide, the inner
with filaments to 1.8 mm wide and connectives prolonged to 0.5 mm beyond
anthers, the inner anthers 0.55-0.8 times as long as stamen; pollen diameter 28
38 |?m. Carpels 13-35, the appendages (7.5-) 9-20 mm long, cream-colored, with
tips often pinkish or purplish, slightly cl?vate, 0.7-1.8 mm in diameter, the stig
matic rays with many projecting acicular sclereids and papillae cells remaining
attached; axial process 1-3 mm long, blunt to sharply acuminate. Fruits rarely
present; seeds 0.6-0.85 mm long, 0.5-0.65 mm in diameter, 1.2-1.3 times as long
as broad, greenish with longitudinal rows of hairs 20-90 [un long, the surface
finely granulate. Tuberiferous flowers absent. Chromosome number: 2n = 28.
Figs. 4, 11, 12, 21.
FIG. 21. Nymphaea gardneriana (Wiersema & Gonz?lez 2215). a. Undersurface of leaf. b.
Opened flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross section of peduncle, f.
Upper surface of leaf.
Phenology. Collected in flower throughout the year in tropical areas, but only
during warmer seasons in subtropical regions further south; collections in fruit,
although quite rare, suggest a similar schedule for fruiting.
Distribution. (Fig. 18). South America, including Venezuela, Brazil, Bolivia,
Paraguay, and Argentina.
Specimens Examined. Argentina. Chaco: Margarita Bel?n, Aguilar 751, 1045 (LIL); Colonia
Ben?tez, Wiersema et al. 2247a (UNA). Corrientes: Depto Concepci?n, 28 km SE Concepci?n, Schi
nini et al. 13270 (CTES); Depto Itat?, Ruta 12, 60 km E Itat?, Ahumada et al. 876 (CTES). Formosa:
near Guayerles, Jorgensen 3124 (GH, LIL, MO, US); Depto Matacos, 1 km N Tolder?a Toba, Arenas
2272 (BACP, UNA). Misiones: Depto Capital, Posadas, Rodrigues 98 (BAB).?Bolivia. Santa Cruz:
Prov. Nuflo de Ch?vez, 20 km S Concepci?n, Krapovickas & Schinini 32094 (CTES); Prov. Chiquitos,
28 km NE San Jos?, Krapovickas & Schinini 36605 (CTES).?Brazil. Acre: Estrada Alemanha,
Cruzeiro do Sul, Prance et al. 11912 (NY, UNA). Goi?s: Mun. Yatai, Macedo 3320 (US); Mun.
Alvorada, BR-020 at Rio Macacos, Hatschbach 39372 (MBM). Mato Grosso: inter Guia et Cuiab?,
14 May 1894, Malme s.n. (S). Minas Gerais: S?o Jo?o del Rey, Schwacke 10015 (P). Para: Taperinha
bei Santar?m, Ginzberger 657 (F, W); locality unknown, 1832, Siber s.n. (BR). S?o Paulo: Moji
Gua?u, Mos?n 1102 (S).?Paraguay. Paraguar?: Plaine de Mbatobi, Balansa 521 (BM, BR, G,
GOET, K, S). San Pedro: Primavera, R?o Tapiracuai, Woolston 1257 (K, NY, US).?Venezuela.
Apure: Mantecal, Ramia 4540 (VEN), Ramia & Montes 4701 (VEN); S of Bruzual, 3 km N of Ca?o
Guaritico, Wiersema & Gonz?lez 2215 (COL, F, MY, UNA, VEN). Barinas: km 30 of hwy Barinas
Cuidad Bolivia, Trujillo 13407 (MY). Bol?var: Sabana Cardona, R?o alto Suapure, Puig 3050 (VEN).
Cojedes: 10 km S of San Carlos, Trujillo 13825 (MY). Gu?rico: estaci?n experimental "Los Llanos,"
Trujillo 8881 (MY); hwy 2, 2 km S R?o Gu?rico near Calabozo, Wiersema 1403 (F, MO, NY, UNA,
US, VEN); hwy 2, 53 km S Calabozo, Wiersema 1412 (UNA, VEN); hwy 2, 9.5 km S R?o Gu?rico
near Calabozo, Wiersema 1428 (UNA, VEN). Portuguesa: hwy 5, 17 km SW of Guanare, 2 km NE of
R?o Tucupido, Wiersema & Gonz?lez 2210 (MY, NY, UNA, VEN); Depto Guanare, terrenos de la
Unellez, Stergios 7077 (PORT, UNA).
Observations on Venezuelan {Wiersema 1428) and Argentine {Wiersema et al.

2247a) material in cultivation indicate that flowers open for two consecutive
nights. On both nights opening commences at dusk, with flowers closing around
midnight. During the first night the stigma is covered with stigmatic fluid; how
ever, access to the stigmatic surface is prevented by the overlying stamens. The
stigma remains moistened the second night and is exposed along with the now
dehiscing stamens. Crossing studies indicate that the stigma is not receptive to
either cross- or self-pollination on the second night. Flowers both nights emit a
strongly aromatic, somewhat sweet-scented fragrance.
Propagation in most populations is the result of stolon formation. Numerous
stolons are produced from a single individual. Although seeds are readily pro
duced from artificial crosses in cultivation, they are rarely found in nature, as
evidenced from my own field observations and their presence on only three of the
herbarium specimens seen. Autogamous seed production has not been observed
in N. gardneriana. The dependence of seed set on cross-pollination suggests that
the low seed set occurring in nature might be related to a shortage of available
pollinators.
From an analysis of chromosome number, phytochemistry, and general mor
phology, N. gardneriana appears clearly related to N. conardii and N. jamesoni
ana. All three display a weblike venation pattern in their leaves that is not found
in other species of subg. Hydrocallis. Its stoloniferous habit distinguishes N. gard
neriana from these other two species which lack stolons. Although seed morphol
ogy would also readily distinguish N. gardneriana, seeds of this species unfortu
nately are rarely encountered in nature. It appears, from the limited evidence
available, that the seeds of N. gardneriana bear some relation to those of N.
belophylla (q.v., no. 5).
Nymphaea gardneriana is highly polymorphic. This may be partly due to
extremely low levels of sexual reproduction, with a consequent reliance on asex
ual reproduction. Noteworthy among this variation is the absence of reddish
coloration in the leaves in certain populations, particularly those in southern
South America. Flowers in these latter populations develop a purplish color in the
carpellary appendages, with the tips of these appendages more incurved than in
populations further north. Several Brazilian populations, as well as plants from
the type locality, exhibit leaves which are noticeably more peltate than those in
other parts of the range. There is evidence from some populations that this
characteristic may be variable, but unfortunately none of these populations has
been studied in the field or in cultivation by me. Overall, they display a similarity
to other populations in Argentina and Venezuela that have been scrutinized more
extensively. The results of this study, particularly with regard to seed morphology
and phytochemistry, support the acceptance of a single variable species.
The specimen at G from the herbarium Delessert has been chosen as the
lectotype of N. gardneriana; it was one of two specimens cited by Planch?n in his
expanded 1853 treatment. The name N. passiftora Lehmann is superfluous, be
cause Lehmann cited the type of N. gardneriana. The collection of C. Beyrich
from Brazil (Serra d'Estrella) also cited by Lehmann (1853) could not be located
by Caspary (1878) in any herbarium, including Lehmann's. Caspary questioned
the authenticity of this reported collection.
4. Nymphaea glandulifera Rodschied, Med. Bern. Kol. Rio Essequebo 76. 1794.
Nymphaea blanda G. Meyer, Prim. fl. esseq. 201. 1818, nom. superfl.
Castalia blanda (G. Meyer) Lawson, Proc. and Trans. Roy. Soc. Canada
6: 117. 1888. Leuconymphaea blanda (G. Meyer) Kuntze, Revis, gen. pi.
1: 11. 1891.?Type: Guyana. Rio Essequibo, Rodschied 155 (lectotype,
here designated: GOET!). The specimen labelled Rodschied 155 at
GOET contains one leaf of N. glandulifera. The flower, described by
both Rodschied and Meyer, which belongs with this leaf and which (fide
Caspary 1878) was once included with this leaf is mounted with Rodschied
281, the type of N. rudgeana G. Meyer.
N. fenzliana Lehmann, Hamburger Garten- Blumenzeitung 9: 205, 217. 1853.
N. blanda var. fenzliana (Lehmann) Caspary in Martius, Fl. bras. 4(2):
173. 1878.?Type: Guatemala. San Juan de Nicaragua, 1841, Friedrichs
thal 1878 (holotype: W, destroyed; isotype: K!). Although Lehmann cited
only "San Jouan de Nicaragua"?apparently taken from a specimen sent
to him by E. Fenzl of the University of Vienna?Caspary (1878) indicates
the cited collection to be that of Friedrichsthal at W and K.
Rhizome ovoid, lacking stolons. Leaf blades to 21 cm long, to 19 cm wide, 1

1.35 times as long as wide, chartaceous, broadly elliptic to suborbicular, entire,
rounded at apex, peltate 2 mm or less from base of sinus, the sinus 0.35-0.5 times
length of blade, the lobes (acute or) obtuse to rounded at apex, often with slight
protuberance; upper surface glossy green, when dried with numerous tiny promi
nences from underlying trichosclereids, often with short elevated lines from un
derlying acicular sclereids; undersurface pale green; venation radiate centrally,
becoming dichotomous and faintly reticulate toward margins, the principal veins
9-15. Petioles to 5 mm in diameter, glabrous or pubescent (in the type) with long
septate hairs. Peduncles to 7 mm in diameter, glabrous or pubescent, usually with
an inner ring of 6 major air canals surrounded by 12 minor ones. Flowers floating,
the sepals and petals in distinct whorls of four. Sepals 3.2-5.7 (-7.5) cm long,
1.1-2.1 (-2.8) cm wide, green, oblong-ovate, acute at apex, when dry often
marked with slender rusty brown lines at base, containing acicular sclereids, in
conspicuously nerved. Petals 12-20, creamy white, containing acicular sclereids,
the outer slightly shorter than sepals, with apex acute, the inner 2-3.5 (-5) cm
long, 0.3-1.6 cm wide, acute to acuminate; transition to stamens gradual. Sta
mens 45-98, creamy white, containing acicular sclereids, the outermost to 3.7 cm
long, with filaments to 8 mm wide, the inner with filaments to 1.1 mm wide and
connectives prolonged to 0.4 mm beyond anthers, the inner anthers (0.4-) 0.5
0.7 times as long as stamen; pollen diameter ca. 36 \im. Carpels 19-39, the
appendages 7-14 mm long, cream-colored, linear to slightly cl?vate, with tips
usually darkening upon drying, to 1.2 mm in diameter, the stigmatic rays lacking
acicular sclereids and with papillae cells mostly remaining attached; axial process
blunt. Fruits normally present; seeds 0.8-1 mm long, 0.5-0.7 mm in diameter,
1.4-1.6 times as long as broad, dark green to dark reddish-brown when dry, with
clusters of long hairs 45-180 (xm long, the fine surface topography smooth. Tuber
iferous flowers absent. Chromosome number unknown. Figs. 6, 10, 22.
Phenology. Collected in flower and fruit throughout the year.
Distribution (Fig. 20). Northern South America and Central America, from
Amazonas, Brazil, west to northern Peru and Ecuador and north to Guatemala,
Belize, Trinidad, and French Guiana; not known from Panama or El Salvador.
Collections from Florida referred to this species (as N. blanda) by Ward (1977)
are N. jamesoniana.
Specimens Examined. Belize. Dist. Toledo, Crique Sarco, Hedger 326 (BM).?Costa Rica.
Prov. unkown: Marais de Cucurba, Estrella Calamanea, Fittier 9376 (BR, US); Monte Verde, Stork
1897 (F). Lim?n: Cahuita, Baker & Burger 26 (F, U). Puntarenas: Playa Blanca, Golfo Dulce, Valerio
408 (F); near Jalaca Station, between head of Golfo Dulce and Palmar Sur, Allen 6278 (F); Osa
Peninsula, Corcovado Nat. Park, Liesner 2892 (MO).?Guatemala. Izabal: near Puerto Barrios,
Standley 72857 (F), Standley 73128 (F); between Puerto Barrios and San Tom?s, 1 xh mi SE Puerto
Barrios, Steyermark 39867 (F).?Honduras. Atl?ntida: vicinity of Tela, Standley 55153 (F, US)?
Nicaragua. Zelaya: 2 km SW of Bluefields, Moreno 24967 (MO, UNA).
Brazil. Acre: Vicinity of Turauac?, France et al. 7350 (INPA, NY, UNA). Amazonas: km 5
Manaus-Porto Velho road, Berg et al. F17628 (INPA, NY, UNA); Jacaretinga, S of Rio Amazonas,
opposite Careiro Lake, France et al. 23302 (NY, UNA). Para: Cacaual Grande, Aguas Pr?tas, Black
52-15458 (IAN).?Colombia. Choc?: Trocha Coredo, Fuchs et al. 22176 (COL).?Ecuador. Los
R?os: km 56 of road Quevedo-Sto. Domingo, R?o Palenque Biol. Sta., Dodson 6172 (MO). Morona
Santiago: Lago de Kumpaka, 12 Aug 1982, Steinitz-Kannan s.n. (UNA), Steinitz-Kannan 409 (UNA).
Prov. unknown: Coast plain, Rimbach 86 (F), Rimbach 288 (US).?French Guiana. Near Kourou,
Schnell 11221 (P).?Guyana. Near Bartices, Essequibo River, Jenman 5746 (K, NY); locality un
known, Jenman 7773 (GH, K, NY), 1824, Parker s.n. (K).?Peru. Amazonas: Prov. Bagua, 208 km
by road NE of Chiriaco, Barbour 4408 (MO). Loreto: Iquitos, Killip & Smith 27173 (F, NY, US),
Asplund 14388 (G, S, US), Martin et al. 1614 (INPA).?Surinam. Brokopondo: Brownsweg, Cramer
LBB14915 (U). Marowijne: E of Perika River, van Donselaar 2548 (U); Maroni, 1901, Lem?e s.n.
(P). Para: Pararacweg at km 25 of road to Zanderij, Cramer LBB14902 (U). Saramacca: Zonneris
Creek near Bitagron, Cramer LBB14876 (U, VEN). Suriname: Lower Suriname R., Plantation Peper
pot, Cramer LBB14855 (U), Florsch?tz & Florsch?tz 1004 (U); Carolina, Middle Suriname R.,
Cramer LBB14856 (U); near Cassipora, Suriname R., Cramer LBB14857 (U); Domburg, Cramer
LBB14908 (NY, U); La Liberte, Beneden, Suriname R., Soeprato 179 (U). Unknown: valley of
Parepou Creek, Oldeman 2859 (U).?Trinidad. 1 mi S of St. Joseph, Hazen 2077 (GH, NY, US).?
Venezuela. Bolivar: Rio Uiri-yuk at El Foco, Maguire et al. 53507 (M, MY, VEN). Fed. Terr. Delta
Amacuro: Aniepo, Gines 4963 (US); Winiquina, Gines 5056 (US); Rio Amacuro, Sierra Imataca,
between Amacuro and mouth of Deadwater Creek Moat, Steyermark 87349 (F, NY).
FIG. 22. Nymphaea glandulifera (Hazen 2077). a. Undersurface of leaf. b. Opened flower, c.
Unopened flower bud. d. Upper surface of leaf.
Information on flowering responses in N. glandulifera comes from published

observations in Surinam (Cramer et al. 1975) and from data furnished by Walter
Pagels on Guyana plants in cultivation (Pagels, pers. comm.). Flowers reportedly
open for two successive nights. On both nights opening commences at dusk and
closure is completed by midnight, although flowers may remain open slightly longer
the first night. Only on the second night are the stamens exposed and, although not
so indicated, dehiscence probably occurs at this time, as in other members of subg.
Hydrocallis. The fruit production observed by Pagels in cultivation suggests that
self-pollination is likely in N. glandulifera and that the stigma remains receptive in
second-day flowers. Individuals of the scarab beetle Cydocephala castanea Olivier
were observed visiting flowers of N. glandulifera in Surinam.
The lack of evidence for stolon formation in N. glandulifera from herbarium
specimens or literature accounts implies that reproduction in this species occurs
entirely via seed production. The frequency with which maturing fruits are ob
served on herbarium specimens and are produced in cultivation indicates that
autogamy is probably involved in seed formation, as is the case in several other
species of subg. Hydrocallis.
On the basis of general morphology and floral biology, N. glandulifera ap
pears similar to N. conardii, N. gardneriana, and N. jamesoniana. These latter
three species normally exhibit a weblike leaf venation pattern that is not present
in N. glandulifera. Of the three, N. conardii is probably most closely related.
Both species have delicate leaves that are green above and below, lack stolons,
produce abundant seeds, and have very similar flowers. In addition to the differ
ence in leaf venation and some slight differences in leaf shape, the presence of
acicular sclereids in the stigmatic rays of N. conardii readily distinguishes it from
N. glandulifera. Also, the smooth seeds of N. glandulifera differ from the finely
granulate seeds of N. conardii. This feature can only be observed from seed testal
whole mounts with the aid of light microscopy. Certain features of N. glandulifera
seeds suggest an association with N. potamophila and N. oxypetala, two sagittate
leaved species. Further phytochemical and karyological study could provide addi
tional insight into this relationship.
The variation in pubescence which prompted the segregation of the glabrous
N. fenzliana, first as a distinct species and later as a variety of N. blanda, is not
considered to be worthy of taxonomic recognition. Similar variation in pubes
cence exists in N. tenerinervia, N. gardneriana, and N. amazonum.
The name Nymphaea blanda has been universally accepted as the correct
name of this species since its publication by Meyer in 1818; however, in publishing
N. blanda Meyer listed "N. glandulifera. Rodsch. Observ. p. 76" as a synonym
but rejected Rodschied's name as descriptively inappropriate. Meyer's citation
was a clear reference to Rodschied's Medizinische und chirurgische Bemerkungen
?ber das Klima, die Lebensweise und Krankheiten der Einwohner der hollaendi
schen Kolonie Rio Essequebo, where on page 76 N. glandulifera appears as
species number 89 with a short Latin description and some notes comparing it
with N. alba. Gerhard Wagenitz, who kindly supplied me with the relevant infor
mation from Rodschied's work in the University Library at GOET, agrees that N.
glandulifera was validly published by Rodschied. Therefore N. blanda must be
viewed as a superfluous name. Rodschied did not indicate a type specimen; his
collection (Rodschied 155) at GOET which is labelled N. glandulifera and con
forms to his description has been designated as the lectotype.
5. Nymphaea belophylla Trickett, Kew Bull. 26(1): 29. 1971.?Type: "a cultivated
specimen in the Kew Herbarium (K) grown at the Royal Botanic Gard
ens, Kew . . . introduced by A. Bleher from the river Guapor?, Brazil"

(holotype: K, not located; isotype: BM, not located).
Rhizome subglobose, sprouting freely. Leaf blades to 30 cm long, to 11 cm

wide, ca. 2 times as long as wide, elliptic-sagittate, entire, acute at apex, sagittate
at base, peltate 2 mm or less from base of sinus, the sinus ca. 0.5 times length of
blade, the lobes long-tapering acute, containing numerous trichosclereids and
acicular sclereids; upper and undersurface green; venation radiate centrally, re
ticulate toward margins, the principal veins 7-9. Petioles to 3 mm in diameter,
glabrous. Peduncles to 5 mm in diameter, glabrous. Flowers floating, the sepals
and petals in distinct whorls of four. Sepals ca. 3-6 cm long, ca. 1.5-2 cm wide,
green, oblong-ovate, acute to somewhat rounded at apex, containing acicular
sclereids, inconspicuously nerved. Petals 16 or 20, containing acicular sclereids,
the outer greenish abaxially, only slightly shorter than sepals, with apex acute to
rounded, the inner ca. 2.5-4.5 cm long, ca. 0.6-1.8 cm wide, becoming creamy
white or light yellow; transition to stamens gradual. Stamens ca. 138, creamy
white to light yellow, containing acicular sclereids, the outermost to 4.5 cm long,
with filaments to 8 mm wide and connectives prolonged 0.5 mm or more beyond
anthers, the inner with filaments to 1.5 mm wide and connectives prolonged to 0.2
mm beyond anthers; pollen diameter unknown. Carpels 29-37, the appendages
10-17 mm long, cream-colored to light yellow, slightly cl?vate, with enlarged tips
ca. 1 mm in diameter, the stigmatic rays with projecting acicular sclereids and
papillae cells remaining attached; axial process blunt. Fruits not seen; seeds
(found embedded in leaf on herbarium specimen) 0.9 mm long, 0.75 mm wide,
with rows of hairs 35-90 \xm long, the surface finely granulate. Tuberiferous
flowers absent. Chromosome number unknown. Fig. 23.
Phenology. Flowering and perhaps fruiting in August in Venezuela, otherwise
unknown.
Distribution. (Fig. 24). Uncertain at this time, but probably restricted to the
interior of the Amazon and Orinoco basins. A collection from Barinas, Venezu
ela?"alrededores de Puerto Nutrias," Trujillo 11430 (MY)?has been tentatively
assigned to this taxon.
Although the remote nature of the type locality precluded my visiting it, an
excursion to the Venezuelan locality was undertaken. A survey of the area in
September of 1981 failed to disclose any Nymphaea with the distinct sagittate
leaves of the earlier (August 1972) Trujillo collection. The only species en
countered in the immediate vicinity was N. conardii, which did not display sagit
tate leaves, nor have they appeared in subsequent cultivations of this material. A
closer inspection of the Trujillo specimen revealed the presence of a few seeds
embedded in the leaf tissue?seeds which appear similar (although slightly larger)
to seeds of N. gardneriana and are not those of N. conardii. As N. gardneriana
was not observed in the immediate area and seldom produces seeds, it is likely
that these seeds are those of N. belophylla.
The flowers are described by Trickett as being nocturnal and inodorous.
Flowers of the Venezuelan plants were observed by Trujillo to emit a disagree
able almondlike odor. Otherwise, flowering responses are unknown in this
species.
A great deal of uncertainty surrounds the type material of N. belophylla. An
X}m(>hata bclopkylla. i, leaf, X {; t,
SSIfeL* h gy:ioccium showing arrangement of
?wii, x 5. AH drawn from A. Blthtr. E
FIG. 23. Nymphaea belophylla. a. Photo of Trujillo 11430 (MY) from Barinas, Venezuela, b. Illustration repro
(Reproduced from Kew Bulletin, Vol. 26, by permission of the Controller of Her Majesty's Stationery Office.)
FIG. 24. Distribution of Nymphaea belophylla, N. potamophila, and N. oxypetala.
empty folder at Kew is labelled N. belophylla; however, the staff was unable to
locate the type specimen or any surviving material of the original 1962 cultivation
at the Royal Botanic Gardens, Kew (reported as E.N. 167-62). No evidence
could be found of any specimen having been deposited at BM. The possibility of a
chance recovery of the type material and the absence of additional collections
which can be assigned with absolute certainty to this taxon make the designation
of a neotype inadvisable at this time.
Further collections from the region of the type locality are desperately needed
to permit a more accurate interpretation of this taxon. In the absence of this
material, the assessment of several critical features missing from the type descrip
tion has been dependent upon the assignment of Trujillo 11430 to N. belophylla.
This collection, which includes a leaf, flower, and attached tuber (Fig. 23a),
appears identical to N. belophylla as illustrated by Trickett (Fig. 23b). Certain
sagittate-leaved Nymphaea that have been collected in South America have car
pellary appendages shorter than those indicated for N. belophylla. These further
differ from the presumed TV. belophylla of Venezuela in lacking acicular sclereids
in the leaf and stigmatic ray tissues, where such structures are abundantly present
in the Venezuelan plants. The above differences form the basis for the segrega
tion of these plants as N. potamophila (q.v., no. 6).
6. Nymphaea potamophila Wiersema, Sida 10: 195. 1984.?Type: Brazil. Amazo

nas: Rio Negro acima do lugar S?o Luis, pouco ?cima de Barcelos, 28 Jun
1979, Maia, Soares, Ramos & Mota 124 (holotype: TNPA!).
Rhizome not seen. Leaf blades to 20 cm or more long, to 10 cm or more wide,

ca. 1.8-2.5 times as long as wide, chartaceous, elliptic-sagittate, entire, acute
tapering to somewhat rounded at apex, sagittate at base, peltate 1-2 mm from base
of sinus, the sinus 0.5-0.65 times length of blade, the lobes acute-tapering; upper
surface green, often strongly variegated with red, when dried with numerous tiny
prominences from underlying trichosclereids, lacking acicular sclereids; undersur
face greenish to variegated with dark red, usually marked with slender, branched
rusty brown lines when dried; venation radiate centrally, scarcely evident toward
margins, the principal veins 7-11, somewhat impressed. Petioles to 3 mm in diame
ter, elongating up to 4 m or more, glabrous. Peduncles to 4 mm in diameter,
elongating up to 4 m or more, glabrous. Flowers floating, the sepals, petals, and
outer stamens in distinct whorls of four. Sepals ca. 4-5 cm long, ca. 1.5-2 cm wide,
green, oblong-ovate, acute at apex, when dry usually marked with slender brown
ish or blackish lines, with few to numerous acicular sclereids, inconspicuously
nerved. Petals ca. 16, containing acicular sclereids, the outer greenish abaxially,
slightly shorter and often narrower than sepals, with apex acute, the inner ca. 2.5
3.7 cm long, ca. 0.8-1.2 cm wide, becoming creamy white, acuminate; transition to
stamens gradual. Stamens ca. 40-60, creamy white, containing abundant acicular
sclereids, the outermost to 3 cm long, with filaments to 8 mm wide, the inner with
filaments to 1.4 mm wide and connectives prolonged to 0.2 mm beyond anthers, the
inner anthers 0.5-0.6 times as long as stamen; pollen diameter 25-35 \xm. Carpels
ca. 19-21, the appendages ca. 7-9 mm long, cream-colored, cl?vate, with enlarged
tips ca. 3 mm long, ca. 1-1.2 mm in diameter, apex blunt, the stigmatic rays lacking
acicular sclereids and with papillae cells remaining attached. Fruits usually present;
seeds 0.85-1.3 mm long, 0.7-1 mm in diameter, 1.1-1.3 times as long as broad,
with clusters of hairs 45-210 ^m long, the fine surface topography smooth. Tuberif
erous flowers absent. Chromosome number unknown. Figs. 1, 6, 25.
Phenology. Both flowers and fruits have been collected in June and in
October.
Distribution. (Fig. 24). Known from the states of Amazonas and Para in
northern Brazil; most collections are from river margins. The long petioles and
peduncles and the collection data indicate N. potamophila can survive in water as
much as 4-5 m deep.
Specimens Examined. Brazil. Amazonas: Rio Negro, Padauiry, Telheiro, Fr?es 22686 (IAN);
Rio Negro, Paran? Ararira, ca. 100 km above Barcelos, Madison et al. 6155 (SEL, NY). Para: S?o
Francisco do Curur? Mission, boca de Igarap? Kenebiit-Taibiri para o Rio Curur?, Sioli 4 (IAN).
No information on flowering responses is available except collectors' observa

tions that the plants are night-blooming. Both pollen viability and seed set in this
species appear to be high, as judged from the limited amount of herbarium
material available. Clearly, seed production is an important means of dispersion
in this species. No evidence is presently available to determine whether autogamy
contributes to seed production.
Although no karyological or phytochemical data have been gathered on N.
potamophila, on the basis of seed morphology it can be grouped with N. glandu
lifera and N. oxypetala. Flowers of N. potamophila are similar to those of N.
glandulifera, the latter having longer carpellary appendages and more numerous
FIG. 25. Nymphaea potamophila (leaf from Madison et al. 6155; flower from Maia et al. 124). a.
Undersurface of leaf. b. Opened flower.
acicular sclereids in the sepals. The floating sagittate leaves of N. potamophila are
reminiscent of N. oxypetala, which has sagittate submersed leaves, but floating
sagittate leaves of the latter species are reduced both in number and size. The
poorly known N. belophylla possesses sagittate floating leaves resembling those of
N. potamophila. Flowers of N. belophylla, as described by Trickett (1971), appear
to have carpellary appendages approaching twice the length of those in N. po
tamophylla. A plant assigned to N. belophylla, Trujillo 11430 (MY) from Barinas,
Venezuela, indeed displays these longer carpellary appendages. This plant is fur
ther distinguished in an abundance of acicular sclereids in the leaf and stigmatic
ray tissues, where these cells are completely absent in N. potamophila. If this
Venezuelan plant is in fact N. belophylla, the observed character differences
support its taxonomic separation from N. potamophila, in view of the conserva
tive nature of these characters among other species of subg. Hydrocallis. Accord
ingly, the two taxa have been segregated at the rank of species. However, addi
tional collection and study of these species in the South American interior are
clearly necessary to provide a better understanding of their relationship. Indeed,
the existence of a plant collected by Garc?a-Barriga (20813 at COL, GH) in
southeastern Colombia (state of Guainia) with leaves like N. potamophila but
flowers tending toward the presumed Venezuelan N. belophylla emphasizes this
need for further research.
7. Nymphaea oxypetala Planch?n, FI. Serres Jard. Eur. 8: 120. 1852. Nymphaea
raja Lehmann, Hamburger Garten- Blumenzeitung 9: 196, 208. 1853,
nom. superfl. Leuconymphaea oxypetala (Planch?n) Kuntze, Revis, gen.
pi. 1: 11. 1891.?Type: Ecuador, "in inundatis circa Guayaquil," Mar
1845, Jameson 334 (holotype: G!; isotypes: B! BM! K!). The specimen at
B (from the herb. Lehmann) contains label of Parker from Guyana and
no information of Jameson (see discussion).
Rhizome subglobose, stolons probably lacking. Blades of floating leaves to 6

cm long, to 2.5 cm wide, 1.9-2.5 times as long as wide, firm, sagittate, entire,
acute to somewhat rounded at apex, scarcely or not at all peltate from base of
sinus, the sinus 0.35-0.42 times length of blade, the lobes acute-tapering, lacking
evident sclereids; upper surface green; undersurface purplish; principal veins 7-9.
Blades of submersed leaves to 40 cm long, to 33 cm wide, 1-1.8 times as long as
wide, submembranaceous to slightly rigid, broadly sagittate, entire, rounded to
somewhat obtuse at apex, peltate 1 mm or less from base of sinus, the sinus 0.5
0.7 times length of blade, the lobes acute to rounded, lacking trichosclereids or
acicular sclereids; upper surface green; undersurface purplish; venation radiate
centrally, dichotomous toward margins, the principal veins 7-12. Petioles to 4 mm
in diameter, glabrous, with 2 major and several minor air canals. Peduncles to 5
mm in diameter, glabrous, with 4 major air canals. Flowers floating, the sepals
and only outermost petals in distinct whorls of four. Sepals 5.4-13.5 cm long, 1
2.5 cm wide, green, lanceolate, long-acuminate at apex, lacking acicular sclereids,
inconspicuously nerved. Petals 16-34, the outer greenish abaxially, mostly shorter
and narrower than sepals, lacking acicular sclereids, with apex acuminate, the
inner 3.7-8 cm long, 0.7-1.7 cm wide, becoming creamy white, containing acicu
lar sclereids, acuminate; transition to stamens gradual. Stamens 66-111, heavily
suffused with purple, containing acicular sclereids, splitting along three longitudi
nal sutures adjacent to anther sacs to release tiny spherical "sclerites," the
outermost to 6.5 cm long, with filaments to 16 mm wide and connectives pro
longed 1.5 mm or more beyond anthers, the inner with filaments 1-2.4 mm wide
and connectives prolonged to 1 mm beyond anthers, the inner anthers 0.58-0.73
times as long as stamen; pollen diameter ca. 40 \im. Carpels 21-27, the append
ages 13-29 mm long, cream-colored, abruptly bent at distal %, linear-clavate, with
enlarged tips 6-20 mm long, 1-2 mm in diameter, the stigmatic rays shortened
axially, lacking acicular sclereids, with papillae cells remaining attached; axial
process absent. Fruits normally present; seeds 0.85-1 mm long, 0.7-0.85 mm in
diameter, 1.1-1.25 times as long as broad, with clusters of hairs 70-120 [xm long,
the fine surface topography smooth. Tuberiferous flowers absent. Chromosome
number: 2n = 84. Figs. 1, 6, 10, 12, 26.
Phenology. Flowers have been collected during March in Ecuador, August in
Venezuela, and April and July in southern Brazil; fruits were obtained during
August in Venezuela and July in Brazil.
Distribution. (Fig. 24). Collected in Ecuador, Venezuela, and southern
Brazil. Material of N. oxyp?tala is also present in mixed collections of Nymphaea
from Cuba and Paraguay, but until additional collections from both areas are
obtained the accuracy of these records is questionable. Attempts at cultivation of
N. oxypetala (from Wiersema & Gonz?lez 2232) indicate a low level of tolerance
to increased turbidity or salinity and decreased temperatures. Although the small
number of collections of N. oxypetala provides little ecological information, this
submersed species is probably best adapted to clean, gently flowing freshwater
habitats.
Specimens Examined. Cuba. Habana: Hoyo Colorado, brook near Laguna Ariguanabo, Ekman
13710 (S, flower only, leaf is N. amazonum).-Brazil. Mato Grosso: Biarra do Rio dos Bugres,
Lindman 2877 (S). Mato Grosso do Sul: Corumba, Malme 1764B (S).?Ecuador. Prov. unknown: in
num. Daule, Mille 1000 (F).?Paraguay. Gran Chaco: Santa Elisa, Hassler 2804 (G, K, both with
flower of N. oxypetala mounted with material of N. prolif?ra).?Venezuela. Apure: near San Fer
nando de Apure, Ramia & Montes 5362 (VEN); Guasdualito, Fern?ndez 1299 (MY), Wiersema &
Gonz?lez 2232 (F, MO, MY, NY, VEN, UNA). Cojedes: R?o Portuguesa, Delascio et al. 4493
(VEN).
No information on flowering responses of N. oxypetala is available. The

flowers reportedly emit a strong etherlike aroma. They are unique among Nym
phaea in their production and release of tiny spherical "sclerites" from their
stamens. This phenomenon was examined anatomically by Chifflot (1902); how
ever, its functional importance has not been determined. Reproduction in N.
oxypetala may be entirely dependent on seed production, as no evidence of stolon
formation has been observed.
Nymphaea oxypetala, with its habit and unique morphology, probably repre
sents the most specialized member of subg. Hydrocallis. On the basis of leaf and
seed morphology, its closest taxonomic relative appears to be N. potamophila,
another sagittate-leaved species. Similarities in seed morphology also suggest a
link with N. glandulifera. Unfortunately, no phytochemical or karyological data
are available on these other two species, data which could provide valuable in
sight into their relationship to each other and to N. oxypetala. An almost total
FIG. 26. Nymphaea oxypetala (Wiersema & Gonz?lez 2232). a. Undersurface of subm
b. Upper surface of floating leaf. c. Opened flower, d. Cross section of petiole, e. Cro
peduncle.
lack of information on the sagittate-leaved N. belophylla prevents an evaluation

of its relation to N. oxypetala. The submersed sagittate leaves and long-acuminate
sepals and petals of N. oxypetala readily distinguish it from other species of
Nymphaea.
The name N. raja Lehmann is superfluous; Lehmann cited the type of N.
oxypetala. His citation of type material reads: "in Columbia collegit Cl. Jameson,
in Chile filius Guilielmus Lehmann." The name N. raja, handwritten by Leh
mann, appears on one of two sheets of the 1845 Jameson collection from Guay
aquil (until 1830 included in Great Columbia) at K and on a specimen from
Lehmann's own herbarium (now at B). This latter specimen bears no information
of Jameson and is now labelled "Mr. Parker, Guiana." Caspary (1878), who
viewed this specimen, makes no mention of the Parker label. He does indicate,
however, that he was unable to find evidence of a collection from Chile by
Lehmann's son either in Lehmann's herbarium or in any other herbarium. As
Parker's collections from Guyana in other herbaria have never included N. oxype
tala (which has never been found in Guyana) and the Jameson collection of this
species is the only one prior to 1890 in existence, Lehmann's specimen is undoubt
edly material collected by Jameson in Ecuador.
8. Nymphaea novogranatensis Wiersema, Brittonia 36: 217. 1984.?Type: Vene

zuela. Guarico: small depression along road from hwy 13 near El Som
brero to Barbacoas, 13 km N of junction with hwy 13, 31 Aug 1981,
Wiersema & Gonz?lez 2234 (holotype: MO!; isotypes: COL! F! MY! NY!
UNA! VEN!).
Rhizome ovoid, stolon-forming only when germinating. Leaf blades to 24 cm

long, to 21 cm wide, 1.05-1.25 times as long as wide, subcoriaceous to almost
skinlike in texture, broadly elliptic-ovate to suborbicular, entire, truncate or more
frequently obtuse-mucronulate at apex, peltate 2 mm or less from base of sinus,
the sinus 0.36-0.46 times length of blade, the lobes obtuse to rounded with slight
protuberance; upper surface glossy-green with numerous purple flecks evident on
younger leaves, when dried usually with numerous tiny prominences and short
elevated lines from underlying trichosclereids and acicular sclereids; undersurface
brownish purple, with dark flecks especially evident on younger leaves; venation
radiate and manifest centrally, faintly reticulate toward margin, the principal
veins 13-22. Petioles to 5 mm in diameter, glabrous, with 2-4 major and 6-8
minor air canals. Peduncles to 7.5 mm in diameter, glabrous, with 5, 6, 15, or 18
major air canals. Flowers floating or slightly emergent, with sepals, petals, and
outermost stamens in distinct whorls of 4. Sepals 3-6 cm long, 1.1-3.4 cm wide,
uniformly green or somewhat brownish purple, ovate, acute to somewhat rounded
at apex, containing acicular sclereids, the nerves slightly evident basally. Petals 16
or 20, white, lacking sclereids, those of the 3 outer whorls broader than inner and
oblong-ovate, the outermost shorter than sepals and acute, those of the inner
whorls 2.5-3.5 cm long, 0.7-1.4 cm wide, lanceolate, acuminate; transition to
stamens gradual. Stamens 53-132, cream-colored, lacking sclereids, the outer
most to 3.2 cm long, with filaments to 12 mm wide and connectives prolonged ca.
1 mm or more beyond anthers, the inner with filaments to 1.1 mm wide, some
times slightly purplish at base, and with connectives prolonged to 0.7 mm beyond
anthers, the inner anthers 0.6-0.9 times as long as stamen; pollen diameter 28-40
\im. Carpels 15-32, the appendages 5-9 mm long, 0.7-1.5 mm in diameter, deep
crimson, cylindric-clavate, with apex slightly incurved and somewhat apiculate,
the stigmatic rays lacking acicular sclereids and with papillae cells separating into
powdery mass; axial process acute or blunt. Fruits normally present; seeds 0.8-1
FIG. 27. Nymphaea novogranatensis (Wiersema & Gonz?lez 2234). a. Undersurface of l

Opened flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross section of pedun
Upper surface of leaf.
mm long, 0.55-0.75 mm in diameter, 1.3-1.6 times as long as broad, with lo

dinal rows of hairs 20-90 |im long, the surface finely granulate. Tuberi
flowers absent. Chromosome number: 2n = 28. Figs. 2, 5, 10, 12, 13, 27.
Distribution. (Fig. 28). Known only from Venezuela and Colombia; co
in stagnant water of both coastal and inland habitats.
:.*f?. 70? ~ 69?.50?.? .
FIG. 28. Distribution of Nymphaea novogranatensis, N. tenerinervia, and N. prolif?ra.
Specimens Examined: Colombia. Bolivar: Soplaviento and vicinity, Killip & Smith 14613 (F, GH,
NY, US). Magdalena: Isla de Salamanca, Los Almendros, Romero-Casta?eda 10869 (COL).?
Venezuela. Falc?n: between Chimpire & El Guay, road Coro to Tucacas, Prance 23722 (NY, UNA);
vicinity of Tucacas, Prance 23724 (NY, UNA); Paraguana, Tamayo 701 (VEN); 1.5-2 km N of
Tucacas, Steyermark & Manara 110570 (NY, VEN); near dam of Jatira, environs of Tucuyo de la
Costa, Trujillo 13467 (F, MY). Gu?rico: environs of El Sombrero, Trujillo 2538 (MY); km 2 between
crossing at Barbacoas and crossing at El Sombrero, Trujillo 11329 (MY); km 10 from the crossing La
Guamita (El Sombrero-Chaguaramas) and Barbacoas toward Barbacoas, Trujillo 11358 (MY); off hwy
2, ca. 2 km S of R?o Gu?rico near Calabozo, Wiersema 1404 (VEN); 25 km S Calabozo, Wiersema
1411 (NY, UNA, US, VEN); 34.8 km S of R?o Gu?rico at Calabozo, Wiersema 1420 (BM, COL, MO,
UNA, VEN).
Observations of Venezuelan material (Wiersema & Gonz?lez 2234) in cultiva

tion indicate that flowers open for two consecutive nights. Flowering responses
are similar to those of the morphologically similar N. tenerinervia. First-day
flowers open only enough to expose the fluid-filled stigma, the sepals and outer
petals reflexing ca. 45 degrees. Opening commences ca. 1-2 hours after dark with
complete closure by midnight of the first night. Second-day flowers begin opening
at dusk, the sepals and petals gradually spreading to 180 degrees. Although the
stamens dehisce in the early evening, they are not exposed until ca. 1-2 a.m. and
continue spreading thereafter. Second-day flowers close at dawn and begin to be
retracted under water. Flowers both nights are only faintly odorous. Although the
stigma is fairly dry by the second night, it is still receptive to pollen, as some
flowers in cultivation produced seeds autogamously. This autogamous seed pro
duction was observed in only a few flowers, however, with the majority of flowers
failing to set seed in the absence of cross-pollination.
No evidence of asexual reproduction exists for N. novogranatensis, which
apparently relies entirely on seed production for its dispersion. The data obtained
indicate that this species is less efficient in autogamous seed production than other
more widespread species of subg. Hydrocallis, such as N. jamesoniana or N.
conardii. This factor may be partly responsible for the restricted distribution of TV.
novogranatensis, which is thus more dependent on pollinating organisms for suc
cessful reproduction.
Nymphaea novogranatensis is most closely related to N. tenerinervia, the two
having almost identical floral morphology and biology. In addition to the differ
ences presented in the key, N. novogranatensis differs in having glabrous petioles
and lacking stolons (vs. glabrate or pubescent petioles and stolon formation in N.
tenerinervia). The mucronulate leaf apices of N. novogranatensis serve to distin
guish it from all other members of the subgenus, which have rounded or truncate
leaf apices.
9. Nymphaea tenerinervia Caspary in Martius, Fl. bras. 4(2): pi. 36. 1878.?Type:
Brazil. Bahia: "prope Juazeiro ad flumen St. Francisco," April 1819,
Martius s.n. (lectotype here designated: M!, mounted on 2 sheets; isolec
totype: B?, see discussion).
Rhizome ovoid, often stolon-forming throughout vegetative period. Leaf

blades to 21 cm long, to 19 cm wide, 1.05-1.26 times as long as wide, somewhat
coriaceous, elliptic-cordate, entire, obtuse to rounded at apex, cordate at base,
peltate 1-4 mm from base of sinus, the sinus 0.34-0.44 (-0.48) times length of
blade, the lobes rounded, with little or no protuberance; upper surface green or
occasionally reddened, often with purple flecks, when dried usually with numer
ous tiny prominences from underlying trichosclereids, also containing forked
acicular sclereids; undersurface purple or red-purple, often with short dark flecks;
venation radiate but scarcely evident centrally, the veins somewhat impressed, the
principal veins 13-19. Petioles to 6 mm in diameter, glabrate or slightly pubes
cent, with 4 or 8 major and 8-12 minor air canals. Peduncles to 9 mm in diame
ter, glabrate or slightly pubescent, with 15 or 18 major air canals. Flowers floating
or slightly emergent, the sepals, petals, and outer stamens in distinct whorls of
four. Sepals 3.5-7.7 cm long, 1.4-3.3 cm wide, green, usually with short purple
streaks, ovate, rounded or somewhat acute at apex, containing acicular sclereids,
inconspicuously nerved. Petals 16 or 20, creamy white to light yellow, lacking
sclereids, those of the three outer whorls broader than inner and oblong-ovate,
the outermost shorter but often wider than sepals, with apex rounded or some
what acute, those of the inner whorls 2.6-4.6 cm long, 0.6-1.4 cm wide, lanceo
late, acute to acuminate; transition to stamens gradual. Stamens 40-100, creamy
white to light yellow, lacking sclereids, the outermost to 4.3 cm long, with fila
ments to 8.5 mm wide and connectives prolonged nearly 1 mm or more beyond

anthers, the inner with filaments to 1.5 mm wide, often suffused with deep purple
at base, and with connectives prolonged to 0.4 mm beyond anthers, the inner
anthers 0.7-0.8 times as long as stamen; pollen diameter 25-35 |im. Carpels 15
25, the appendages 4-10 mm long, 0.7-1.2 mm in diameter, deep purple, cylindric
clavate, with apex slightly incurved and somewhat apiculate, the stigmatic rays
lacking acicular sclereids and with papillae cells separating into powdery mass;
axial process acute or blunt. Fruits only occasionally present; seeds 0.8-1 mm
long, 0.6-0.7 mm in diameter, 1.3-1.5 times as long as broad, with longitudinal
rows of hairs 55-90 [xm long, the surface finely granulate. Tuberiferous flowers
absent. Chromosome number: 2n = 20. Figs. 4, 10, 12, 29.
Phenology. Collections of flowering material have been made throughout the
year; fruiting probably occurs throughout the year as well, as fruiting specimens
have been obtained both in June and November.
Distribution. (Fig. 28). Known only from northern Brazil, mostly north of
Bahia and northern Goi?s and east of the Amazon Basin, although also collected
in Roraima. This species is especially common in natural swamps and marshes in
northern Piaui and Cear?.
Specimens Examined. Brazil. Cear?: BR-222 at km 16,16 mi W Fortaleza, Wiersema et al. 2314
(GUA, IPA, MO, UNA). Goi?s: Brejinho de Nazare, entre Fatima e Alian?a do Norte, 2 km N de
Rio Crixas, Camargo de Abreu 25 (NY, SP). Maranh?o: BR-316 at km 598 between Teresina and
Caxias, Wiersema et al. 2307 (INPA, IPA, UB, UNA). Para: 12 km de Bragan?a, Maciel & Bou?as 43
(IAN). Piaui: BR-343 Teresina-Sobral at km 303, Rio Suruben, Wiersema et al. 2310 (IPA, NY, SP,
UNA); BR-343 at km 260 E of Campo Meior, Rio Jenipapo, Wiersema et al. 2311 (IPA, UNA).
Roraima: Rio Branco, Surumu, Ule 8120 (G, K, not B).
Observations of Brazilian material (Wiersema et al. 2307) in cultivation indi

cate that flowering responses in this species are very similar to those of N. novo
granatensis, as is the flower morphology. Flowers open two consecutive nights.
First-day flowers open only enough to expose the fluid-filled stigma, with only the
sepals and outer petals reflexed. Opening occurs about one hour after dusk, with
complete closure by midnight of the first night. Second-day flowers begin opening
at dusk, remaining open until dawn. Sepals and petals of second-day flowers are
reflexed more than 180 degrees when flowers are fully open a few hours after
dark. The stamens have not reflexed by midnight but have already dehisced; they
become fully exposed before dawn when closure is effected and the flowers begin
to be retracted under water. Flowers observed in outdoor cultivation were only
faintly aromatic. In view of its flowering schedule, successful cross-pollination
among N. tenerinervia is most likely during the two to three hours before mid
night that first-day flowers are accessible to potential pollinators. No evidence of
self-pollination was observed in this species.
Propagation in most populations appears to be the result of stolon formation.
Numerous elongate stolons are produced from individual plants in cultivation.
Seed production was observed in only a few natural populations. As autogamous
seed production was not observed in cultivation and seeds were produced follow
ing artificial pollination, seed production in natural populations may depend on
outcrossing and thus on the availability of pollinators.
The relationships of N. tenerinervia to other species of subg. Hydrocallis are
FIG. 29. Nymphaea tenerinervia (Wiersema et al. 2314). a. Undersurface of leaf. b. Opened
flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross section of peduncle, f. Upper
surface of leaf.
interesting from an evolutionary point of view, as N. tenerinervia may represent a

link between two groups of species. With respect to its floral morphology and
biology it is obviously related to N. novogranatensis of the 2n = 28 phyletic line. On
the basis of phytochemistry, TV. tenerinervia is most closely allied to TV. amazonum
of the 2n = 18 line. The 2n = 20 observed for TV. tenerinervia supports its position
between these two groups. A unique anatomical feature observed in TV. tenerinervia
renders identifications of this species relatively easy. Nymphaea tenerinervia is the
only Nymphaea known to have forked acicular sclereids in the leaf mesophyll.
These structures are readily visible both in living or in dried leaf material.
The spelling "tenuinervia" accompanies Caspary's description; however, N.
tenerinervia appears with the plate, in two other places in Caspary's manuscript,
and on his 1876 annotation of the type. This latter spelling was also adopted by
Conard (1905). Caspary's description of TV. tenerinervia was based on material
from the Munich herbarium (now mounted on two sheets at M, one with two
opened flowers and the other with two leaves and an unopened flower) and from
the herbarium Lehmann. Two specimens at B from the herbarium Lehmann
contain material of N. tenerinervia. A flower of this species is mounted with a leaf
fragment of TV. lasiophylla, but Caspary's notes are absent from this specimen,
which is labelled TV. lasiophylla and contains no collector information. A second
specimen, labelled N. maximiliani Lehmann and also lacking collection informa
tion, contains a leaf of N. tenerinervia (so annotated by Caspary) and a flower of
TV. odorata. Caspary (1878), in rejecting Lehmann's name, indicated that Leh
mann's description was based on this specimen, although features described by
Lehmann are missing from the specimen. Caspary suggested that Lehmann may
have fabricated portions of his description, and furthermore accused Lehmann of
purloining the TV. tenerinervia material from Martius's herbarium. Caspary indi
cates that Martius himself interrogated Lehmann on this matter, but that Leh
mann stated that he had received the then unidentified specimen from Dr. Rudol
phi of Ratzeburg. It seems likely that Caspary's assertions are probably correct, as
the leaf fragment of Lehmann's N. lasiophylla specimen compares remarkably
well with the leaf portion missing from Martius's collection of this species at M. In
citing TV. tenerinervia material in Lehmann's herbarium, Caspary, although not
clearly indicating such, was probably referring to the leaf represented in the type
specimen of TV. maximiliani, although the origin of this specimen cannot be estab
lished with absolute certainty.
To preserve Caspary's segregation of TV. tenerinervia from the previously pub
lished TV. maximiliani requires lectotypification of the ambiguous TV. maximiliani
on its TV. odorata element (see Doubtful and Excluded Names).
10. Nymphaea amazonum Martius & Zuccarini, Abh. Math.-Phys. Cl. K?nigl.
Bayer. Akad. Wiss. 1: 363. 1832. Nymphaea blanda ? amazonum (Mar
tius & Zuccarini) Planch?n, FI. Serres Jard. Eur. 11: 21. 1856. Nymphaea
rudgeana ? amazonum (Martius & Zuccarini) Grisebach, Fl. Brit. W.
Ind. 12. 1864. Leuconymphaea amazonum (Martius & Zuccarini) Kuntze,
Revis, gen. pi. 1: 11. 1891. Castalia amazonum (Martius & Zuccarini)
Britton & P. Wilson, Bot. Porto Rico 5(2): 305. 1924.?Type: Brazil.
Para: "in aquis stagnantibus prope urbem Para," May-June (probably
1819), Martius exsicc. 3313 (holotype: M!, mounted on 3 sheets).
Rhizome to 4 cm or more in diameter, ovoid to cylindric, sprouting freely,

often stolon-forming throughout vegetative period. Leaf blades to 32 cm long, to
26 cm wide, 1-1.5 times as long as wide, subcoriaceous, broadly ovate to elliptic,
entire, rounded at apex, peltate 1-8 mm from base of sinus, the sinus 0.32-0.5
times length of blade, the lobes slightly tapering to broadly rounded; venation
radiate centrally, reticulate toward margins, the veins slightly to evidently raised,
the principal veins 13-21. Petioles to 8 mm in diameter, glabrous or occasionally
pubescent, the apex usually with ring of long septate hairs, with 2 major and 6
minor air canals. Peduncles to 10 mm in diameter, glabrous or rarely with ring of
pubescence at apex, with 6 or 18 major and 12 or 24 minor air canals. Flowers
floating, the sepals, petals, and outer stamens in distinct whorls of four. Sepals
green with short black streaks, ovate, rounded or somewhat acute at apex, lacking
acicular sclereids, the nerves often faintly evident. Petals 16, 20, or 24 (28),
creamy white, lacking acicular sclereids, the outer shorter and often narrower
than sepals, with rounded apex, the inner 2.5-5.9 cm long, 1-2.2 cm wide, acute
to rounded; transition to stamens gradual. Stamens 84-324, creamy white, with
acicular sclereids rare and very fine, the outermost to 5.2 cm long, with filaments
to 20 mm wide and connectives prolonged to 1.2 mm beyond anthers, the inner
with filaments to 1.3 mm wide and connectives prolonged 0.4 mm or less beyond
anthers, the inner anthers 0.5-0.85 times as long as stamen. Carpels 20-47, the
appendages 5-20 mm long, cream-colored or becoming slightly pinkish or pur
plish, strongly incurved, with tips strongly cl?vate, the enlarged portion 2-6 mm
long, (1.1-) 1.4-3 mm in diameter, the stigmatic rays lacking acicular sclereids
and with papillae cells separating into powdery mass; axial process to 3 mm long,
acute to widely obtuse. Seeds 0.95-1.35 mm long, 0.5-0.85 mm in diameter, 1.4
1.9 times as long as broad, with longitudinal rows of hairs 90-190 \im long, the
surface finely granulate. Tuberiferous flowers absent.
Nymphaea amazonum, on the basis of phytochemistry, general morphology,
seed morphology, and chromosome number, appears most closely allied taxo
nomically to N. prolif?ra and N. tenerinervia. Putative hybrids between N. amazo
num and both N. prolif?ra and N. lasiophylla have been observed. Differences in
morphology, reproductive biology, and flavonoid chemistry support the recogni
tion of at least two subspecies of N. amazonum. One of these, N. amazonum
subsp. pedersenii, displays some phytochemical similarity to N. rudgeana.
Key to the Subspecies of Nymphaea amazonum
1. Leaf blades 1-1.35 times as long as wide, with trichosclereids dense on upper surface,
acicular sclereids absent from mesophyll; stamens mostly less than 200.
10a. N. amazonum subsp. amazonum.
1. Leaf blades 1.25-1.5 times as long as wide, with trichosclereids sparse on upper surface,
acicular sclereids often present in mesophyll; stamens usually more than 225.
10b. N. amazonum subsp. pedersenii.
10a. Nymphaea amazonum subsp. amazonum.

Nymphaea albo-viridis Saint-Hilaire, Voy. distr. diam. 2: 426. 1833.?Type:
Brazil. Rio de Janeiro: "dans les marres d'eau voisines de Taguahy,"
1816-1821, Saint-Hilaire s.n. (holotype: P!, mounted on 3 sheets).
Nymphaea goudotiana Planch?n, FI. Serres Jard. Eur. 8: 120. 1852. Nym
phaea amazonum f. goudotiana (Planch?n) Caspary in Martius, Fl. bras.

4(2): 169-170. 1878. Leuconymphaea goudotiana (Planch?n) Kuntze, Re
vis, gen. pi. 1: 11. 1891. Nymphaea amazonum var. goudotiana (Plan
ch?n) Conard, Publ. Carnegie Inst. Wash. 4: 204. 1905.?Type: Co
lombia. "Llanos d'lbag" [probably collected near Ibaqu? in the state of
Tolima], 1844, Goudot s.n. (holotype: P!).
Nymphaea candolleana Lehmann, Hamburger Garten- Blumenzeitung 9: 203.
1853.?Type: Surinam. Splitgerber s.n. (lectotype, here designated: B!).
Leaf blades to 32 cm long, to 26 cm wide, 1-1.35 times as long as wide,

broadly ovate-elliptic, peltate 1-8 mm from base of sinus, the sinus 0.32-0.5
times length of blade; upper surface green with purple spots, when dried densely
covered with tiny prominences from underlying trichosclereids, lacking acicular
sclereids; undersurface purple with small dark flecks; venation slightly raised
centrally, the principal veins 13-21. Petioles to 8 mm in diameter, the apex with
ring of pubescence, the remainder glabrous or occasionally pubescent. Peduncles
to 10 mm in diameter, glabrous or rarely with ring of pubescence at apex. Sepals
3-8 cm long, 1.2-3 cm wide, rounded or somewhat acute at apex, the nerves
often faintly evident. Petals 16, 20, or 24 (28), the outer shorter and often nar
rower than sepals, the inner 2.5-5.5 cm long, 1-2.2 cm wide, with apex acute to
rounded. Stamens 84-201 (-250), the outermost to 5 cm long with filaments to 20
mm wide and connectives prolonged less than 0.5 mm beyond anthers, the inner
with filaments to 0.9 mm wide and connectives prolonged 0.1 mm or less beyond
\im. Carpels 20-38, the appendages 5-12 (-16) mm long, the cl?vate tips 2-5 mm
long, (1.1-) 1.4-2.3 mm in diameter; axial process to 3 mm long, acute. Fruits
nearly always present due to autogamous seed production. Chromosome number:
2n = 18. Figs. 4, 10, 30a-f.
Distribution. (Fig. 31). Common in tropical belt of South America and Carib
bean Islands, particularly in lowlands adjacent to coastal areas; rare in Central
America and Mexico; mostly in standing water, sometimes in slightly brackish
water.
Representative Specimens. Cuba. Habana: Guatao, Le?n 11522 (NY); Laguna de Perdig?n,
Coralillo, Le?n & Nivaro 7609 (NY); Hoyo Colorado, near Laguna Ariguanabo, Ekman 13710 (S).?
Dominican Republic. Saman?: Peninsula de Saman?, S?nchez, Ekman 15949 (G, S).?Guadeloupe.
Pointe Noire, Duss 3702 (F, GH, MO, NY, US); Vieux-Bourg, Raynal-Roques & J?r?mie 21101
(P).?Haiti. Dept. du Nord, Bayeux, between Bord de la Mer and Morne ? Madame, Ekman 2680
(S).?Jamaica. Locality unknown, March 25 (GH, K).?Martinique. Le Marin, entre Cap Ferr? et
Barri?re la Croix, Raynal-Roques & J?r?mie 21169 (P).?Puerto Rico. Punta Santiago, Stevens 2463
(NY); Sauturee, Britton & Britton 7086 (NY); Sabana Atiajo, Britton & Britton 9234 (NY); Vega
Baja, Britton et al. 6762 (F, NY, US).
Mexico. Tabasco: Laguna del Corcho, near Ocuiltzapotl?n, Rovirosa 674 (NY, PH, US).
El Salvador. Laguna de Zapotitan, Fassen 29326 (WIS).?Guatemala. Peten: near Flores,
Hedger 75 (BM).?Nicaragua. Chontales: road from Acoyapa to San Carlos, 2 km NW of Rio Oyate,
Haynes 8593 (MO, UNA).?Panama. Canal Zone: Col?n, Lehmann 1007 (BM). Panam?: road be
tween Panam? and Chepo, Dodge et al. 16713 (MO).
Bolivia. Santa Cruz: Santa Cruz, Kuntze 340 (NY).?Brazil. Acre: 4 km from Brasileia, Lowrie
et al. 720 (INPA). Alagoas: BR-316, 1 km E Maribondo, Wiersema et al. 2291 (IPA, UNA). Bahia:
FIG. 30. Nymphaea amazonum. a-f. N. amazonum subsp. amazonum (Wiersema et al. 2304). a.
Undersurface of leaf. b. Opened flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross
section of peduncle, f. Upper surface of leaf, g, h. N. amazonum subsp. pedersenii (Wiersema et al.
2270). g. Undersurface of leaf. h. Upper surface of leaf.
Lac da Con?eic?o, Blanchet 11 (G); 10 km NE Salvador, Lagoa de Abaete, Anderson 224 (INPA).
Cear?: Mun. Fortaleza, Barra do Cear?, Drouet 2511 (F, GH); km 204 of BR-222 between Sobral
and Fortaleza, Wiersema et al. 2312 (IPA, UNA). Maranh?o: BR-316 at km 613, 8 km W of Rio
Parnaiba, Wiersema et al. 2304 (IPA, UNA). Mato Grosso: Mun. Pocon?, km 62 Rodovia Transpanta
neira, Rio Pixaim, Macedo et al. 452 (INPA), km 63, da Silva 452 (UFMT). Minas Gerais: Mun. Santa
FIG. 31. Distribution of Nymphaea amazonum subsp. amazonum and subsp. pedersenii.
Luzia, Lagoa Santa, Warming 1750 (C), Barreto 7230 (BHMH). Para: Mun. Soure, Porto Santo
Antonio, Black 50-9086 (IAN); Rio Cupari, Fl?chai, Black 47-2105 (IAN). Pernambuco: Dois Ir
mois, near Recife, Wiersema et al. 2284 (IPA, UNA). Piaui: Lake at Parnagua, Gardner 2474 (BM);
BR-316 at km 240 between Picos and Teresina, Wiersema et al. 2303 (IPA, UNA). Rio de Janeiro:
Mun. Cabo Fr?o, Pontal Beach, Smith et al. 6647 (NY, US), Restinga 1307 (NY); Mun. S?o Jo?o da
Barra, between S?o Jo?o da Barra and Atafona, Restinga 418 (US); Mun. Campos, Lagoa Feia,
Araujo & Maciel 3822 (GUA). Rond?nia: Rio Guapor?, Faz. Cruz Verde, Black & Cordeiro 52
15053 (IAN).?Colombia. Cauca: Locality unknown, Holton 689 (G, GH, NY, PH); Valles du Cauca,
1851-1857, Triana s.n. (BM, K, P). Magdalena: Valles de Magdalena, Triana 5020 (COL). Tolima:
llanos de Ibaqu?, 1851-1857, Triana s.n. (BM). Valle: entre Jamundi y R?o Pance, Cuatrecasas 19656
(COL, F, G); km 31 of hwy Cali-Popayan, Maas & Plowman 2159 (COL, U, VEN) .?Ecuador.
Guayas: hwy 25, 21 km S Jujan, Wiersema et al. 2027 (UNA, USF); hwy 25, 8 km N Guayaquil,
Wiersema et al. 2029 (AAU, F, NY, UNA, US, USF). Morona-Santiago: Lago de Kumpaka, 12 Aug
1982, Steinitz-Kannan s.n. (UNA).?French Guiana. Sinnamary, Raynal-Roques 19861 (P); Kaw, 60
km SE de Cayenne, Raynal-Roques 21545 (P).?Guyana. Coast Canals, Jenman 5747 (BM, K, NY,
U); dam road from Mon Repos to Water Conservancy, Cramer 18 (U); Georgetown, Cramer 23 (U);
Berbice River, new Dageraad, Maas et al. 5565 (UNA).?Peru. Loreto: Balsopuerto, Killip & Smith
28682 (NY, US). Madre de Dios: 30 km SSW Puerto Maldonado, Barbour 4885 (MO, UNA).?.
Surinam. Commewijne: Plantation Koonenburg, Lower Commewijne River, Cramer LBB14849 (U).
Coronie: Inverness, km 129.5, Cramer LBB14851 (COL, U). Nickerie: Prins Bernard-polder, Hek
king 940 (C, GH, K, NY, U, VEN). Para: km 24 of road to Zanderij, Cramer LBB14846 (U).
Saramacca: Zonneris Creek near Bitagron, Cramer LBB14877 (F, U). Suriname: Weg naar Zei, NW
of Paramaribo, Cramer LBB14847 (U).?Tobago. The Whein, Broadway 3366 (BM, F, MO, NY,
PH); Studley Park, Broadway 4560 (BM, F, G, GH, K, NY, U, US).?Venezuela. Apure: Sabanas
de Arichuna, Ramia & Montes 5532 (VEN). Aragua: Dist. Urdaneta, 10 km SE Camatagua, Wier
sema 1436 (NY, UNA, VEN). Delta Amacuro: near Tucupita, Trujillo 12860 (MY). Gu?rico: Dist.
Miranda, 12 km N Calabozo, Wiersema 1400 (UNA, US, VEN). Miranda: Dist. Bri?n, sabana de
Guardalagua E of Sotillo, Wiersema & Gonz?lez 2203 (F, MY, NY, UNA, VEN). Monagas: Isla de
Guara, Depto. Sotillo, Mpio Uracoa, Trujillo 12946 (MY). Sucre: Laguna de Campoma, Mpio Cari
aco, Trujillo 11917 (MY); Trujillo 14380 (MY).
Flowers of this subspecies are floating and open two successive nights (Con
ard 1905; W. Pagels, pers. comm.; Wiersema, unpubl.). First-day flowers remain
fully closed until 2-3 hours before dawn. The sepals and outer petals then open
very rapidly to provide an aperture ca. 1 cm in diameter to the center of the
flower about one hour before dawn. The stigmatic surface is covered with fluid
and with the powdery remains of papillae cells. Closure of the flower is very rapid
and completed by dawn. Second-day flowers begin opening at dusk but become
fully open only about one hour before dawn. All the stamens are now fully
exposed, their anthers dehiscing at this time. The carpellary appendages are in
curved over the stigmatic surface, which is now dry. Closure of second-day
flowers begins at dawn and is completed in about one hour, the flower being
slowly retracted under water. Flowers both nights emit a strong aromatic odor
that has been variously described as resembling turpentine, xylol, petrol, ben
zene, xylene, PDB, benzol, or acetone. In Surinam they are reportedly visited by
the scarab beetle Cyclocephala verticalis Burmeister (Cramer et al. 1975). From
observations of plants in cultivation (from Wiersema 1400), it is evident that seeds
are produced autogamously. The stigma therefore remains receptive in second
day flowers.
Nymphaea amazonum subsp. amazonum relies mainly on autogamous seed
production for dispersion. Plants also propagate by the formation of elongate
stolons or short sprouts from the rhizome throughout the growing season.
A sterile Brazilian population in southwestern Pernambuco (Wiersema et al.
2296 at IP A, UNA) appears, on the basis of morphological and chemical evi
dence, to represent a natural hybrid between N. amazonum subsp. amazonum
and N. lasiophylla. Other plants collected from the nearby state of Minas Gerais
(Mos?n 803, Regnell 2036, both at S) may also represent hybrids involving N.
amazonum subsp. amazonum and TV. lasiophylla.
Several collections from central Colombia display a ring of pubescence at the
peduncular apex in addition to that generally found at the petiolar apex. In the
absence of other detectable morphological difference, no special taxonomic status
is being assigned to these plants. They have previously been regarded as a distinct
species, N. goudotiana, or as a variety or form of N. amazonum. The variety
forma-submersa reported by Sagot (1881) from French Guiana is actually N.
rudgeana.
Lehmann (1853) based his N. candolleana on N. ampia sensu de Candolle,
but explicitly excluded the basionym Castalia ampia Salisbury from his taxon. In a
footnote he mentioned a specimen of Splitgerber from Surinam as being the most
representative of his species. Presumably the lectotype at B is this specimen,
which is clearly N. amazonum subsp. amazonum.
10b. Nymphaea amazonum subsp. pedersenii Wiersema, subsp. nov.?Type:

Argentina. Corrientes: Depto. Itati, lagoon along Ruta 12, 5 km W of
junction with access road to Itati, 15 Apr 1982, Wiersema, Vanni & Schi-'
nini2270 (holotype: UNA!; isotype: CTES!).
Subspecies haec ab subsp. amazonum differt foliis 1.25-1.5-plo longioribus

quam latioribus (vice 1-1.35-plo in subsp. amazonum), supra trichosclereidibus
paucis manifestis et sclereidibus acicularibus saepe praesentibus in mesophyllo,
staminibus numerosioribus (plerumque plus quam 225), et floribus non autogamis.
Leaf blades to 31 cm long, to 22 cm wide, 1.25-1.5 times as long as wide,
broadly elliptic, peltate 1-4 mm from base of sinus, the sinus 0.34-0.47 times
length of blade; upper surface green, usually variegated with red, when dried
sparsely covered with tiny prominences from underlying trichosclereids, also often
with acicular sclereids; undersurface green or somewhat reddened, often splotched
with deeper red; venation prominent centrally, the principal veins 13-19. Petioles
to 5 mm in diameter, the apex usually with ring of pubescence, the remainder
glabrous. Peduncles to 9 mm in diameter, glabrous. Sepals 5.5-8.5 cm long, 2-4 cm
wide, subcoriaceous, rounded at apex, inconspicuously nerved. Petals 16,20, or 24,
the outer distinctly shorter and somewhat narrower than sepals, the inner 3.8-5.9
cm long, 1.3-2.1 cm wide, with acute apex. Stamens 224-324, the outermost to 5.2
cm long, with filaments to 18 mm wide and connectives prolonged 0.3-1.2 mm
beyond anthers, the inner with filaments to 1.3 mm wide and connectives pro
longed to 0.4 mm beyond anthers, the inner anthers 0.65-0.85 times as long as
stamen; pollen diameter 35-45 \im. Carpels 28-47, the appendages 10-20 mm
long, the cl?vate tips 3-6 mm long, 1.8-3 mm in diameter; axial process minute,
widely obtuse. Fruits and seeds not collected in nature. Chromosome number: 2n
= 18. Figs. 12, 30g, h.
Phenology. Collected in flower throughout the year, although only during
warmer months in the subtropical part of the range; not yet collected in fruit.
Distribution. (Fig. 31). Subtropical belt of Argentina and southern Brazil;
probably also in adjacent Uruguay and Paraguay.
Specimens Examined. Argentina. Corrientes: Depto Mburucuy?, Estancia Santa Teresa, Peder
sen 4494 (C, CORD, K, NY, P, S), Schinini & Crist?bal 10791 (CTES); Depto Mercedes, Rio
Mirinay, Krapovickas & Crist?bal 21733 (CTES, G, MO, WIS), Colonia Pellegrini, Quarin & Schinini
1063 (CTES); Depto Monte Caseros, Monte Caseros, arroyo Timboa, Nicora 7132 (CTES); Ruta 127
and arroyo Curuz? Cuatia, Schinini et al. 17395 (CTES); Depto San Mart?n, Colonia Pellegrini,
Laguna Ibera, Krapovickas et al. 29505 (CTES).?Brazil. Mato Grosso: Pocon?, Macedo &
Assumps?o 688 (NY); Transpantaneira Highway, Fazenda Jofre, Prance et al. 26205 (NY), da Silva
395 (UFMT); Rodavia Transpantaneira km 20, da Silva 440, 441 (UFMT), km 22, da Silva 459
(UFMT); Mun. C?ceres, Ilha Taiama, da Silva 212 (UFMT).
Flowers of this subspecies observed in cultivation (from Wiersema et al. 2270)

are similar in their responses to those of the nominate subspecies with one impor
tant exception: they do not produce seeds autogamously. The stigma does not
appear to retain its receptivity to pollen on the second day of flowering. Prance
(1980) reports members of this subspecies in southern Brazil to be pollinated by
the scarab beetle Cyclocephala mollis Endr?di.
In N. amazonum subsp. pedersenii seeds have not been observed on herbar
ium specimens, in Argentine field populations, or on cultivated plants in the

absence of cross-pollination, which suggests that this subspecies is an obligate
outcrosser. This apparent absence of seed production in field populations may be
due to a shortage or absence of pollinators in many of these southern populations.
Although some limited amount of seed production in field populations may have
escaped detection, the primary means of local dispersal in Argentina appears to
involve the production of stolons.
Several asexually reproducing forms which exhibit morphologies intermediate
to N. prolif?ra have been observed in northern Argentina. They are further
discussed under N. prolif?ra (no. 11).
Certain populations in Apure, Venezuela (Wiersema & Gonz?lez 2221 A, 2225
at UNA) exhibit the elliptic leaves, with extensive red pigmentation and fewer
trichosclereids, and flowers with more numerous stamens that are typical of
subsp. pedersenii in Argentina. The lack of foliar acicular sclereids in the Vene
zuelan plants differs from typical subsp. pedersenii. Seed production was evident
in these plants, although whether autogamy was involved is uncertain. An analysis
of flavonoid compounds in these plants suggests they are more closely related to
subsp. pedersenii than to subsp. amazonum and perhaps ought to be included
here.
Nymphaea amazonum subsp. pedersenii is named for Troels Myndel Peder
sen, whose field assistance and excellent specimens of Argentine Nymphaea have
contributed greatly to my understanding of this subspecies.
11. Nymphaea prolif?ra Wiersema, Brittonia 36: 219. 1984.?Type: Argentina.

Corrientes: Depto Mburucuy?, along small dam N from residence at
Estancia Santa Teresa, 8 Apr 1982, Wiersema & Pedersen 2248 (holotype:
NY!; isotypes: CTES! UNA!).
Rhizome ovoid to cylindric, stolon-forming only when germinating. Leaf

blades to 22 (-30) cm long, to 21 (-29) cm wide, 1-1.25 times as long as wide,
subcoriaceous, elliptic-ovate to suborbicular, entire, rounded to somewhat ob
tuse-angled at apex, peltate 1-10 mm from base of sinus, the sinus 0.34-0.45
times length of blade, the lobes slightly tapering to broadly rounded; upper sur
face green, often flecked with purple, when dried densely covered with tiny
face green or purple-tinged with dark flecks; venation radiate and slightly raised
centrally, becoming reticulate toward margins, the principal veins 13-19. Petioles
to 6 mm in diameter, glabrous, with 8 major and 12 minor air canals. Peduncles to
7.5 mm in diameter, glabrous, usually with 6 or 18 major and 12 or 24 minor air
canals. Flowers floating, the sepals and outer petals in distinct whorls of four.
Sepals 2.5-7.5 cm long, 1.3-3.2 cm wide, green, sometimes suffused with purple,
usually with short black streaks, ovate, rounded at apex, lacking acicular scle
reids, the nerves only faintly evident. Petals 19-35, creamy white, lacking acicular
sclereids, the outer mostly equalling or longer and somewhat narrower to broader
than sepals, with apex rounded, the inner 2.8-5.8 cm long, 0.6-2.9 cm wide,
acuminate to rounded; transition to stamens gradual. Stamens 99-248, outer
cream-colored becoming indigo- or purple-tinged within, sparsely containing fine
acicular sclereids, the outermost to 5.5 cm long, with filaments to 14 mm wide and
connectives prolonged 0.5-2 mm or more beyond anthers, the inner with fila
ments to 1 mm wide and connectives prolonged to 0.6 mm beyond anthers, the
inner anthers 0.6-0.85 times as long as stamen; pollen diameter 30-40 |im. Car
pels 20-39, the appendages 6-15 mm long, indigo or purple, with cl?vate tips
1.5-4 mm long, 0.6-1.9 mm in diameter, the stigmatic rays lacking acicular
sclereids and with papillae cells separating into powdery mass; axial process ca. 1
mm long, acute to obtuse. Fruits and seeds rarely if ever produced. Seeds from
artificial hybridization finely granulate, with longitudinal rows of hairs. Tuberifer
ous flowers usually present, submersed; outer sepaloid bracts (3-) 4, 3-6 cm long,
1.8-2.3 cm wide, greenish to reddish, usually with short black streaks, subcoria
ceous, ovate-triangular with truncate base; tuber globose, densely wooly, produc
ing alternating whorls of tuberous flowers and leaves in (3's) 4's; peduncle of
tuberiferous flowers to 7.5 mm in diameter, glabrous, abcission layer forming
near apex. Chromosome number: 2n = 18. Figs. 10, 12, 32.
Phenology. Collected in flower from November through June, although
flowering appears to be of shorter duration further south; no collections of fruits
from natural populations are known.
Distribution. (Fig. 28). Collected only in northern Argentina, Paraguay,
southern Brazil, western Ecuador, and in Costa Rica and El Salvador; primarily
in lowland savannahs.
Representative Specimens. Costa Rica. Guanacaste: Comelco, 5 km NW Bagaces, Opler 926

(F); Palo Verde, Bagaces, Janzen T-68 (MO).?El Salvador. Santa Ana: N side of Lago de
Metap?n, Fassen 28738 (WIS); Lagunita San Dieguito, 5 km SW of Metap?n, Fassen 29381 (WIS).
Argentina. Corrientes: Depto Capital, 15 km E Corrientes, Ruta 12, Ahumada 1521 (CTES),
along Ruta 12 near airport NE of Corrientes, Wiersema et al. 2241 (UNA); Depto Concepci?n, Paso
Puc?, Pedersen 8094 (C, K, NY, P); Depto Itat?, Ramada Paso, Schinini 8766 (CTES), along Ruta 12,
37 km E of access road to Itati, Wiersema et al. 2273 (CTES, UNA); Depto Mburucuy?, Estancia
Santa Teresa, Ca?ada Portillo, Pedersen 7743 (C, CORD, K, NY, S); Depto San Luis, near Ramones,
Pedersen 8335 (C, K, NY, P). Formosa: Depto Pilcomayo, S of Clorinda on Ruta 11, Krapovickas &
Crist?bal 33903 (CTES), Louia Hora, Morel 1810 (LIL).?Brazil. Mato Grosso: Mun. Pocon?,
Rodavia Transpantaneira km 13, da Silva 219 (UFMT). Rio Grande do Sul: Mun. Uruguayana,
Palacios-Cuezzo 223 (LIL).?Ecuador. El Oro: between Santa Rosa and M?chala, 6 Jun 1959,
Guldmann s.n. (K). Guayas: Nobol, Asplund 16024 (K, NY); along hwy 25 between Babahoyo and
Guayaquil, Wiersema et al. 2028 (AAU, LIL, MO, UNA, USF).?Paraguay. Boquer?n: Santa Elisa,
Hassler 2804 (BM, G-some sheets mixed, GH, K-mixed collection, LIL, MO, NY, S). La Cordillera:
San Bernardino, Sparre & Vervoorst 2336 (LIL, W). Neembuc?: Dist. Yataity, Alonso, Walter 136
(BM). Paraguar?: Laguna Cora, on the plain of Mbatobi, near Paraguari, Balansa 522 (G, P). Depto
unknown: regione circus superioris fluminis Apa, Hassler 8471 (BM, G, NY).
Observations of plants obtained from Argentina (Wiersema et al. 2273) in

cultivation indicate that flowers open for two successive nights. First-day flowers
begin opening about one hour after dusk. The sepals and outer petals become
broadly reflexed, with the inner petals and stamens spreading to provide a small
opening to the fluid-filled stigma within an hour of initial opening. Flowers emit a
somewhat unpleasant, etherlike odor. Closure of first-day flowers does not occur
until 3-4 a.m. Second-day flowers open at dusk. The sepals and outer petals again
open wide, but the now-dehising stamens and inner petals do not become reflexed
until 1-2 hours after midnight. The stamens are fully exposed from about 2-4
a.m., with the subsequent closing of the flower being completed by 1-2 hours
FIG. 32. Nymphaea prolif?ra (Wiersema & Pedersen 2248). a. Undersurface of leaf. b. Upper
surface of young leaf. c. Opened flower, d. Cross section of petiole, e. Cross section of peduncle, f.
Upper surface of leaf. g. Tuberiferous flower, h. Cross section of peduncle of tuberiferous flower.
before dawn. Autogamous seed production has not been observed in N. prolif?ra,
which suggests that the stigma is no longer receptive in second-day flowers.
Although seeds have been produced from an artificial hybridization in cultiva
tion, no evidence of seed production has been observed in field populations. This
is apparently due to an absence of available pollinators in many of the regions N.
prolif?ra now inhabits. In northern Argentina, the subtemperate nature of the
climate may exclude potential pollinators of N. prolif?ra from tropical regions to
the north. A similar phenomenon may be affecting N. amazonum. Existing polli

nators of N. prolif?ra could likely be observed in northern Paraguay or southern
Brazil, as they do not appear to have migrated to Ecuadorian or Central Ameri
can populations. In most populations, N. prolif?ra seems to rely exclusively on its
unusual form of asexual reproduction for its dispersion. This process, which in
Nymphaea commonly occurs elsewhere only in N. lasiophylla (see that species,
no. 12), involves the production of tuber-bearing flowers. It is an effective
method of propagation in N. prolif?ra, as hundreds of tiny tubers, many already
complete with juvenile leaves and roots, can be released from a single individual.
Nymphaea prolifera's closest taxonomic relative appears to be N. amazonum,
based on phenetic comparisons and phytochemistry. Indeed, potential hybrids are
suspected between these two species in Argentina, where N. amazonum subsp.
pedersenii and N. prolif?ra both occur. Several asexual forms exist in this region,
which propagate entirely via the production of tuberiferous flowers. Chromosome
counts obtained by Quarin (pers. comm.) and by me from three such forms
suggest that they are triploid (3n = 27). Several morphological characteristics
tending toward N. amazonum are evident in these forms, such as pubescence at
the apex of the petioles, sepals that are distinctly longer and broader than the
petals, and carpellary appendages with more thickened cl?vate tips. The charac
teristics just mentioned for these asexual forms serve to distinguish them from N.
prolif?ra, which lacks pubescence on the petioles, has sepals that are usually
shorter and narrower than the outer petals, and has carpellary appendages with
tips mostly less than 1.5 mm thick. In N. amazonum, which never produces
tuberiferous flowers, a ring of pubescence is evident at the apex of the petiole, the
sepals are always longer and often broader than the petals, and the carpellary
appendages have tips mostly more than 2 mm thick. Although certain of these
putative hybrids also commonly produce normal flowers, fruits are never devel
oped. Most of these plants, however, are characterized by a reduced production
of normal flowers.
12. Nymphaea lasiophylla Martius & Zuccarini, Abh. Math.-Phys. Cl. K?nigl.
Bayer. Akad. Wiss. 1: 364. 1832, quoad folium, flos est N. tenerinervia
Caspary. Leuconymphaea lasiophylla (Martius & Zuccarini) Kuntze, Re
vis, gen. pi. 1: 11. 1891.?Type: Brazil. Bahia: "in aquis stagnantibus ad
flumen St. Francisci prope Juazeiro," April-May (probably 1819), Mar
tius exsicc. 2377 (lectotype, the leaves only designated: M!, on 2 sheets;
isolectotype: B!-fragment).
Rhizome subglobose to cylindric, stolon-forming throughout vegetative pe

riod. Leaf blades to 33 cm long, to 29 cm wide, 1-1.2 times as long as wide,
subcoriaceous, broadly elliptic to suborbicular, entire, rounded to somewhat ob
tuse at apex, peltate 0.5-11 mm from base of sinus, the sinus 0.37-0.48 times
length of blade, the lobes rounded or with small obtuse protuberance; upper
surface green, often with darker flecks, when dried densely covered with tiny
face green with slight reddish tint; venation radiate, the veins slightly raised
centrally, faintly reticulate toward margins, the principal veins 13-23. Petioles to
7 mm in diameter, glabrous, with 8-12 major air canals. Peduncles to 9.5 mm in
diameter, glabrous, with 6 or 18 major air canals. Flowers floating, the sepals and
outer petals in distinct whorls of four. Sepals 3.5-7 cm long, 1.3-3.3 cm wide,
uniformly green or yellow-green, ovate, acute or somewhat rounded at apex,
lacking acicular sclereids, inconspicuously nerved. Petals 20-26, creamy white
with lace work venation, lacking acicular sclereids, the outer only slightly shorter
than sepals, with apex rounded to somewhat acute, the inner 3.1-5.9 cm long; 1
2.5 cm wide, acute or somewhat rounded; transition to stamens abrupt. Stamens
65-130, creamy white, containing fine acicular sclereids, the outermost to 4.7 cm
long, with filaments to 9 mm wide and connectives prolonged 0.5-1.5 mm beyond
anthers, the inner with filaments to 1.8 mm wide and connectives prolonged 0.2
mm or less beyond anthers, the inner anthers 0.4-0.6 (-0.7) times as long as
stamen; pollen diameter 35-40 ^m. Carpels 19-33, the appendages 5-14 mm
long, to 3 mm wide near base, with reddish purple tips, lingulate-tapering, the
stigmatic rays lacking acicular sclereids and with papillae cells remaining attached;
axial process minute or absent. Fruits seldom develop; seeds 0.9-1.05 mm long,
0.6-0.7 mm in diameter, 1.4-1.6 times as long as broad, with longitudinal rows
of hairs 55-110 \xm long, the surface finely ruminate. Tuberiferous flowers usually
present, submersed; outer sepaloid bracts 3-4, uniformly yellow-green or green;
inner bracts 0-4; tuber globose, densely wooly, usually producing alternating
whorls of tuberous flowers and leaves in 3's; peduncle of tuberiferous flower to 7
mm in diameter, glabrous, abscission layer forming near apex. Chromosome num
ber: 2n = 18. Figs. 3, 10, 12, 33.
Phenology. Collected in flower throughout the year; fruiting collections are
rare, in June and August.
Distribution. (Fig. 34). Known only from coastal states of eastern Brazil, Rio
de Janeiro north to Piaui, with a disjunct population, probably introduced, on Isla
de Margarita, Venezuela; usually in stagnant water, often in artificial ponds.
Specimens Examined. Brazil. Alagoas: between Palmeira dos Indios and Bom Conselho, 12 km
S of division Alagoas-Pernambuco, Wiersema et al. 2293 (IPA, MO, UNA). Bahia: E of Rio Salitre,
between Juazeiro and Sobradinho, Wiersema et al. 2298 (IPA, UNA). Cear?: Mun. Fortaleza, Lagoa
Mecejana, Mecejana, Drouet 2140 (F, GH, S, US), Lagoa do Tanape, Drouet 2292 (F, GH, US),
Drouet2563 (GH), ponds of Damas, Drouet 2486 (F, GH, S, US); Tat?, L?fgren 729 (S); km 195 of
BR-222 between Sobral and Fortaleza, Wiersema et al. 2313 (INPA, IPA, UNA, US); km 14 of BR
304 between Fortaleza and Mossor?, Wiersema et al. 2316 (IPA, MO, SP, UNA, US). Paraiba: a?ude
Novo, Campina Grande, Wright 1 (GH); a?ude Puxanana, Wright 4 (F, GH). Pernambuco: in a river,
Tapera, Pickel 3725 (F, GH, IPA, NY); km 327-8 of BR-232 between Arcoverde and serre Talhada,
Wiersema et al. 2294 (IPA, UNA); km 475 of BR-232 E of Salqueiro, Wiersema et al. 2295 (IPA, NY,
?B, UNA); Mun. Rajada, km 45 of PE-407 between Petrolina and Paulistana, Wiersema etal. 2302.
Rio de Janeiro: Mun. Campos, BR-356, 1 km de Campos, Casari et al. 523 (GUA). Rio Grande do
Norte: Mossor?, Wiersema etal. 2317 (F, GUA, IPA, MO, UNA).?Venezuela. Nueva Esparta: Isla
de Margarita, between Boca del R?o and Boca del Pozo, Steyermark et al. 130978 (MO, UNA).
The floral behavior has not been studied; however, flowers open fully when
placed in darkness and emit a strong aromatic odor.
In northeastern Brazil reproduction in N. lasiophylla is primarily asexual.
Propagation in most populations is by means of abortive tuberiferous flowers, a
phenomenon among Nymphaea common only in this species and N. prolif?ra,
although rare occurrences of similar structures have been reported for other syn
carpous Nymphaea (Mohan Ram & Nayyar 1974; Mitra & Subramanyam 1982).
FIG. 33. Nymphaea lasiophylla (leaf from Wiersema et al. 2317; flower from Wiersema et al.
2316). a. Undersurface of leaf. b. Opened flower, c. Cross section of petiole, d. Cross section of
peduncle, e. Upper surface of leaf.
These abortive flowers, their status indicated by the peduncular air canal anatomy
and the production of outer sepaloid bracts, develop an enlarged tuber, which
produces whorls of leaves and additional secondary tuberiferous flowers. A few
orders of branching may result from similar development of secondary and even
tertiary tuberiferous flowers, resulting in the production of a substantial number
FIG. 34. Distribution of Nymphaea lasiophylla and N. lingulata.
of tubers. The tubers detach readily, remain afloat briefly, and eventually develop
into adult plants. Leaves and roots are often well-developed prior to abscission of
these tubers, which suggests the application of the term vivipary to this reproduc
tive strategy. The ratio of tuberous to normal flowers varies from one population
to another. Although tuberous flowers were present in all field populations
studied, normal flowers were not observed in some populations. Another form of
vegetative propagation, the production of stolons, has been observed in certain
populations as well.
Seed production was evident in only one of ca. 20 field populations examined.
Coincidentally, in only this and another nearby population (ca. 1 km away) was
there evidence of pollinating beetles, these found trapped within some of the
diurnally closed flowers. This suggests that cross-pollination may be necessary for
seed production, the absence of seeds from most populations being attributable to
an absence of pollinators. Further study is necessary to verify this hypothesis.
Nymphaea lasiophylla is clearly linked to N. lingulata; the two species are
unique among members of subg. Hydrocallis in their possession of tapering (as
opposed to cl?vate) carpellary appendages. Karyology, phytochemistry, and seed
morphology all support this close relationship. Although not apparent phyto
chemically, N. lasiophylla must also bear some relation to N. prolif?ra, as both
species exhibit a form of vegetative reproduction not observed elsewhere in the
subgenus. A putative hybrid involving N. lasiophylla and N. amazonum subsp.
amazonum is discussed under this subspecies (no. 10a).
The type description, drawn from Martius's field notes and the type speci
mens, is based on a mixed collection (gathered in 1819 fide Caspary, 1878). The
name N. lasiophylla, although descriptively inappropriate (the pubescence alluded

to is actually a layer of algae), has been retained for the two leaves of this
collection, which are now mounted on separate sheets. Flowers are also mounted
on both sheets, but no connecting rhizomes are present. These flowers are assign
able to N. tenerinervia, which Caspary (1878) subsequently described from a
second Martius collection from Juazeiro. It should be noted that I am hesitant in
my assignment of the N. lasiophylla type material. Though the flowers are virtu
ally identical to those of N. tenerinervia, the leaves, although otherwise compara
ble, are slightly more peltate than those in other populations of the species to
which they have been linked. This species is now flourishing in several localities
near Juazeiro and would likely have been encountered by Martius. In most of
these localities it produces only submerged tuberous flowers, but these could
easily have been overlooked by Martius, who collected only floating portions of
the plant. However, it is still possible, albeit improbable, that Martius's N. lasio
phylla type represents an authentic combination of characters in a plant (perhaps
even of hybrid origin) that has not been encountered since.
The specimen at B from the herbarium Lehmann is labelled N. lasiophylla but
contains no collection information. It contains one flower of N. tenerinervia and a
leaf fragment of N. lasiophylla. This leaf fragment matches that portion missing
from a leaf of the Martius type at M; therefore the components of this specimen
are believed to have once been a part of the original Martius collection.
13. Nymphaea lingulata Wiersema, Brittonia 36: 215. 1984.?Type: Brazil.

Maranh?o: pond along BR-316 between Teresina and Caxias at km 598,
ca. 23 km W of Rio Parnaiba, 28 Jun 1982, Wiersema, Horn & de Ataide
Silva 2308 (holotype: IPA!; isotypes: NY! UNA!).
Rhizome ovoid to subglobose, sprouting freely, often stolon-forming through

out vegetative period. Leaf blades to 25 cm long, to 21 cm wide, 1-1.3 times as
long as wide, subcoriaceous, broadly ovate to suborbicular, entire, rounded at
apex, peltate 1-6 mm from base of sinus, the sinus 0.37-0.45 times length of
blade, the lobes rounded or with small obtuse protuberance; upper surface green,
when dried ornate with starshaped expressions of underlying trichosclereids, also
containing acicular sclereids; undersurface red to reddish purple; venation radiate
but scarcely evident centrally, somewhat impressed, the principal veins 15-21.
Petioles to 6 mm in diameter, glabrous, with 2, 4, or 6 major air canals. Peduncles
to 7 mm in diameter, glabrous, with an inner ring of 5-6 major air canals sur
rounded by 10-12 minor ones. Flowers floating, the sepals and outer petals in
distinct whorls of four. Sepals 5-8 cm long, 1.5-3.1 cm wide, yellowish green,
sometimes flecked with black, subcoriaceous, ovate, acute to rounded at apex,
containing acicular sclereids, inconspicuously nerved. Petals 8-14, creamy white,
chartaceous, with lace work venation, acicular sclereids present but often incon
spicuous, the outer petals shorter than sepals, with apex rounded or somewhat
acute, the inner 4-7 cm long, 1.3-2.7 cm wide, acute to acuminate; transition to
stamens abrupt. Stamens 80-130, creamy white with dark purple filaments, con
taining acicular sclereids, the outermost to 5.6 cm long, with filaments to 9 mm
wide and connectives prolonged 2 mm or more beyond anthers, the inner with
filaments to 1.8 mm wide and connectives prolonged 0.5-1 mm or more beyond
FIG. 35. Nymphaea lingulata (Wiersema et al. 2308). a. Undersurface of leaf. b. Opened flower,
c. Unopened flower bud. d. Cross section of petiole, e. Cross section of peduncle, f. Upper surface of
leaf.
pm. Carpels 20-28, the appendages 5-14 mm long, to 2 mm wide at base, creamy
white with purplish base abaxially, lingulate-tapering, the stigmatic rays lacking
acicular sclereids and with papillae cells remaining attached; axial process ca. 1
mm long and blunt. Fruits known only from type collection; seeds 1.05-1.2 mm
long, 0.8-0.9 mm in diameter, ca. 1.3 times as long as broad, with longitudinal
rows of short hairs 30-70 \xm long, the surface finely ruminate. Tuberiferous
flowers absent. Chromosome number: 2n = 18. Figs. 3, 10, 12, 13, 35.
Phenology. Flowering material scanty, collected in February, April, June, and
August; fruits known only from my own June collection.
Distribution. (Fig. 34). Collected only in northeastern Brazil where observa
tions indicate it is probably quite rare. Collections from Paraguay which were
previously assigned to this species (Wiersema 1984a) are now believed to repre
sent N. gardneriana.
Specimens Examined: Brazil. Cear?: Mun. Maranguape, a?ude S?o Bento, Drouet 2174 (F,
GH). Minas Gerais: Mun. Pirapora, Alagoas, Barreto 11319 (BHMH, UNA); Mun. Ituiutaba, L.
Teyiuba, Macedo 2279 (MO, S); Lagoa da Babylonia, Glaziou 12414 (C, G, K, P).
Cultivation of Brazilian material (Wiersema et al. 2308) indicates that flowers

open for two successive nights. On both nights flowers begin opening about 1/4-2
hours after dark, close at ca. 2-3 a.m., and are only faintly odorous. The thick,
coriaceous sepals prevent the petals from reflexing more than ca. 30 degrees on
either night; however, both the stamens and fluid-filled stigma are fully exposed.
In all flowers observed in cultivation stamen dehiscence occurred on the first
night, a phenomenon not observed in the other species of subg. Hydrocallis
studied. Abundant seeds were produced autogamously from each flower.
Nymphaea lingulata propagates mainly via seed production, the majority of
this production is probably autogamous. Asexual reproduction, through stolon
formation, is also found in this species.
Nymphaea lingulata is obviously related to N. lasiophylla; both species exhibit
tapered instead of cl?vate carpellary appendages. Interestingly, N. lingulata dis
plays some similarity in flavonoid chemistry to members of subgenera Nymphaea
and Brachy ceras, which also have tapered carpellary appendages.
14. Nymphaea rudgeana G. Meyer, Prim. fl. esseq. 198. 1818. Nymphaea ampia ?
rudgeana (G. Meyer) de Candolle, Syst. nat. 2: 54. 1821. Castalia rudge
ana (G. Meyer) Britton & P. Wilson, Bot. Porto Rico 5(2): 305. 1924.?
Type: Guyana. "In rivulis continentis," Rodschied 281 (lectotype, desig
nated by Caspary, 1878: GOET!). The type sheet contains one leaf and
one flower of N. rudgeana and one flower of N. glandulifera, the latter
undoubtedly from Rodschied 155, which has apparently been mismounted
from the original specimen of the Grisebach Herbarium.
Nymphaea sinuata Salzmann ex Lehmann, Hamburger Garten- Blumenzei
tung 9: 203. 1853.?Type: Brazil. Para: Lago d'Obidos, Spruce 479 (lec
totype, here designated: K!).
Nymphaea amazonum var. forma-submersa Sagot, Ann. Sei. Nat. Bot., ser.
6, 2: 142. 1881.?Type: French Guiana. Karouany [Acarouani], 1855,
Sagot 25 (lectotype, here designated: K!).
Rhizome to 7 cm or more long, to 8 cm or more in diameter, ovoid to

subglobose, lacking stolons. Leaf blades to 36 (-45) cm long and wide, 0.8-1.2
times as long as wide, subcoriaceous, suborbicular, irregularly dentate with obtuse
teeth, truncate to rounded at apex, peltate 0-12 mm from base of sinus, the sinus
0.35-0.51 times length of blade, the lobes with dentate apex, the margins often
slightly upturned in larger leaves; upper surface green, sometimes purplish, occa
sionally with purple flecks, when dried densely covered with tiny prominences
from underlying trichosclereids, lacking acicular sclereids; undersurface greenish
or brownish purple, sometimes with darker flecks; venation prominent and radi
ate centrally, becoming conspicuously reticulate toward margins, the principal
veins (11-) 15-23. Petioles to 9 (-11) mm in diameter, glabrous, with 2 major

and 4-numerous minor air canals. Peduncles to 12 mm in diameter, glabrous,
usually with a ring of 6 (-7) major air canals surrounded by 12 minor ones.
Flowers floating or, more commonly, emergent, the sepals, petals, and outer
stamens in distinct whorls of four. Sepals 3.5-8 cm long, 1.1-3.2 cm wide, green
ish above, somewhat yellowish below, subcoriaceous, ovate, rounded or some
what acute at apex, containing fine acicular sclereids, inconspicuously nerved.
Petals 12-29, creamy white to light yellow, with pinkish tinge particularly on
older flowers, subcoriaceous, with acicular sclereids present but inconspicuous,
the outer slightly longer or mostly shorter and slightly narrower or often wider
than sepals, with apex rounded to somewhat acute, the inner 2.2-5.6 cm long,
0.7-1.7 (-2.2) cm wide, acute; transition to stamens gradual. Stamens 39-186,
creamy white to light yellow, containing fine acicular sclereids, the outermost to 5
cm long, with filaments to 16 mm wide and connectives prolonged to 1 mm
beyond anthers, the inner with filaments to 1.5 mm wide and connectives pro
longed 0.2 mm or less beyond anthers, the inner anthers 0.31-0.6 times as long as
stamen; pollen diameter 35-45 \im. Carpels 11-31, the appendages 6-16 mm
long, creamy white to light yellow, with strongly cl?vate tips inwardly coiled, 2-7
mm long, 0.8-1.8 mm in diameter, the stigmatic rays lacking acicular sclereids
and with papillae cells separating into powdery mass; axial process to 2 mm long,
obtuse. Fruits normally present; mature seeds 1.2-2.4 mm long, 0.85-1.8 mm in
diameter, 1.2-1.6 times as long as broad, with longitudinal rows of hairs 100-350
|xm long, the surface topography smooth. Tuberiferous flowers absent. Chromo
some number: 2n = 42? Figs. 1, 3, 10, 36.
Distribution. (Fig. 37). Eastern and northern South America north of the
state of Paran?, Brazil, and east of northern Colombia; also in Jamaica, Cuba,
Guadeloupe, Martinique, Trinidad, and Nicaragua; mostly in lowland coastal
habitats, some of these somewhat brackish, in stagnant or flowing water.
Representative Specimens. Cuba. Santa Clara: city of Santa Clara, Britton etal. 6129 (NY, US);
near Lomo Cruz, Britton et al. 10238 (NY).?Guadeloupe. Locality unknown, Duchassaing 16
(GOET).?Jamaica. St. Elizabeth: Black River, Adams 13218 (UCWI).?Martinique. Le Robert,
NE de Trou Terre, Raynal-Roques & J?r?mie 21178 (P).
Nicaragua. San Juan del Norte (Greytown), Stevens 20848 (MO).
Brazil. Amap?: Fazenda Mungubas, Black & Lobato 50-9534 (IAN); lago grande, Goeldi 1204
(BM, INPA, U). Amazonas: Taperinha bei Santar?m, Ginzberger & Hagmann 658 (F, W); Mun.
Humayta, near Livramento, Krukoff6969 (BM, BR, F, GH, K, LIL, MO, NY, U, US). Bahia: 5 km S
of Santa Cruz, Harley et al. 17129 (MO, NY, P, U); 10 km NE of Salvador, Anderson 223 (INPA); Rio
Salitre, between Juazeiro and Sobradinho, Wiersema etal. 2299 (IPA, UNA). Cear?: Mun. Fortaleza,
Barra do Cear?, Drouet 2512 (F, GH, S, US). Maranh?o: S?o Luis, Anderson 161 (NY, INPA); km 583
of BR-316, 38 km W Rio Parnaiba, Wiersema et al. 2309 (IPA, UNA). Para: lina do Mosquiero, near
Para, Killip & Smith 30531 (NY, US); Mun. Bel?m, Lagoa Agua Pr?ta (Utinga), Drouet 1955 (F, GH,
US); Rio Arapiuns, Lago Mentai, Pires & Silva 4359 (IAN, US). Paran?: Mun. Paranagua, Pontal do
Sul, Hatschbach 16387 (US). Pernambuco: Tapera, Pickel 2343 (F, GH, US); Dois Irmois near Recife,
Wiersema et al. 2285 (IPA, ?B, UNA); km 45 of PE-60 S of Cabo, Wiersema et al. 2286 (IPA, NY,
UNA). Rio de Janeiro: Mun. Campos, Farol de S?o Tom?, Araujo & Maciel 4056 (GUA); Mun. de
Casimiro de Abreu, Lagoa Iodada, Casari 500 (GUA). S?o Paulo: Santos, Mos?n 3339 (BR, C, GH,
GOET, K, M, S, U).?Colombia. Magdalena: Santa Marta, Smith 2108 (BM, BR, G, GH, K, NY, P,
PH, US).?French Guiana. Vicinity of Cayenne, Broadway 276 (GH, NY, US); Cayenne, Badocul,
Broadway 919 (GH, NY, US); Savane Gabrielle, 10 km ESE de Stoupan, Raynal-Roques & J?r?mie
FIG. 36. Nymphaea rudgeana (Wiersema & Gonz?lez 2205). a. Undersurface of leaf. b. Opened
flower, c. Unopened flower bud. d. Cross section of petiole, e. Cross section of peduncle, f. Upper
surface of leaf.
21291 (P); Kaw, 60 km SE de Cayenne, Raynal-Roques 21554 (P).?Guyana. Demerara, 1824, Parker
s.n. (K, NY); Barina River, Jenman 7230 (K); Kara-Kara Creek, Wismar, Demarara River, Jenman
7779 (K); Tapacooma Lake, Jenman 7892 (K); Kaieteur Plateau, Cowan & Soderstrom 2224 (K, NY,
US); Amacuro River, NW District, de la Cruz 3563 (GH, US); Wanama River, NW District, de la Cruz
3969 (F, GH, MO, NY, PH, US); Rapununi, Moreru Lake, Goodland 1069 (US); Lamaha stop-off, SE
of Georgetown, Hitchcock 16868 (NY, S, US).?Surinam. Brokopondo: Kwakoegron, Saramacca
River, Maguire 25004 (F, GH, K, NY, US). Commewijne: Voorburg, Lower Suriname River, Cramer
LBB14861 (U). Coronie: Sarah-Leasowes, Cramer LBB14865 (MO, U). Marowijne: Maengo, Cramer
FIG. 37. Distribution of Nymphaea rudgeana.
LBB14913 (F, U). Nickerie: Nieuw-Nickerie, Hekking 955 (GH, K, NY, U). Para: Coropina Creek,
Cramer LBB14859 (COL, NY, U). Suriname: 2 km E of Meerzorg, Cramer LBB14860 (U).?Trinidad.
1 mi S St. Joseph, Hazen 2079 (NY); S of Arima, near Caroni River, Britton & Britton 2887 (GH, K,
NY, US).?Venezuela. Anzoategui: R?o Tigre, ESE of San Tom?, Pittier 1408 (VEN); km 70-71 de
carretera Sn. Mateo-Anaco, Rutkis 169 (VEN). Bol?var: near Urull?n, at foot of Auyantepui, Vareschi
& Foldats 4532 (NY, VEN); drainage of R?o Aponguao, 152 km S of El Dorado, Steyermark &
Dunsterville 104108 (NY, VEN); Depto Roscio, near Luepa, Badillo & Holmquist 6239 (MY). Delta
Amacuro: downstream from mouth of Yarakita River, Sierra Imataca, Steyermark 87446 (K, NY, SP,
VEN). Miranda: Dist. Bri?n, Rio Capaya near Sotillo, Wiersema & Gonz?lez 2205 (F, MY, NY, UNA,
VEN). Monagas: R?o Tigrito, al SSW de San Tom?, Rutkis 158 (VEN); R?o Mapirito, carretera
Matur?n-Temblador, Rutkis 159 (VEN). Zulia: Dist. Col?n, entre Boca Escalante y Pta. Pausada,
Bunting & Luis 6497 (MO).
Field studies of N. rudgeana (Cramer et al. 1975; Prance & Anderson 1976)
and observations of plants in cultivation (from Wiersema <? Gonzalez 2205) indi
cate that flowers are protogynous, opening for two or three successive nights.
Flowers reportedly open at dusk each night. The stigma is receptive, covered with
stigmatic fluid, the first night, and the stamens dehisce the second night. Flowers
both nights emit a strong "fruity" or "aniselike" odor. They are visited by the
scarab beetles Cyclocephala verticalis Burmeister and C. castanea Olivier. Al
though closure of flowers is not discussed in these studies, Conard (1905) reports
flowers to close at dawn, at 2-3 a.m., or before midnight. In view of the flowering
responses of other species of subg. Hydrocallis, it is likely that the different
closing times given correspond to flowers of different ages. The stigma clearly
remains receptive the second night, as seeds are produced autogamously.
Nymphaea rudgeana reproduces entirely via seed production, much of this
autogamous; no form of asexual reproduction has been observed.
Based on evidence from phytochemical and crossing studies, N. amazonum
subsp. pedersenii appears to be the closest taxonomic relative to N. rudgeana
within subg. Hydrocallis. The seed morphology, general morphology, and
chromosome number of N. rudgeana all suggest a relationship to subg. Lotos and
indicate that it may be of polyphyletic origin. Nymphaea rudgeana is the only
member of subg. Hydrocallis with dentate leaves.
Plants growing in flowing water often produce an abundance of submersed
leaves. These leaves are more delicate and often have entire margins. Sagot
(1881) described one such population in French Guinea as N. amazonum var.
forma-submersa.
Doubtful and Excluded Names
Nymphaea foetida Gardner ex Lehmann, Hamburger Garten- Blumenzeitung

9: 205. 1853, pro syn.
Nymphaea fragrans Gardner ex Caspary in Martius, Fl. bras. 4(2): 176. 1878,
pro syn.
Nymphaea integrifolia Salzmann ex Lehmann, Hamburger Garten- Blumen

zeitung 9: 205. 1853, pro syn.
Nymphaea leiboldiana Lehmann, Hamburger Garten- Blumenzeitung 9: 197,

209. 1853.?Type: "in terris Mexicanis," no date, Leibold s.n. (not located).
Lehmann also cited: "in Asia australi Burke?" Caspary (1878) viewed an original
specimen of the American plant (apparently the Leibold collection, although not
so indicated) at W (now destroyed) and referred it to N. ampia var. speciosa
(Martius & Zuccarini) Caspary. Another "speciminis originalis" of N. leiboldiana
in the herbarium Lehmann contained, according to Caspary, a flower of N. rudge
ana and a leaf of N. ampia. Conard (1905) referred to specimens labelled N.
leiboldiana "ex herb. Lehm." at K and B, which he identified as N. capensis. I
saw none of the above specimens; presumably the holotype was the specimen
formerly at W. Until an isotype (if one exists) can be located, the application of
this name will remain questionable. The type description agrees well with Cas
pary's identification of the holotype.
Nymphaea maximiliani Lehmann, Hamburger Garten- Blumenzeitung 9: 204,

215-216. 1853.?Type: "prope Bahiam," no date, no collector (lectotype, the
flower only here designated: B!). A specimen at B from Lehmann's herbarium
bears only the name N. maximiliani and no other information. According to
Caspary (1878), this specimen is the sole basis for Lehmann's description of N.
maximiliani. It includes one flower of N. odorata and a leaf of N. tenerinervia. It
has been lectotypified on the N. odorata element to preserve Caspary's treatment
of N. tenerinervia (see discussion under N. tenerinervia, no. 9).
Nymphaea nocturnea March ex Hooker, Bot. Mag. 80: t.4823. 1854, pro syn.
Nymphaea tropaeolifolia Lehmann, Hamburger Garten- Blumenzeitung 9:

197, 209-210. 1853. According to Caspary (1878) the specimen in Lehmann's
herbarium which bore this name (and which I was unable to locate at B) was a
mixed collection composed of a leaf of N. rudgeana and a flower of N. ampia.
Lehmann's citation in the protologue reads: "in Brasilia prope Bahiam in aquis
stagnantibus et in Surinamia." The second part of Lehmann's citation may refer
to a collection from Surinam by Hostmann (565); the leaves of this collection at K
were identified by Lehmann as his N. tropaeolifolia. This collection is actually N.
amazonum. The first part may refer to a collection of Salzmann (381), which
bears identical locality information and displays the same mixture of taxa (as that
observed by Caspary) in several European herbaria (BM, G, K, P). Indeed,
Lehmann mentions the manuscript name of Salzmann, N. sinuata, appearing on
these other specimens in his discussion and his description matches the particular
combination of species observed by Caspary. In the same manuscript Lehmann,
citing a collection of Spruce from Brazil but not Salzmann's collection, validates
Salzmann's name with a reference to a Plumier description and plate.
ACKNOWLEDGMENTS
I am exceedingly grateful to Robert R. Haynes, my doctoral advisor, for his professional guid
ance, assistance, and support without which this project could not have been completed. Other
members of my dissertation committee, Robert Krai, C. Earle Smith, Jr., Joseph F. Scheiring, and
George P. Whittle, provided editorial comments on the original manuscript, as did Edward L.
Schneider for the section on seed morphology. The constructive comments of Christiane Anderson
and Daniel J. Crawford on the submitted manuscript are appreciated as well. I am also indebted to
Susan Wittlake and Jan A. Lay for preparing some of the illustrations, Steven R. Hill for providing a
Latin description and assisting with several translations, M. Carol Argo for her efforts in typing the
original manuscript, James L. Reveal for securing funds to purchase reprints, and above all Lori
Wittlake Wiersema for her emotional support throughout. I also thank those who contributed to the
research phase of this project: Harriet E. Smith-Somerville and Deborah S. Clayton for supervising
the SEM seed and pollen research, Thomas R. Bauman for aiding in the preparation of seed sections,
Marvin L. Roberts for valuable tutelage in flavonoid chemistry, Ronald G. Lindahl for answering
several karyological inquiries, Brett C. Ratcliffe for identification of Cyclocephala beetles, and fellow
graduate students Lawrence J. Davenport, Charles N. Horn, and David L. Lentz for providing a
variety of skills and services.
Field work was made possible through the generous cooperation of Angel Gonz?lez, Antonio
Krapovickas, Carmen Crist?bal, Aurelio Schinini, Ricardo Vanni, Troels Myndel Pedersen, Marcelo
de Ataide Silva, and other staff members at VEN, CTES, and IP A, as well as Richard P. Wunderlin
at USF. Appreciation is also expressed to the curators of herbaria (B, BAB, BHMH, BM, BR, C,
COL, CORD, CTES, ENCB, F, FLAS, G, GH, GUA, IAN, INPA, K, LIL, M, MBM, MO, MY,
NY, P, PH, S, SEL, SP, U, UB, UCWI, UNA, US, VEN, W, WIS) for the loan of specimens and the
staffs of several European herbaria (B, BM, G, GOET, K, M, P) for facilitating visits to examine type
material.
Financial support for field work was received from the Latin American Studies Program, Univer
sity of Alabama, and the National Science Foundation (Grant No. DEB-8111024 to the University of
Alabama). Laboratory work was aided by a grant from the National Science Foundation (No. DEB
7818402 to the University of Alabama). This is Scientific Article No. A-4331, Contribution No. 7320
of the Maryland Agricultural Experiment Station.
LITERATURE CITED
Bate-Smith, E. C. 1963. The phenolic constituents of plants and their taxonomic significance. I.
Dicotyledons. J. Linn. Soc, Bot. 58: 95-173.
-. 1968. Chemotaxonomy of Nuphar l?tea (L.) Sm. Phytochemistry 7: 459.
Bendz, G., and B. Jonsson. 1971. Anthocyanins in leaves o? Nymphaea candida. Phytochemistry 10:
471-472.
Britten, J. 1888. The nomenclature o? Nymphaea, &c. J. Bot. 26: 6-11.
Candolle, A. P. de. 1821. Nymphaea. In Regni vegetabilis systema nattirale, by A. P. de Candolle, 2:
49-59. Paris: Treuttel et W?rtz.
Capperino, M. E., and E. L. Schneider. 1985. Floral biology o? Nymphaea mexicana Zuce. (Nympha
eaceae). Aquatic Bot. 23: 83-93.
Caspary R. 1865. Nymphaeaceae. Ann. Mus. Bot. Lugduno-Batavum 2: 241-253.
-. 1878. Nymphaeaceae. in Flora brasiliensis, eds. C. F. P. Martius and A. G. Eichler, 4(2):
129-184. Leipzig: Friedrich Fleischer.
-. 1888. Nymphaeaceae. In Die nat?rlichen Pflanzenfamilien, eds. A. Engler and K. Prantl, 3(2):
1-10. Leipzig: Wilhelm Engelmann.
Chaturvedi, M. 1974. Pollen grains of Nymphaea stellata L. (var. August Coach). Journal of Palynol
ogy 10(2): 126-131.
Chifflot, J. B. J. 1902. Contributions, a l'?tude de la classe de Nymph?in?es. Ann. Univ. Lyon, s?r. 2,
10: 1-294.
Collinson, M. E. 1980. Recent and Tertiary seeds of the Nymphaeaceae sensu lato with a revision of
Brasenia ovula (Brong.) Reid and Chandler. Ann. Bot. 46: 603-632.
Conard, H. S. 1905. The waterlilies: a monograph of the genus Nymphaea. Publ. Carnegie Inst. Wash.
4: 1-279.
-. 1916. Nymphaea and Nuphar again. Rhodora 18: 161-164.
Cramer, J. M., A. D. J. Meeuse, and P. A. Teunissen. 1975. A note on the pollination of nocturnally
flowering species of Nymphaea. Acta. Bot. Neerl. 24: 489-490.
Gaudet, J. 1960. Ontogeny of the foliar sclereids in Nymphaea odorata. Amer. J. Bot. 47: 525-532.
Gornall, R. J., B. A. B?hm, and R. Dahlgren. 1979. The distribution of flavonoids in the angio
sperms. Bot. Not. 132: 1-30.
Greene, E. L. 1887a. Bibliographical notes on well-known plants. III. Nymphaea and Nuphar. Bull.
Torrey Bot. Club 14: 177-179.
-. 1887b. Bibliographical notes on well-known plants. VI. Nymphaea and Nuphar. Bull. Torrey
Bot. Club 14: 257-258.
-. 1888. Bibliographical notes on well-known plants. VIL Castalia and Nymphaea. Bull. Torrey
Bot. Club 15: 84-85.
Gupta, P. P. 1978. Cytogenetics of aquatic ornamentals. II. Cytology of Nymphaeas. Cytologia 43:
477-484.
-. 1980. Cytogenetics of aquatic ornamentals. VI. Evolutionary trends and relationships in the
genus Nymphaea. Cytologia 45: 307-314.
Harborne, J. B., T. J. Mabry, and H. Mabry, eds. 1975. The flavonoids. New York: Academic Press.
Harborne, J. B., and C. A. Williams. 1975. Flavone and flavonol glycosides. In The flavonoids, eds. J.
B. Harborne, T. J. Mabry, and H. Mabry, 376-441. New York: Academic Press.
-. 1982. Flavone and flavonol glycosides. In The flavonoids. Advances in research, eds. J. B.
Harborne and T. J. Mabry, 261-311. New York: Chapman and Hall.
Henkel, F., F. Rehnelt, and L. Dittmann. 1907. Das Buch der Nymphaeaceen oder Seerosengew?chse.
Darmstadt: Friedrich Henkel.
Howard, G., and T. J. Mabry. 1970. Myricetin 3-0-methyl ether 3'-0-8-D-glucoside, the major
flavonoid of Oenothera speciosa (Onagraceae). Phytochemistry 9: 2413-2414.
Jones, K. 1976. Multiple Robertsonian fusions in the evolution of a plant genus. In Current chromoso
mal research, eds. K. Jones and P. E. Brandham, 220-221. Amsterdam: Elsevier/North-Holland
Biom?dical Press.
Jones, M. R., and G. C. S. Clarke. 1981. The Northwest European pollen flora, 25. Nymphaeaceae.
Review of Palaeobotany.and Palynology 33: 57-67.
Khanna, P. 1967. Morphological and embryological studies in Nymphaeaceae. III. Victoria cruziana
D'Orb. and Nymphaea stellata Willd. Bot. Mag. (Tokyo) 80: 305-312.
Kuprianova, L. A. 1976. Pollen morphology of some species of Nymphaea from the European part of
the USSR. Bot. Zum. (Moscow & Leningrad) 61: 1558-1563 (in Russian).
Langlet, O., and E. S?derberg. 1927. ?ber die Chromosomenzahlen einiger Nymphaeaceen. Acta
Horti Berg. 9: 85-104.
Lawson, G. 1888. On the Nymphaeaceae. Proc. & Trans. Roy. Soc. Canada 6: 97-125.
Lehmann, J. G. C. 1853. ?ber die^Gattung Nymphaea. Hamburger Garten-Blumenzeitung 9:193-218.
-. 1854. Index seminum in horto bot?nico hamburgensi. Ann. Sei. Nat. Bot., ser 4, 1: 323-327.
Linnaeus, C. 1753. Species plantarum. Stockholm.
Lynch, S. P., and G. L. Webster. 1975. A new technique of preparing pollen for scanning electron
microscopy. Grana 15: 127-136.
Mabry, T. J., K. R. Markham, and M. B. Thomas. 1970. The systematic identification of flavonoids.
New York: Springer.
Mackenzie, K. K. 1927. Proper use of the name Nymphaea. Rhodora 29: 234-237.
Malaviya, M. 1962. A study of sclereids in three species of Nymphaea. Proc. Indian Acad. Sei. 56B:
232-236.
Mauve, A. A. 1967. Water-lilies in South Africa. Fauna Flora (Transvaal) 18: 31-35.
Meeuse, B. J. D., and E. L. Schneider. 1980. Nymphaea revisited: a preliminary communication.
Israel J. Bot. 28: 65-79.
Meyer, N. R. 1964. Palynological studies in Nymphaeaceae. Bot. Zum. (Moscow & Leningrad) 49:
1421-1429 (in Russian).
Miele, F., A. Fabian, ancf N. Bodocan. 1969. Anthesis in Victoria regia. N. Stud. Univ. Babes-Bolyai,
Ser. Biol. 14: 57-62 (in Rumanian). (Not seen, information from Harborne et al. 1975.)
Mitra, R. L., and K. Subramanyam. 1982. Is Nymphaea rubra Roxb. ex Andrews an apomict? Bull.
Bot. Surv. India 24: 83-86.
Mohan Ram, H. Y., and V.L. Nayyar. 1974. A case of reversion of flower of Nymphaea mexicana to
vegetative condition. Curr. Sei. 43: 290-291.
Moseley, M. F., Jr. 1961. Morphological studies of the Nymphaeaceae. II. The flower of Nymphaea.
Bot. Gaz. (Crawfordsville) 122: 233-259.
Planch?n, J. E. 1852. Enumeration succincte des esp?ces de la famille des Nymph?ac?es. FI. Serres
Jard. Eur. 8: 117-120.
-. 1853. Etudes sur les Nymph?ac?es. Ann. Sei. Nat. Bot., s?r. 3, 19: 17-63.
Prance, G. T. 1980. A note on the pollination of Nymphaea amazonum Mart. & Zuce. (Nymphaea
ceae). Brittonia 32(4): 505-507.
Prance, G. T., and A. B. Anderson. 1976. Studies of the floral biology of neotropical Nymphaeaceae.
Acta Amaz?nica 6: 163-170.
Pring, G. H. 1934a. Hybrid Nymphaeas. Missouri Bot. Gard. Bull. 22: 47-90.
-. 1934b. Hybrid Nymphaeas: Their creation, propagation, and cultivation. Missouri Bot. Gard.
Bull. 22: 93-108.
-. 1941. Tropical water-lilies. Missouri Bot. Gard. Bull. 29: 57-76.
-. 1949. Water-lilies. Missouri Bot. Gard. Bull. 37: 65-88.
Pring, G. W. 1960. Tropical water lilies. Missouri Bot. Gard. Bull. 48: 93-98.
Radford, A. E., W. C. Dickison, J. R. Massey, and C. R. Bell, eds. 1974. Vascular plant systematics.
New York: Harper & Row.
Rao, P. S., and T. R. Seshadri. 1941. Constitution of cannabiscitrin. Proc. Indian Acad. Sei. 14A:
265-269.
Rao, T. A., and B. C. Banerjee. 1979. Foliar sclereids in the Nymphaeaceae sensu lato and their use
in familial classification. Proc. Indian Acad. Sei. 88B: 413-422.
Rizk, A. M., A. M. Youssef, M. A. Diab, and H. M. Salem. 1976. Constituents of Egyptian Euphor
biaceae. Part 2: Flavonoids of Euphorbia paralias. Pharmazie 31: 405.
Rohlf, F. J., J. Kishpaugh, and D. Kirk. 1977. Numerical taxonomy system of multivariate statistical
programs. Program printout.
Sagot, P. 1881. Catalogue des plantes phan?rogames et cryptogames vasculaires de la Guyane
Fran?aise. Ann. Sei. Nat. Bot., s?r. 6, 2: 134-180.
Salisbury, R. A. 1806. Description of the natural order of Nymphaeaceae. Ann. Bot. (K?nig & Sims)
2: 69-76.
Schmucker, T. 1932. Physiologische und ?kologische Untersuchungen an Bl?ten tropischer Nym
phaea-Arten. Planta 16: 376-412.
-. 1933. Zur Bl?tenbiologie tropischer Nymphaea-Arten. II. Bor als entscheidender Faktor.
Planta 18: 641-650.
-. 1935. ?ber den Einfluss von Bors?ure auf Pflanzen, insbesondere keimende Pollenk?rner.
Planta 23: 264-283.
Schneider, E. L. 1978. Morphological studies of the Nymphaeaceae. IX. The seed of Barclaya longifo
lia Wall. Bot. Gaz. (Crawfordsville) 139: 223-230.
-. 1979. Pollination biology of the Nymphaeaceae. Proc. IVth Int. Sym. on Pollination. Mary
land Agrie. Exp. Sta. Special Misc. Publ. 1: 419-429.
-. 1982a. Notes on the floral biology of Nymphaea elegans (Nymphaeaceae) in Texas. Aquatic
Bot. 12: 197-200.
-. 1982b. Observations on the pollination biology of Nymphaea gigantea W. J. Hooker (Nym
phaeaceae). W. Austral. Naturalist 15(3): 71-72.
Schneider, E. L., and T. Chaney. 1981. The floral biology o? Nymphaea odorata (Nymphaeaceae).
Southw. Naturalist 26: 159-165.
Schneider, E. L., and E. G. Ford. 1978. Morphological studies of the Nymphaeaceae. X. The seed of
Ondinea purpurea den Hartog. Bull. Torrey Bot. Club 105: 192-200.
Smith, J. E. 1809. Nymphaea, Nuphar. In Florae graecae prodromus, eds. J. Sibthorp and J. E. Smith,
1: 360-362. London.
Stebbins, G. L. 1971. Chromosomal evolution in higher plants. London: Edward Arnold Ltd.
Taku, E. P., A. I. Tikhonov, and V. I. Litvinenko. 1970. Flavonoids o? Nymphaea alba. Khim. Prir.
Soedin. 6(5): 629-630 (in Russian).
Trickett, R. S. 1971. A new tropical American waterlily, Nymphaea belophylla. Kew Bull. 26: 29-31.
Ueno, J., and S. Kitaguchi. 1961. On the fine structure of the pollen walls of Angiospermae. I.
Nymphaeaceae. J. Biol. Osaka City Univ. 12: 83-90.
Ule, E. 1915. Nymphaeaceae. Notizbl. K?nigl. Bot. Gart. Berlin 6: 293-295.
Wagner, J. 1964. ?ber Inhaltsstoffe des Rosskastaniensamens. III. Untersuchungen an Flavonolen.
Hoppe-Seylers Z. Physiol. Chem. 335: 232-249.
Walker, J. W. 1974. Aperture evolution in the pollen of primitive angiosperms. Amer. J. Bot. 61:
1112-1136.
Walker, J. W., and J. A. Doyle. 1975. The bases of angiosperm phylogeny: palynology. Ann. Mis
souri Bot. Gard. 62: 664-723.
Walpers, W. G. 1842. Repertorium botanices systematicae, vol. 1. Leipzig: Friederich Hofmeister.
Ward, D. B. 1977. Night-flowering waterlilies in Florida. Florida Scientist 40: 155-159.
Weberbauer, A. 1894. Beitr?ge zur Samenanatomie der Nymphaeaceen. Bot. Jahrb. Syst. 18: 213
258.
White, M. J. D. 1973. Animal cytology and evolution. Cambridge: Cambridge University Press.
Wiersema, J. H. 1982. Distributional records for Nymphaea lotus (Nymphaeaceae) in the western
hemisphere. Sida 9: 230-234.
-. 1984a. Systematics o? Nymphaea subgenus Hydrocallis (Nymphaeaceae). I. Four new species
from the neotropics. Brittonia 36: 213-232.
-. 1984b. A revision o? Nymphaea subgenus Hydrocallis (Nymphaeaceae). Ph. D. thesis, Univer
sity of Alabama.
Wodehouse, R. P. 1935. Pollen grains. New York: McGraw-Hill Book Company, Inc.
Wohlfart, V. R., and R. Gademann. 1974. ?ber das Pigmentmuster einiger Arten aus der Familie der
Nymphaeaceae. Deutsche Apotheker-Zeitung 114: 1279-1281.
Wood, C. E., Jr. 1959. The genera of Nymphaeaceae and Ceratophyllaceae in the southeastern
United States. J. Arnold Arbor. 40: 94-112.
APPENDIX 1
Chromosome numbers of species of Nymphaea subg. Hydrocallis; vouchers are deposited at

UNA.
N. amazonum subsp. amazonum: 2n = 18; Gu?rico, Venezuela, Wiersema 1400; Guayas, Ecua
dor, Wiersema et al. 2013 (cultivated), 2027 (cultivated).
M amazonum subsp. pedersenii: 2n = 18; Corrientes, Argentina, Wiersema et al. 2270 (culti
vated).
N. conardii: 2n = 28; Barinas, Venezuela, Wiersema & Gonzalez 2214 (cultivated); Chontales,
Nicaragua, Haynes 8588 (cultivated).
N. gardneriana: 2n = 28; Gu?rico, Venezuela, Wiersema 1403 (cultivated), 1412 (cultivated),
1428; Chaco, Argentina, Wiersema etal. 2247a (cultivated).
N. jamesoniana: 2n = 28; Guayas, Ecuador, Wiersema et al. 2026 (cultivated); Florida, U.S.A.,
Wiersema 2031 (cultivated); Corrientes, Argentina, Wiersema et al. 2277 (cultivated).
N. lasiophylla: 2n = 18; Alagoas, Brazil, Wiersema et al. 2293 (cultivated).
N. lingulata: 2n = 18; Maranhao, Brazil, Wiersema et al. 2308 (cultivated).
N. novogranatensis: 2n = 28; Gu?rico, Venezuela, Wiersema 1420, Wiersema & Gonz?lez 2234
(cultivated).
N. oxyp?tala: 2n = 84; Apure, Venezuela, Wiersema & Gonz?lez 2232 (cultivated).
N. prolif?ra: 2n = 18; Guayas, Ecuador, Wiersema et al. 2028; Corrientes, Argentina, Wiersema
et al. 2273 (cultivated).
N. rudgeana: 2n = 42?; Miranda Venezuela, Wiersema & Gonz?lez 2205 (cultivated).
N. tenerinervia: 2n = 20; Maranh?o, Brazil, Wiersema et al. 2307 (cultivated).
?N. amazonum x N. prolif?ra: 3n = 27; Corrientes, Argentina, Wiersema etal. 2238 (cultivated),
2249 (cultivated), Schinini & Crist?bal 14253 (this count obtained by Camilo Quarin of Corrientes,
Argentina).
APPENDIX 2
Structure and occurrence of compounds isolated from leaves of species of Nymphaea. Unless
indicated otherwise, all collection numbers are those of the author; vouchers are deposited at UNA.
C-Glycosylflavones (Fig. 38a).

Cl-Ri = C-glucosyl, R2 = H, R3 = H (vitexin).
C2-Ri = C-glucosyl, R2 = H, R3 = OH (orientin).
C3-Ri = H, R2 = C-glucosyl, R3 = OH (isorientin).
Kaempferols (Fig. 38b).

Kl-R = rhamnosyl (kaempferol 3-O-rhamnoside).
K2-R = acylrhamnosyl (kaempferol 3-O-acylrhamnoside).
K3-R = acylgalactosyl I (kaempferol 3-O-acylgalactoside I).
K4-R = acylgalactosyl II (kaempferol 3-O-acylgalactoside II).
K5-R = acylgalactosyl III (kaempferol 3-O-acylgalactoside III).
K6-R = acylgalactosyl IV (kaempferol 3-O-acylgalactoside IV).
Quercetins (Fig. 38c).

Ql-Ri = glucosyl; R2, R3, R4 = H (quercetin 3-O-glucoside).
Q2-R3 = glucosyl; Ri, R2, R4 = H (quercetin 3'-0-glucoside).
Q3-R4 = glucosyl; Ri, R2, R3 = H (quercetin 4'-0-glucoside).
Q4-R3 = diglucosyl; Rls R2, R4 = H (quercetin 3'-0-diglucoside).
Q5-Ri = galactosyl; R2, R3, R4 = H (quercetin 3-O-galactoside).
Qfr^R? = acylgalactosyl I; R2, R3, R4 = H (quercetin 3-O-acylgalactoside I).
Q7-Ri = acylgalactosyl II; R2, R3, R4 = H (quercetin 3-O-acylgalactoside II).
Q8-R1 = acylgalactosyl III; R2, R3, R4 = H (quercetin 3-O-acylgalactoside III).
Q9-R! = rhamnosyl; R2, R3, R4 = H (quercetin 3-O-rhamnoside; quercetin).
Q10-R3 =xylosyl; Ri, R2, R4 = H (quercetrin 3'-0-xyloside).
Qll-Ri = rhamnoglucosyl; R2, R3, R4 = H (quercetin 3-O-rutinoside).
Il-Ri = rhamnosyl; R3 = Me; R2, R4 = H (isorhamnetin 3-O-rhamnoside).
12-Ri = acylrhamnosyl; R3 = Me; R2, R4 = H (isorhamnetin 3-O-acylrhamnoside).
I3-fti =galactosyl; R3 = Me; R2, R4 = H (isorhamnetin 3-O-galactoside).
I4-Ri = acylgalactosyl I; R3 = Me; R2, R4 = H (isorhamnetin 3-O-acylgalactoside I).
I5-Ri = acylgalactosyl II; R3 = Me; R2, R4 = H (isorhamnetin 3-O-acylgalactoside II).
16-Rx = acylgalactosyl III; R3 = Me; R2, R4 = H (isorhamnetin 3-O-acylgalactoside III).
OH O
OH O OR2 O
FIG. 38. Base structures for compounds
ferols. c. Quercetins. d. Myricetins. e. Unk
I7-Ri = diglucosyl; R3 = Me; R2, R4 = H

Rl-Ri = glucosyl; R2 = Me; R3, R4 = H
R2-Ri = rhamnosyl; R2 = Me; R3, R4 =
R3-Ri = rhamnoglucosyl; R2 = Me; R3,
mQl-Ri = Me; R2 = glucosyl; R3, R4 =
mQ2-R! = Me; R3 = xylosyl; R2, R4 =
mQ3-Ri = rhamnosyl; R2, R3 = Me; R4
Myricetins (Fig. 38d).

Ml-Ri = glucosyl; R2, R3, R4 = H (my
M2-R3 = glucosyl; Ri, R2, R4 = H (my
M3-Ri = rhamnosyl; R2, R3, R4 = H (m
M4-Ri = arabinosyl; R2, R3, R4 = H (m
M5-Ri = acylgalactosyl; R2, R3, R4 = H
M6-R3 = xylosyl; Ri, R2, R4 = H (myr
M7-R3 = diglucosyl; Ri, R2, R4 = H (m
M8-R1 = rhamnoglucosyl; R2, R3, R4 = H (myricetin 3-O-rutinoside).

mMl-R! = Me; R2, R3, R4 = H (3-methylmyricetin).
mM2-Ri = Me; R2 = glucosyl; R3, R4 = H (3-methylmyricetin 7-O-glucoside).
mM3-Ri = Me; R3 = xylosyl; R2, R4 = H (3-methylmyricetin 3'-0-xyloside).
mM4-Ri = Me; R3 = diglucosyl; R2, R4 = H (3-methylmyricetin 3'-0-diglucoside).
mM5-Ri = Me; R3 = glucosyl; R4 = xylosyl; R2 = H (3-methylmyricetin 3'-0-glucoside,
5-O-xyloside).
Unknowns (Fig. 38e).

A-xanthone?
B-xanthone?
C-?
Fl-R!, R2 = Me?; R3, R4, R5, R? = ?
F2-R!, R2 = Me?; R3, R4, R5, R? = ?
F3-R!, R2 = Me?; R3, R4, R5, R? = ?
F4-R!, R2 = Me?; R3, R4, R5, R? = ?
D-?
I8-R4 = Me; R2, R5, R? = H; Rlf R3 = ?
mQ4-R3 = glucosyl; R2, R5 = H; Ri = Me; R4 = Me?; R? = ? (3,3'-dimethylquerc
glucoside?).
Subgenus Hydrocallis.
N. amazonum subsp. amazonum. 1436 (Venezuela): C2, Ql, Q2, Ml, M5, A, B, C. 2013
(Ecuador): C2, Ql, Ml, M5, A, B, C. 2027 (Ecuador): C2, C3, Ql, Ml, M5, A, C. 2291 (Brazil): C2,
Ql, Ml, M5, A, B, C. 2221A (Venezuela): C2, Q2, Q3, A, B, C.
N. amazonum subsp. pedersenii. 2270 (Argentina): Cl, C2, C3, Ql, Q2, Q3, Q10, Qll, Ml,
M6?, M8, A, B, C.
N. amazonum subsp. amazonum x N. lasiophylla. 2296 (Brazil): C2, C3, Ql, Q2, Q4, Q7, Q8,
A, B, C, D.
N. conardii. 2214 (Venezuela): Cl, K3, K4, K5, K6. 2230 (Venezuela): 13, 14, 15, 16. 2289
(Brazil): Cl, II, 12, 13, 17,18. Haynes 8588 (Nicaragua): Cl, Kl, K2, Q9. Haynes 8612 (Nicaragua):
Cl, Kl, K2.
N. gardneriana. 1403 (Venezuela): C2, Ql, Q9, RI, R3, Ml, M2. 1412 (Venezuela): Ql, Q9,
Qll, Ml. 1428 (Venezuela): Ql, Q2, Q9, R2, R3, mQ3. 2210 (Venezuela): Ql, Q9, R2, Ml. 2247A
(Argentina): C2, Ql, Q9, R2.
N. jamesoniana. 1343 (Mexico): Ql, Q9. 2030 (Ecuador): Ql, Q9. 2032 (Florida): Ql, Q9. 2265
(Argentina): Ql, Q9. 2290 (Brazil): Ql, Q9.
N. lasiophylla. 2293 (Brazil): M2, M7, mM3, mM4, A, B, C, FI, F2, F3, F4, mQ4. 2316 (Brazil):
M2, M7, mM3, mM4, A, B, C, FI, F2, F3, F4, mQ4.
N. lingulata. 2283A (Brazil): C3, mQl, mQ2, mMl, mM2, mM3, mM5, A, B, C. 2308 (Brazil):
mQl, mQ2, mMl, mM2, mM3, mM5, A, B, C.
N. novogranatensis. 1404 (Venezuela): C2, Ql, Q5, Q6, Q10, Qll, A, B, C. 1411 (Venezuela):
C2, Ql, Q5, Q6, Q9, Q10, Qll, A, B, C. 1420 (Venezuela): C2, Ql, Q5, Q6, Qll, A, B, C. 2234
(Venezuela): C2, Ql, Q5, Q6, Q9, Qll, A, B, C.
N. oxypetala. 2232 (Venezuela): Ql, Q9, mQ2?, Ml, A, B, C.
N. prolif?ra. 2028 (Ecuador): C2, Ql, Q8?, Qll, Ml, M4, M8, A, B, C. 2241 (Argentina): C2,
Ql, Q8?, Qll, Ml, M4, M8, A, B, C. 2273 (Argentina): C2, Ql, Q8?, Qll, Ml, M4, M8, A, B, C.
N. prolif?ra x N. amazonum subsp. pedersenii? 2238 (Argentina): C2, Ql, Qll, Ml, M4, M8,
A, B, C. 2240 (Argentina): C2, Ql, Q10, Qll, Ml, M4, M6, M8, A, B, C. 2249 (Argentina): C2, Ql,
Qll, Ml, M4, M8, A, B, C. 2254 (Argentina): C2, Ql, Qll, Ml, M4, M8, A, B, C. 2255 (Argen
tina): C2, Ql, Qll, Ml, M4, M8, A, B, C.
N. rudgeana. 2205 (Venezuela): Cl, C2, C3, Q2, Q3, Q9, A, B, C. 2285 (Brazil): Cl, C2, C3,
Ql, Q2, Q3, Q9, A, B, C. 2299 (Brazil): Cl, C2, Ql, Q2, Q3, Q9, A, B, C.
N. tenerinervia. 2307 (Brazil): C2, Ql, Ml, M5, M6, A, C. 2310 (Brazil): C2, Ql, Q10, Ml, M6,
A, B, C. 2311 (Brazil): C2, Ql, Q10, Ml, M5, M6, A, B, C. 2314 (Brazil): C2, Ql, Ml, M5, M6, A,
B, C.
Subgenus Brachyceras.
N. ampia. 1345 (Mexico): C2, Ql, Q9, Q10. 1452 (Florida): C2, Ql, Q9, Q10, mQ2, mMl,
mM3, A, C. 2021 (Ecuador): C2, Ql, Q9, Q10, mQ2, A, C.
N. elegans. 1981 (Florida): C2, Q9, Q10, mQ2, M3, M6, mMl, mM3, A, B.
Subgenus Lotos.
N. lotos. 1442 (Florida): Cl, C2, M6, A, B, C.
Subgenus Nymphaea.
N. candida. 2196A (Germany): Q9, mQ2, M6, mM3, A, B.
N. mexicana. 281 (Alabama): C2, Ql, Qll.
N. odorata. 274 (Alabama): C2, Q9, mQ2, Ml?, M3, M6, mMl, mM3, A, B.
NUMERICAL LIST OF SPECIES
1. Nymphaea jamesoniana 9. Nymphaea tenerinervia

2. Nymphaea conardii 10a. Nymphaea amazonum subsp. amazonum
3. Nymphaea gardneriana 10b. Nymphaea amazonum subsp. pedersenii
4. Nymphaea glandulifera 11. Nymphaea prolif?ra
5. Nymphaea belophylla 12. Nymphaea lasiophylla
6. Nymphaea potamophila 13. Nymphaea lingulata
7. Nymphaea oxypetala 14. Nymphaea rudgeana
8. Nymphaea novogranatensis
INDEX TO NUMBERED COLLECTIONS EXAMINED
The numbers in parentheses refer to the corresponding species in the text and in the Numerical
List of Species presented above.
Adams, C. D. 13218 (14). Breedlove, D. E. 26918 (1); 27347 (2); 37370

Aguilar, R. M. 751 (3); 1045 (3). (1).
Ahumada, O. 1521 (11). Breedlove & Thome 20917 (1).
Ahumada et al. 876 (3). Britton, E. G. 5093 (10a).
Allen, P. H. 6278 (4). Britton & Britton 2887 (14); 7086 (10a); 9234
Anderson, A. B. 159 (14); 161 (14); 223 (14); (10a).
224 (10a). Britton et al. 6129 (14); 6762 (10a); 10238 (14).
Andr?, E. F. 2567 (10a). Broadway, W. E. 276 (14); 919 (14); 3366 (10a);
Anisits, J. D. 2703 (3); 2811 (3). 4560 (10a).
Araujo & Maciel 3822 (10a); 4056 (14). Bunting & Luis 6497 (14).
Arenas, P. 2060 (1); 2272 (3). Burchell, W. J. 8569 (3); 9497 (10a); 9556 (14).
Asplund, E. 14388 (4); 16024 (11). Burger et al. 11307 (11).
Badillo & Holmquist 6239 (14). Camargo de Abreu, L. 15 (14); 25 (9).
Baker & Burger 26 (4). Carreira, L. 182 (14).
Balansa, M. 521 (3); 522 (11). Casari, M. B. 470 (14); 500 (14).
Barbour, P. J. 4408 (4); 4885 (10a). Casari et al. 507 (14); 509 (14); 523 (12).
Barreto, M. 11319 (13); 7230 (10a). Cooley et al. 7151 (1).
Berg et al. P17628 (4). Cowan & Soderstrom 2224 (14).
Black, G. A. 47-2105 (10a); 50-9086 (10a); 52 Cramer, J. M. 15 (14); 18 (10a); 23 (10a);
15458 (4). LBB14846 (10a); LBB14847 (10a);
Black & Cordeiro 52-15053 (10a). LBB14848 (10a); LBB14849 (10a);
Black & Lobato 50-9534 (14). LBB14850 (10a); LBB14851 (10a);
Blanchet, N. G. 11 (10a); 1029 (10a). LBB14854 (10a); LBB14855 (4); LBB14856
Brantjes, N. B. M. 706902 (14); 707701 (14); (4); LBB14857 (4); LBB14858 (14);
707801 (2); 707901 (14). LBB14859 (14); LBB14860 (14); LBB14861
(14); LBB14862 (14); LBB14863 (14); Hassler, E. 2804 (7).

LBB14865 (14); LBB14866 (14); LBB14867 Hatschbach, G. 16387 (14); 39372 (3).
(14); LBB14868 (14); LBB14876 (4); Haught, O. 4431 (2).
LBB14877 (10a); LBB14902 (4); LBB14907 Haynes, R. R. 5441 (1); 8430 (1); 8572 (1); 8588
(10a); LBB14908 (4); LBB14913 (14); (2); 8593 (10a); 8612 (2).
LBB14915 (4); LBB14919 (14). Hazen, T. E. 2077 (4); 2079 (14).
Croat, T. B. 12232 (2); 16702 (2). Hedger, J. N. 75 (10a); 326 (4).
Cuatrecasas, J. 19656 (10a). Hekking, W. H. A. 782 (14); 940 (10a); 954
Cumana Campos & Lampe 531 (10a). (10a); 955 (14).
da Silva, C. J. 212 (10b); 219 (11); 395 (10b); Heringer et al. 868 (1).
427 (1); 440 (10b); 441 (10b); 452 (10a); Hitchcock, A. S. 16868 (14).
459 (10b). Holton, I. F. 689 (10a).
Davidson & Donahue 8971 (4). Hostmann 565 (10a).
de la Cruz, J. S. 3563 (14); 3969 (14). Howard & Howard 9846 (2).
Delascio et al. 4493 (7). Jameson, W. 334 (7); 546 (1).
Dirven, Ir. J. G. P. LP431 (10a). Janzen, D. H. T-68 (11).
Dodge et al. 16712 (2); 16713 (10a). Jenman, G. S. 5746 (4); 5747 (10a); 5748 (14);
Dodson, C. H. 6172 (4). 7230 (14); 7773 (4); 7775 (10a); 7776 (10a);
Don, G. 146 (14). 7777 (10a); 7778 (14); 7779 (14); 7892 (14);
Drouet, F. 1955 (14); 2073 (14); 2140 (12); 2174 7986 (14).
(13); 2292 (12); 2486 (12); 2511 (10a); 2512 Jobert 337 (14).
(14); 2563 (12). Jonker-Verhoef & Jonker 213 (14).
Duchassaing, E. P. 16 (14). Jorgensen, P. 3124 (3).
Duke, J. A. 5888 (2). Junk, W. 145 (14).
Duke & Bristan 283 (2). Kegel, H. A. H. 932 (10a).
Duss, P. 3702 (10a). Killip & Smith 14613 (8); 14821 (2); 27173 (4);
Ebinger, J. E. 553 (2). 28682 (10a); 30531 (14).
Ekman, E. L. 2679 (2); 2680 (10a); 10035 (2); Krapovickas & Crist?bal 21733 (10b); 33903
13277 (2); 13330 (10a); 13710 (10a); 13710 (11).
(7); 15949 (10a); 17007 (2); 17857 (2). Krapovickas & Schinini 32094 (3); 36605 (3).
Elias, H. 1387 (1). Krapovickas et al. 29505 (10b).
Elias & Kirkbride 1618 (2). Krukoff, B. A. 6969 (14).
Fassett, N. C. 28279 (2); 28565 (1); 28615 (1); Kuntze, O. 340 (10a).
28738 (11); 29099 (2); 29326 (10a); 29381 Lanjouw & Lindemann 630 (14); 1507 (14);
(11). 3055 (14); 3146 (10a).
Fern?ndez, A. 1299 (7). Lehmann, F. C. 1007 (10a).
Fern?ndez, F. 306 (10a). Lentz, D. L. 1187 (1).
Florsch?tz & Florsch?tz 1004 (4). Le?n, Bro. 7523 (10a); 11522 (10a).
Forest Dept. M412 (10a). Le?n & Nivaro 7609 (10a).
Friedrichsthal, E. von 1878 (4). Les, D. H. 114 (1); 140 (1); 159 (1).
Fr?es, R. L. 22686 (6); 26805 (14); 26806 (14). Liesner, R. 2892 (4).
Fuchs et al. 22176 (4). Lindemann, J. C. 4332 (14).
Gardner, G. 1236 (10a); 2474 (10a); 2476 (3); Lindman, C. A. M. 2877 (7).
2477 (10a); 3568 (3). L?fgren, A. 729 (12).
Gamier, A. 298 (1). Lowrie et al. 720 (10a).
Gentry, A. 6194 (2). L?tzelburg, P. von 20502 (1); 20973 (9).
Gentry & Tyson 1696a (2). Maas & Plowman 2159 (10a).
Gibert, M. 53 (1); 198 (10a). Mass et al. 5565 (10a).
Gines, H. 4963 (4); 5056 (4). Macedo, H. 2279 (13); 3320 (3).
Ginzberger, A. 657 (3). Macedo & Assumps?o 688 (10b).
Ginzberger & Hagmann 658 (14). Macedo et al. 452 (10a).
Glaziou, A. F. M. 12414 (13). Maciel & Bou?as 43 (9).
Goeldi, E. A. 1204 (14). Madison et al. 6155 (6).
Goodland, R. 1069 (14). Maguire, B. 25004 (14).
Hammel & D'Arcy 5007 (2). Maguire & Fanshawe 23405 (14).
Harley et al. 17129 (14). Maguire et al. 53507 (4).
Harris, S. A. TP458 (10a); TP502 (14). Maia et al. 124 (6).
Malme, G. O. 1764B (7). Schnell, R. 11221 (4).

March, M. 24 (14); 25 (10a). Schreiter, R. 2156 (1); 11035 (1).
Martin et al. 1614 (4). Schuhes, R. E. 10315 (14).
Martinez-Achenbach, G. 97 (1); 159 (1). Schuhes & L?pez 10135 (14); 10327 (14).
Martius, C. F. P. von, exsicc. 2377 (12); exsicc. Schwacke, C. A. W. 10015 (3).
3313 (10a). Shattuck, O. 1153 (2).
Matuda, E. 161210 (2); 16972 (2). Silva & Rosario 5078 (14).
Men?ndez, F. 255 (1). Silva et al. 975 (14).
Mennega, A. M. W. 306 (14). Sintenis, P. 5688 (2).
M?le, L. 1000 (7). Sioli, H. 4 (6); 49 (14).
Molina & Molina 22704 (2). Smith, H. H. 2108 (14).
Morel, I. 531 (1); 1810 (11). Smith et al. 6647 (10a).
Moreno, P. P. 11862 (2); 24967 (4). Soeprato 41G (14); 112 (14); 126 (14); 179 (4).
Morong, T. 1028 (1). Sparre & Vervoorst 2336 (11).
Mos?n, H. 1102 (3); 3339 (14). Spruce, R. 479 (14).
Nicora, E. G. 7132 (10b). Standley, P. C. 55153 (4); 72857 (4); 73128 (4).
Oldeman, R. A. A. 2859 (4). Steinitz-Kannan, M. 409 (4).
Opler, P. A. 926 (11). Stergios, B. 7077 (3).
Palacios-Cuezzo 223 (11). Stevens, F. L. 2463 (10a).
Pedersen, T. M. 4494 (10b); 7741 (1); 7743 (11); Stevens, W. D. 20848 (14).
8094 (11); 8335 (11). Steyermark, J. A. 39867 (4); 57693 (2); 87349
Pickel, D. B. 2343 (14); 3725 (12). (4); 87446 (14); 111374 (14).
Pires & Suva 4359 (14). Steyermark & Dunsterville 104108 (14).
Pittier, H. F. 1408 (14); 4560 (2); 6805 (2); 9376 Steyermark & Fern?ndez 99560 (2).
(4). Steyermark & Liesner 115121 (14); 119303 (2);
Poeppig, E. F. 3033 (14). 127627 (14); 127628 (14).
Prance, G. T. 23722 (8); 23724 (8). Steyermark & Manara 110570 (8).
Prance & Ramos 23521 (14). Steyermark & Rabe 96622 (2).
Prance et al. 7350 (4); 11912 (3); 23302 (4); Steyermark et al. 108549 (10a).
23384 (14); P25039 (14); 26205 (10b). Stork, H. E. 1897 (4).
Puig, F. C. 3050 (3). Tamayo, F. 701 (8); 3160 (14).
Quarin & Schinini 1063 (10b). Triana, J. 5020 (10a).
Ramia, M. 4540 (3). Trujillo, B. 2538 (8); 8881 (3); 11181 (2); 11329
Ramia & Montes 4701 (3); 5362 (7); 5532 (10a). (8); 11358 (8); 11430 (5); 11917 (10a);
Raynal-Roques, A. 19861 (10a); 21545 (10a); 12860 (10a); 12946 (10a); 13407 (3); 13467
21554 (14). (8); 13825 (3); 14380 (10a).
Raynal-Roques & J?r?mie 21101 (10a); 21127 Tyson & Clewell 5905 (2); 5922 (2).
(10a); 21169 (10a); 21178 (14); 21291 (14). Ule, E. H. G. 8120 (3); 8120 (9).
Restinga 418 (10a); 1307 (10a). Valerio, M. 408 (4).
Rimbach, A. 288 (4); 86 (4). van Donselaar, J. 2472 (14); 2548 (4); 2555 (14).
Rodrigues, F. M. 98 (3). Vareschi & Foldats 4532 (14).
Rodr?guez, J. V. 3431 (2). Ventura, F. 7681 (1).
Rodschied, E. K. 155 (4); 281 (14). Venturi, S. 1701 (1).
Romero-Casta?eda, R. 10869 (8). Walter, M. A. 135 (11); 136 (11).
Rose & Russell 21405 (10a). Warming, E. 1750 (10a).
Rovirosa, J. N. 674 (10a). Weddell, M. A. 3275 (11).
Rutkis, E. 158 (14); 159 (14); 164 (14); 169 (14); Wiersema, J. H. 1343 (1); 1400 (10a); 1403 (3);
175 (14). 1404 (8); 1411 (8); 1412 (3); 1420 (8); 1428
Sagot, P. 25 (14). (3); 1436 (10a); 2031 (1); 2032 (1).
Salazar, L. F. 27 (2). Wiersema & Gonz?lez 2203 (10a); 2205 (14);
Salzmann, P. 381 (14). 2210 (3); 2214 (2); 2215 (3); 2230 (2); 2232
Schinini, A. 8766 (11); 8769 (11); 9485 (11); (7); 2234 (8).
22006 (1). Wiersema & Pedersen 2248 (11).
Schinini & Crist?bal 10791 (10b). Wiersema & Vanni 2265 (1).
Schinini et al. 13270 (3); 17395 (10b). Wiersema et al. 2026 (1); 2027 (10a); 2028 (11);
Schipp, W. A. 1026 (2). 2029 (10a); 2030 (1); 2241 (11); 2247a (3);
NYMPHAEA SUBGENUS HYDROCALLIS 111
2270 (10b); 2273 (11); 2277 (1); 2284 (10a); 2310 (9); 2311 (9); 2312 (10a); 2313 (12);
2285 (14); 2286 (14); 2287 (10a); 2289 (2); 2314 (9); 2315 (1); 2316 (12); 2317 (12).
2290 (1); 2291 (10a); 2293 (12); 2294 (12); Woodson et al. 1225 (2).
2295 (12); 2297 (1); 2298 (12); 2299 (14); Woolston, A. L. 1257 (3).
2300 (1); 2302 (12); 2303 (10a); 2304 (10a); Wright, C. 2491 (10a); 3192 (2).
2306 (1); 2307 (9); 2308 (13); 2309 (14); Wright, S. 1 (12); 4 (12).
INDEX TO SCIENTIFIC NAMES
Accepted names are in roman type; the main entry for each is in boldface. Synonyms are in
italics.
Aesculus Linnaeus 44 12, 15, 21, 22, 23, 24, 26, 27, 28, 31, 43,
Barclaya Wallich 12 45, 46, 96, 108
Brasenia Schreber 24 subg. Hydrocallis (Planch?n) Conard 49-50
Cannabis Linnaeus 44 subg. Lotos (de Candolle) Conard 3, 5, 7, 8,
Castalia Salisbury 2, 48 12, 15, 21, 23, 24, 25, 26, 27, 28, 29, 32,
amazonum (Martius & Zuccarini) Britton & 37, 43, 44, 45, 46, 47, 48, 100, 108
P. Wilson 80 subg. Nymphaea 3, 4, 8, 9, 11, 12, 15, 21, 22,
ampia Salisbury 49, 85 23, 25, 26, 27, 28, 31, 43, 44, 45, 46, 47,
blanda (G. Meyer) Lawson 62 96, 108
gibertii Morong 51 subsect. Hydrocallis (Planch?n) Caspary 49
jamesoniana (Planch?n) Britton & P. Wilson tribe Chamaehtos Lehmann 49
51 alba Linnaeus 2, 25, 37, 44, 48, 65
rudgeana (G. Meyer) Britton & P. Wilson 96 albo-viridis Saint-Hilaire 81
Cyclocephala Latreille 32, 35 amazonum Martius & Zuccarini 2, 3, 8, 9, 12,
castanea Olivier 32, 65, 99 15, 25, 29, 32, 33, 34, 36, 38, 39, 40, 41,
mollis Endr?di 32, 86 42, 43, 44, 48, 49, 51, 65, 80-87, 90, 101
p?trida Burmeister 35 f. goudotiana (Planch?n) Caspary 82
verticalis Burmeister 32, 35, 57, 85, 99 subsp. amazonum 8, 17, 27, 33, 35, 81-85,
Euphorbia Linnaeus 44 86, 87, 93, 104, 107, 108
Euryale Salisbury 37 subsp. pedersenii Wiersema 8, 9, 18, 30,
ferox Salisbury 37 31, 33, 43, 44, 48, 81, 83, 84, 86-87, 90,
Hydrostemma Wallich 12 100, 104, 107, 108
Leuconymphaea Ludwig ex Kuntze 48 var. forma-submersa Sagot 85, 96, 100
amazonum (Martius & Zuccarini) Kuntze 80 var. goudotiana (Planch?n) Conard 82
blanda (G. Meyer) Kuntze 62 ampia (Salisbury) de Candolle 2, 12, 15, 22,
gardneriana (Planch?n) Kuntze 52 23, 26, 28, 39, 40, 43, 44, 49, 51, 59, 85,
goudotiana (Planch?n) Kuntze 82 100, 101, 108
jamesoniana (Planch?n) Kuntze 51 var. rudgeana (G. Meyer) de Candolle 96
lasiophylla (Martius & Zuccarini) Kuntze 90 var. speciosa (Martius & Zuccarini) Cas
oxyp?tala (Planch?n) Kuntze 71 pary 100
Nelumbo Adanson 2 belophylla Trickett 3, 7, 8, 24, 31, 43, 47, 48,
Nuphar Smith 2 50, 61, 65-68, 71, 73, 108
lutea (Linnaeus) Smith 37 blanda G. Meyer 2, 3, 12, 62, 63, 65
Nymphaea Linnaeus 48-49 var. amazonum (Martius & Zuccarini)
sect. Castalia (Salisbury) de Candolle 2 Planch?n 80
sect. Hydrocallis Planch?n 2, 49 var. fenzliana (Lehmann) Caspary 62
sect. Lotos de Candolle 2 caerulea Savigny 37
sect. Lytopleura Caspary 2 candida Presl 25, 37, 39, 40, 43, 44, 108
sect. Symphytopleura Caspary 2 candolleana Lehmann 82, 85
subg. Anecphya (Caspary) Conard 12, 21, 24, capensis Thunberg 26, 28, 37, 100
25, 26, 27, 31, 45, 46 conardii Wiersema 7, 8, 9, 11, 12, 15, 18, 19,
subg. Brachyceras (Caspary) Conard 3, 5, 9, 21, 28, 29, 30, 32, 34, 35, 36, 38, 39, 40,
112 SYSTEMATIC BOTANY MONOGRAPHS
41, 42, 47, 51, 55-58, 61, 65, 66, 77, odorata Ait?n 4, 15, 22, 23, 25, 26, 28, 31,
104, 107, 108 39, 40, 43, 44, 80, 100, 108
dentata Schumacher 25 oxypetala Planch?n 3, 7, 8, 9, 10, 12, 18, 20,
elegans Hooker 15, 22, 23, 26, 28, 31, 39, 40, 27, 30, 31, 35, 38, 39, 40, 43, 46, 47, 50,
43, 44, 108 65, 68, 69, 71-74, 104, 107, 108
fenzliana Lehmann 62, 65 passiflora Lehmann 62
foetida Gardner ex Lehmann 100 potamophila Wiersema 7, 8, 10, 18, 20, 31,
fragrans Gardner ex Caspary 100 43, 47, 50, 65, 68-71, 72, 108
gardneriana Planch?n 3, 7, 8, 9, 17, 18, 28, prolif?ra Wiersema 8, 9, 11, 18, 24, 27, 29,
29, 30, 34, 35, 36, 38, 39, 40, 41, 42, 47, 30, 31, 32, 33, 34, 36, 38, 39, 40, 41, 42,
51, 54, 58-62, 65, 66, 96, 104, 107, 108 48, 51, 76, 81, 87-90, 91, 93, 104, 107,
gibertii (Morong) Conard 2, 51 108
gigantea Hooker 25, 31 raja Lehmann 71, 73
glandulifera Rodschied 7, 8, 9, 12, 18, 20, 21, rubra Roxburgh 37
27, 31, 32, 34, 43, 47, 51, 58, 62-65, 69, rudgeana G. Meyer 2, 3, 6, 7, 8, 9, 10, 12,
72, 96, 108 13, 15, 16, 23, 26, 27, 29, 31, 32, 34, 36,
goudotiana Planch?n 81, 85 37, 38, 39, 40, 43, 46, 48, 50, 62, 81, 85,
integrifolia Salzmann ex Lehmann 100 96-100, 104, 107, 108
jamesoniana Planch?n 3, 7, 8, 9, 11, 12, 13, var. amazonum (Martius & Zuccarini)
18, 19, 21, 23, 26, 28, 29, 30, 32, 34, 35, Grisebach 80
36, 37, 38, 39, 40, 42, 47, 50, 51-55, 58, sagittariaefolia Lehmann 51, 55
61, 63, 65, 77, 104, 107, 108 sinuata Salzmann ex Lehmann 96, 101
lasiophylla Martius & Zuccarini 2, 3, 7, 8, 9, stellata Willdenow 25
11, 15, 16, 26, 27, 29, 30, 35, 38, 39, 40, stenaspidota Caspary 3, 59
42, 43, 44, 45, 48, 50, 80, 81, 85, 90-94, tenerinervia Caspary 3, 8, 9, 12, 15, 17, 27,
96, 104, 107, 108 29, 30, 32, 33, 34, 35, 36, 38, 39, 40, 42,
leiboldiana Lehmann 100 43, 48, 51, 59, 65, 76, 77-80, 81, 90, 94,
lingulata Wiersema 7, 8, 9, 15, 16, 26, 27, 29, 100, 104, 107, 108
30, 32, 33, 34, 35, 36, 37, 38, 39, 40, 42, tetragona Georgi 25
43, 44, 48, 50, 93, 94-96, 104, 107, 108 tropaeolifolia Lehmann 101
lotus Linnaeus 4, 15, 23, 24, 25, 26, 28, 36, viol?cea Lehmann 25
37, 39, 40, 43, 44, 108 wittiana Ule 3, 59
maximiliani Lehmann 80, 100 Oenothera Linnaeus 44
mexicana Zuccarini 12, 25, 26, 28, 31, 39, 40, Ondinea Hartog 12, 24
46, 108 Victoria Lindley 37
nocturnea March ex Hooker 101 amaz?nica (Poeppig) Sowerby 37
novogranatensis Wiersema 8, 9, 14, 18, 19, regia Lindley 37
21, 27, 29, 30, 32, 33, 34, 35, 36, 37, 38,
39, 40, 42, 43, 47, 48, 51, 74-77, 78,
104, 107

A Monograph of Nymphaea Subgenus Hydrocallis (Nymphaeaceae)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Monograph of Nymphaea Subgenus Hydrocallis (Nymphaeaceae)

Uploaded by

Copyright:

Available Formats

A Monograph of Nymphaea Subgenus Hydrocallis (Nymphaeaceae)

Author(s): John H. Wiersema

American Society of Plant Taxonomists is collaborating with JSTOR to digitize, preserve

The nearly cosmopolitan genus Nymphaea Linnaeus is composed of aquatic

addition to the lack of sufficient distributional and general morphological infor

The genus Nymphaea was established by Linneaus (1753) but originally in

amazonum, N. blanda, N. gardneriana Planch?n, N. jamesoniana Planch?n, N.

MATERIALS AND METHODS

number of shared compounds present

of detecting similarities among different compounds, such as different glycosides

GENERAL MORPHOLOGY AND LIFE HISTORY

Nymphaea display anatomical differentiation from typical nutritive roots. The

conardii, N. gardneriana, N. novogranatensis, N. glandulifera, N. belophylla, N.

leaves, as both are pure green in N. jamesoniana, N. conardii, and N. glandulif

In N. prolif?ra and N. lasiophylla, abortive tuberous flowers are produced.

Another feature of Nymphaea seeds is the aril, a bellshaped outgrowth of the

Results and Discussion

of other members of subg. Hydrocallis is too scanty to enable an assessment of

transparent layer surrounds the starch-filled perisperm. Perisperm occupies ca.

(Fig. 4d-f ), N. gardneriana (Fig. 4g-i), and N. prolif?ra, is distinguished in hav

k* <sr k\ c>j k\ k> ? k\ *-i

FIG. 7. Hypothetical evolutionary scheme

referable to N. glandulifera from

An examination of available seeds from other subgenera of Nymphaea pro

The pollen of Nymphaea has been described as both monosulcate or monopo

Results and Discussion

The other species of subg. Hydrocallis examined, N. oxyp?tala, appears to be

FLORAL BIOLOGY AND ARTIFICIAL HYBRIDIZATION

A wealth of information has been gathered on the floral or pollination biology

FIG. 13. Night-blooming flowers of Nymphaea subg. Hydrocallis in cultivation, a. Fi

their flowering times. Those of N. amazonum appear to be the most unusu

FIRST-DAY FLOWERS SECOND-DAY FLOWERS

?- JAMESONIANA, CONARDII, GARDNERIANA, GLANDULIFERA

aNumber of crosses. bEstimated percentage of normally developed seeds.

Mostly fragmentary information exists on the flavonoid composition of nym

Results and Discussion

A considerable number and diversity of flavonoid compounds were en

N. amazonum. In particular, they produce flowers with a greater number of

used to support the establishment of subspecific categories. This would seem to be

putative parents based on morphological considerations. This cluster is chemically

found only in N. lasiophylla and a suspected hybrid involving N. lasiophylla, the

Position of subg. Hydrocallis within Nymphaea. The phylogeny developed by

Anecphya Brachyceras Hydrocallis

Carpel fusion incomplete incomplete complete

Flowering time diurnal diurnal nocturnal

Carpellary absent slightly developed, well-developed,

Leaf margins dentate entire to dentate entire (dentate

Pubescence of absent absent absent

Sepal venation inconspicuous inconspicuous inconspicuous

Rhizome erect erect erect

Pollen surface puckered slightly puckered smooth

Doughnutlike absent absent present

Seed hairs present present present

Stolons absent absent often present

Peduncular 6 (5-7) 6 (5-7) 6 (5) or 18 (4

shared by subgenera Lotos and Hydrocallis, the hypothesis of a shared ancestry

designated: Castalia ampia Salisbury [=Nymphaea ampia (Salisbury) de

Perennial, aquatic herbs. Rhizome erect or horizontal, ovoid to cylindric,

Nymphaea subg. Hydrocallis (Planch?n) Conard, Publ. Carnegie Inst. Wash. 4:

Rhizome erect, tuberous, ovoid to cylindric, usually with contractile roots

fused, mottled, or spotted with reddish or purplish pigment, particularly below,

Key to the Species of Nymphaea Subgenus Hydrocallis

9. Plants producing stolons only during germination of tubers; flowers au

1. Nymphaea jamesoniana Planch?n, FI. Serres Jard. Eur. 8: 120. 185