Skip to main content

Full text of "Phytoneuron"

See other formats

wld/domesticated Cucurbitaceae. Phytoneuron 2011-13: 

Guy L. Nesom 

2925 Hartwood Drive 

Fort Worth, Texas 76109 

www . guynes om. com 


A taxonomic review and formal nomenclatural summaries are provided for certain species 
and species complexes of CitruIIus, Cucumis, and Cucurbita that occur in the FNANM region in an 
attempt to apply consistent taxonomic ranking to those including both wild and closely related 
domesticated forms. Taken here as an archetypical starting point are examples in Cucurbita where 
two paired taxa are morphologically distinct but molecularly identical or near-identical, the 
domesticate with some certainty derived from the wild form — these are recognized as conspecific 
subspecies. CitruIIus. Molecular studies show that the lineage including the cultivated watermelon 
(CitruIIus lanatus sensu stricto) and the lineage including the citron melon (C. lanatus var. citroides) 
have a sister relationship but distinct haplotypes, and they apparently diverged from a common 
ancestor between 0.6-0.9 million years ago. The citron melon is appropriately recognized at specific 
rank — as CitruIIus coffer Schrad. — rather than a variety of C. lanatus. At varietal rank, it is 
commonly identified as CitruIIus lanatus var. citroides (L.H. Bailey) Mansf, but the correct name 
instead is CitruIIus lanatus var. caffrorum (Alef.) Fosberg. A lectotype is designated for CitruIIus 
caffer Schrad. and for CitruIIus vulgaris var. citroides L.H. Bailey, and a photograph of each type is 
provided. The sister and probable progenitor of cultivated watermelons (subsp. lanatus) apparently is 
similar to west African wild types identified as C lanatus subsp. mucosospermus. Cucumis. 
Cultivated forms of Cucumis melo (e.g., canteloupe) have long been known to be similar to 
morphologically distinct wild and feral races generally identified as C. melo subsp. agrestis and 
recent studies have confirmed their molecular identity or near-identity. The forms comprising subsp. 
melo are known strictly as cultivars and almost certainly have arisen from subsp. agrestis, thus their 
taxonomic treatment as conspecific subspecies is appropriate, although multiple independent 
domestications from within subsp. agrestis complicate the interpretation. The domesticated Cucumis 
sativus sensu stricto and the free-living C hardwickii are closely related and perhaps justifiably 
interpreted as sister taxa. Cucumis sativus, however, is highly variable in morphology 7 and some feral 
expressions intergrade with C. hardwickii, but the latter is isozymically distinct and its recognition at 
varietal rank seems appropriate, as C sativus var. hardwickii. The relationship apparently is not one 
of wild progenitor/domesticate. Cucurbita. Closely related pairs of Cucurbita taxa (regarded as wild 
progenitor/domesticate) with identical haplotypes and hypothesized to have diverged in response to 
artificial selection are treated as conspecific subspecies — e.g., C. argyrosperma (1866, cultivated) 
includes C sororia (1943, wild) as C argyrosperma subsp. sororia. Cucurbita pepo sensu stricto 
comprises cultivars and landraces of southern Mexico and Guatemala and its wild ancestor has not 
been identified. Its haplotype differs from plants of the eastern USA and northeastern Mexico 
generally treated as C. pepo subsp. ovifera, from which a separate group of domesticates has arisen. 
Subsp. ovifera (nom. superfl. s 11 eg. ; the earlier subsp. texana is correct) is treated here at specific rank, 
as Cucurbita melopepo L., including var. melopepo (the domesticates), var. ozarkasia (Decker- 
Walters) G.L. Nesom, comb, nov., var. texana (Scheele) G.L. Nesom, comb, nov., and var. fraterna 
(L.H. Bailey) G.L. Nesom, comb. nov. A lectotype is designated for C. melopepo L. and for C pepo 
var. condensa L.H. Bailey. Subsp. texana (Scheele) G.L. Nesom, comb, nov., includes var. texana, 
var. ozarkana, and var. fraterna; subsp. melopepo includes var. melopepo. 
KEY WORDS: CitruIIus, Cucumis, Cucurbita, taxonomic rank, typification 

: Nomenclature in domesticated/wild Cucurbitaceae 2 

In preparation of the taxonomic treatment of Cucurbitaceae for the Flora of North America 
North of Mexico, inconsistencies became apparent in concepts of species and infraspecific taxa, 
including cultivars and cultivar groups. Differences in biology and patterns of evolution, as well as 
degrees of subjectivity in judgement, make it difficult to apply a consistent system of taxonomic ranks 
across all genera. Nevertheless, whatever consistency can be attained is useful. 

Emphasizing examples from the Cucurbitaceae, Jeffrey (1968) proposed a system for 
cultivated plants that incorporates a hierarchy of names — species, subspecioid, convar, provar, and 
cultivar. Similarly, Grebenscikov (1953) presented a detailed hierarchical classification for Cucumis 
cultivars, but mostly, even in the Cucurbitaceae, application of horticultural names seems to have 
remained somewhat idiosyncratic from one genus to the next. The current review deals primarily 
with names and ranks that are applicable at levels where both wild plants and horticultural taxa are 
involved in the classification. 

Particularly challenging for classification are instances where a wild progenitor/domesticated 
derivative relationship (e.g., see Crawford 2010) is hypothesized to exist as a result of artificial 
selection. Taken here as an archetypical starting point, an incipient convention, for consistency are 
several examples in Cucurbita (see details below) where two paired taxa are morphologically distinct 
but molecularly identical or near-identical, the domesticate with some certainty derived from the wild 
form. These are recognized as conspecific subspecies. 


Molecular studies have shown that the cultivated watermelon, Citrullus lanatus (Thunb.) 
Matsum. & Nakai sensu stricto, and the citron melon, generally identified as C. lanatus var. citroides 
(Bailey) Mansf, represent closely related but distinct lineages (Navot & Zamir 1987; Jarret & 
Newman 2000; Levi et al. 2001; Dane & Lang 2004; Dane et al. 2004; Dane & Liu 2007; Mujaju et 
al. 2010). The two taxa have distinct haplotypes and appear to have evolved from a common ancestor 
perhaps closely similar to C. ecirrhosus Cogn., with which they share several substitutions and indels 
at non-coding cpDNA regions (Dane & Lang 2004). This evolutionary divergence is estimated to 
have occurred at least 0.8-0.9 million years ago (Dane & Lang 2004) or at least 0.6 million years ago 
Dane & Liu 2007). 

In Citrullus as a whole, which includes 5 species (as interpreted here), two main clades are 
evident: (a) C colocynthis (L.) Schrad. (the bitter apple) and (b) a lineage in which C rehmii De 
Winter is sister to a clade comprising C. ecirrhosus and C. lanatus lato (Dane & Lang 2004), 
the latter regarded here as C. lanatus sensu stricto plus var. citroides (see comments below). An 
earlier IT'S analysis of Citrullus phylogeny, however, placed C. rehmii as sister to C. lanatus sensu 
lato (Jarret & Newman 2000), and Dane et al. (2004) found that C rehmii showed almost the same 
haplotype as C. lanatus var. citroides with the exception of a unique insertion at a cpSSR site. 
Citrullus rehmii (annual) and C. ecirrhosus (perennial) are endemic to the Kalahari Desert region, of 

Citrullus lanatus sensu stricto — domesticated sweet melons and the egusi-type melons. 

Citrullus lanatus sensu stricto includes two morphological types — subsp. lanatus (the 
cultivated, dessert watermelons, including red sweet watermelon; Fig. 5) and subsp. mucosospermus 
Fursa (the egusi-type melons, including wild, semi-cultivated, and cultivated forms in west Africa, 
which are grown primarily for their large, soft, oil- and protein-rich seeds). The egusi melons are 
relatively small and have white, bitter flesh similar to the citron melons, but the egusi/citron 
similarities apparently are plesiomorphic. 

: Nomenclature in domesticated/wild Cucurbitaceae 3 

The cultivated watermelon and the egusi-type melon have the same cpDNA haplotype. Both 
are reported to lack molecular variation at cpDNA regions (Dane & Lang 2004) and also have a 
narrow genetic base in other kinds of assessments (e.g., Levi et al. 2001; Levi & Thomas 2005). 
Mujaju et ai. (2010). however, reported that cultivated sw^eet melons in Zimbabwe are as diverse 
molecularly as the citron melons; Nimmakayala et al. (2010) identified 583 AFLP bands that are 
polymorphic within var. lanatus; Levi et al. (2004) found high polymorphisms among watermelon 
heirloom cultivars using inter-simple sequence repeat (ISSR) and amplified fragment-length 
polymorphism (AFLP) markers. 

Citrullus lanatus var. citroides — citron melons and tsamma melons. 

Fursa (1972) recognized another entity closely similar to var. citroides — C lanatus var. 
coffer (see nomenclature below), tire wild ci tsamma" melon of the Kalahari Desert. cpDNA studies, 
however, have shown that when the citron/tsamma melons are considered together, three haplotypes 
exist among them (Dane & Liu 2007). These three entities apparently have not been unambiguously 
recognized by formal nomenclature, or at least the genetic affinities of names potentially associated 
with these haploptypes are not known. The ancestral citron/tsamma haplotype is known from 
Swaziland and South Africa; each of the other two ranges across southern Africa. A number of 
distinct landraces that are cultivated in Kalahari region (including the tsamma melons) may represent 
early forms of domestication, as also suggested by Maggs-Kolling et al. (2000). 

Two species rather than one. 

The domesticated watermelon and the primarily wild-type egusi melon, which are identical in 
cpDNA haplotype and essentially so in other molecular features, are recognized here as conspecific 
subspecies within Citrullus lanatus. The citron/tsamma melon, divergent both morphologically and 
molecularly from C. lanatus, is recognized as a separate species. 

At specific rank the citron/tsamma melons are correctly identified as Citrullus coffer Schrad. 
ex Eckl. & Zeyh., 1834; if at varietal rank, then as C. lanatus var. cqffrorum (Alef) Fosberg. The 
epithet "caflrorum" at varietal rank was first validly published by Alefeld in 1866 as "Citrullus 
vulgaris var. caffrorum," preceding Bailey's proposal of "Citrullus vulgaris var. citroides" in 1930. 

In the LISA the two species can be distinguished by the following 

1. Leaf blades ovate to lanceolate-ovate or ovate-triangular in outline, mostly 8-20 cm; fruits globose 
to oblong-ellipsoid, 12-35 cm (or more) in diam., rind hard but not durable, flesh juicy, red, yellow, 

or greenish, sw^eet; seeds commonly black Citrullus lanatus 

1. Leaf blades ovate in outline, 3-i0(-14) cm: fruits globose to globose-ovoid, 14-25 cm in diam., 
rind hard and durable, flesh dry, whitish, bitter; seeds tan to brown or reddish Citrullus caffer 

Without mature fruits, the distinction may be difficult. Leaf shape seems generally to separate them, 
but leaves overlap in size and shape. Stems are villous in both species. Habitat is a good clue, since 
C lanatus is almost always found where seeds were recently discarded by people and the plants 
rarely if ever form reproductive populations, Citrullus caffer is free-living and occurs in a wider 
range of habitats. 

Several African accessions morphologically identified as Citrullus differ were found to have 
the cpDNA haplotype of C. lanatus (Dane et al. 2004; Mujaju et al. 2010), possibly as a result of 
introgression, and Levi et al. (2001) found four plants of C. lanatus with genes perhaps introgressed 
from C. caffer (three from Africa, one from China). The citron melon also has been observed to form 
spontaneous hybrids with C. colocynthis (bitter apple) — in the USA (Fulks et al. 1979), in India 
(Singh 1978), and in Australia (Herrington et al. 1986). 

: Nomenclature in domesticated/wild Cucurbitaceae 4 

CITRULLUS LANATUS (Thunb.) Matsum. & Nakai, Cat. Sem. Spor. Hort. Bot. Univ. Imp. Tokyo 
30, no. 854. 1916. Momordica lanata Thunberg, Prodr. PI. Cap. 1: 13. 1794. TYPE: South 
Africa, Cape Province (as cited by Jeffrey 1967) (holotype: UPS-THUNB microfiche!). The 
protologue has only this; "M. [Momordica] foliis ternato-pinnatifidis scabris, fructo lanato." 

a. Subspecies lanatus 

Cucurbita citrullus L., Sp. PI. 2: 1010. 1753. Anguria citrullus (L.) Mill, Gard. Diet. (ed. 8). 1768. 
Cucumis citrullus (L.) Ser. in DC, Prodr. 3: 301. 1828. Citrullus citrullus (L.) H. Karsten, 
Deut. Fl. [Pharm. med. Bot] 889. 1882. Colocynthis citrullus (L.) Kuntze, Rev. Gen. PI. 1: 
256. 1891. TYPE: [Italy.] "Apulia, Calabria, Sicilia" Not designated (fide Jan-is 2008). As 
noted by Jarvis, Jeffrey (1967) designated as lectotype: LINN-1151.5 (digital image!), but 
this was not original material for the name. Recognizing this, Jeffrey later noted (1980, p. 
791) that "This is perhaps better to be regarded as a neotype, since although annotated as this 
species by Linnaeus, it may not have been studied by him prior to the writing of the 'Species 
Plantarum."' Jarvis, however, observed that because original material is in existence, 
designation of LINN-1 151.5 as a neotype would be contrary to ICBN Art. 9. 11. 

Citrullus vulgaris Schrad, Linnaea 38: 413. 1838 (not Enum. PI. Afric. Austral. 2: 279. 1836). 
Citrullus lanatus subsp. vulgaris (ScSirad. ex Eckl. & Zeyh.) Fursa, Bot. Zhurn. (Moscow & 
Leningrad) 57: 37. 1972. Apparently intended by Schrader as a replacement name for 
Cucurbita citrullus L. The full 1836 entry for the species is this: "1790. CITRULLUS vulgaris 
Schrad. Mss. Cucurbita citrullus L., sp. 1435. — In cultorum vicinitate intra coloniam sponte 
quasi proveniens. Ex Europa lllatus. Jan." 

Authorities of names in Citrullus proposed in Ecklon & Zeyhefs Enumeratio Plantarum 
Africae Australis Extratropicae sometimes have been cited as "Schrader ex Ecklon & Zeyher" but the 
author of the whole Cucurbitaceae treatment (pp. 275-280) is explicitly noted (p. 275) to be H.A. 
Schrader. Morphological descriptions were provided for Citrullus (the genus) and Citrullus amarus, 
thus the authority for those is correctly cited as "Schrad. in Eckl. & Zeyh." or simply as "Schrad." 
Citrullus vulgaris and Citrullus coffer were not validly published until 1838 in a posthumous 
manuscript, with the authority from that publication also as "Schrad." or "Schrad. in Schlecht." 
(Linnaea vol. 38, see notes below). 


Citrullus lanatus subsp. mucosospermus Fursa, Bot. Zhurn. (Moscow & Leningrad) 57: 38. 1972. 
Citrullus mucosospermus (Fursa) Fursa, Trudy Prikl. Bot. Genet. Selek. 81:111. 1983. TYPE 
(as cited by Fursa 1972): Ghana. [No specific locality cited]. 5 Aug 1957, N.P. Oltarshevskyi 
3833 (WIR, cat. num. WIR k-3742). 

CITRULLUS CAFFER Schrad, Linnaea 12: 413. 1838. Citrullus vulgaris var. caffrorum Alef, 
Landw. FL, 210. 1866. Citrullus lanatus var. caffer (Schrad.) Mansf, Kuiturpfi., Beiheft 2 
[Verzeichnis], 421. 1959. Citrullus lanatus var.' caffrorum (Alef) Fosberg in Fosberg & 
Sachet, Smithsonian Contr. Bot. 45: 15. 1980. LECTOTYPE (designated here): probably from 
the Gottingen Botanical Garden, herbarium of C.A. Fischer (GOET 007221!, Fig. 1; 
isolectotypes: GOET 007222!, GOET 007223!, Figs. 2 and 3). These three sheets were 
separated and mounted from a folder in the herbarium of C.A. Fischer- (1785-1836), Inspector 
of the Gottingen Botanical Garden ca. 1821-1836 — his herbarium was accessioned by 
GOET in 1837 (J. Heinrichs, GOET Curator, pers. comm. 2011). 

Schrader first published the name Citrullus caffer in a seed catalogue (Ind. Sem. 
Hort. Goetting, 2. 1834), indicating that the name was based on "Cucurbita caffra Eckl." 
from another 1834 seed catalogue (Coll. Sem. Capensium), with the implication that original 

: Nomenclature in domesticated/wild Cucurbitaceae 5 

material should be sought among collections by Ecklon or Ecklon & Zeyher. A still later 
entry by Schrader (in Ecklon & Zeyher's Enum. PI. Afric. Austral. 2: 279. 1836), however, 
cited both names from 1834, implying that the concept of original material would be 
broadened to include specimens seen by Schrader, especially since the name is formally 
credited only to him. 

The entry in Index Seminum Horti Academici Gottingensis, 1834: "Citrullus coffer 
Schrad. Synonym, est: Cucurbita caffra Eckl, coll, sem. capensium. Plura de hac aliisque 
Cucurbitaceis alio loco." This seed catalogue entry was quoted in Linnaea 10 (Litteratur- 
Bericht 1835 & 1836]: 109. 1836. Provided by the Editor, D.F.L. von Schlechtendal). It 
seems that Schrader in 1834 was acknowledging that his use of the epithet was based on the 
slightly earlier seed catalogue publication by Ecklon [& Zeyher] of Cucurbita caffra (see 
below), thus implying that the name in Citrullus would be correctly cited as C. coffer (Eckl.) 
Schrad., 1834. This interpretation also is indicated by the 1836 entry in Enum. PI. Afric. 
Austral. — "ind. sem. hort. Goetting. 1834" was cited as the place of publication of Citrullus 
coffer, and jinniedutels following that, the citation continues with "Cucurbita caffra Eckl. et 
Zeyh. coll. sem. 1834." 

Peter Goldblatt (pers. comm.) notes that he does not know about the "sem. 
capensium" catalogue but that "Ecklon and Zeyher distributed herbarium collections and 
seeds under names they often invented or had in mind to publish. Quite a few of their names 
remain 'in manuscript'." The Goettingen seed catalogue provided neither a description nor 
specimen citation for Citrullus coffer, and it is assumed here that the entry in the 
"Capensium" catalogue was similar. 

In any case, neither of the seed catalogue entries for Citrullus coffer in 1834 nor the 
entry in 1836 provided a description and the name was not validly published until 1838. 
Shortly after Schradefs death in 1836, a brief biography — with various incomplete 
taxonomic manuscripts appended, including one on Cucurbitaceae — was published in 
Linnaea, presumably by the editor Schlechtendal (Reliquiae Schraderianae. Linnaea 38: 353- 
476. 1838; Cucurbitaceae, pp. 401-423). There, a morphological description of C. coffer was 
provided. Reliquiae Schraderianae also was published as a separate in 1838 (see Google 

In the 1836 entry for Citrullus coffer, the provenance was cited as "In cultis 
provinciarum 'Uitenhage et Albany' et in terra 'Kafferland. ' Jan. Incolis: 
'Kafferwatermelon. '" "Caffraria (Cafferland and variants) was the eastern part of 'Cape 
Colony' and is now Eastern Cape Province. Uitenhage was the first town there and usually 
meant in that district. Albany was one of magisterial districts in past time (of Cape of Good 
Hope), when it was a province of South Africa and no longer has any political meaning" 
(Peter Goldblatt, pers. comm., February 2011). 

In treating this entity as "var. caffrorum" within Citrullus vulgaris, Alefeld (1866) 
cited Citrullus caffrorum Schrader and Cucurbita caffra Eckl. & Zeyh. immediately 
following his proposal of the varietal name, which presumably was intended to be homotypic 
with both of the latter names. 

The epithet "caffrorum," however, appeared in Schrader s 1838 entry (Linnaea 12: 
413. 1838) for Citrullus coffer — as "2. Citrullus (C. caffrorum) caffer." — and it is not clear 
why Schrader included the parenthetical element, especially since immediately following in 
the'entry was the citation "Curcurbita caffra Eckl. et Zeyh. Coll. sem. 1833 seq. Verz. 1833." 
It might appear that Schrader was attempting to correct his earlier citation of the Eckl. & 
Zeyh. name from an 1834 seed catalog, with the intention of attributing authorship of the 
basionym to Ecklon & Zeyher, again indicating the correct name to be "Citrullus caffer 
(Eckl. & Zeyh.) Schrad., 1834." In any case, it appears that "caffrorum" as an epithet at 
varietal rank was first validly published by Alefeld in 1866 (as also explicitly interpreted by 
Fosberg, who noted "Var. caffrorum Alefeld here lectotypified by Cucurbita caffra Ecklon & 

: Nomenclature in domesticated/wild Cucurbitaceae 6 

Zeyher seems to be the earliest varietal epithet for the common red-fleshed cultivated 
watermelon."). Fosberg's interpretation of the biological identity this entity was mistaken, as 
these names refer to the citron melon, but at varietal rank within Citrullus lanatus, the correct 
name for the citron melon is C. lanatus var. caffrorum (Alef.) Fosberg. 

Citrullus vulgaris var. ciiroides L.H. Bailey, Hortus (ed. 1), 152. 1930. Citrullus lanatus var. 
citroides (L.H. Bailey) Mansf., Kulturpfl., Beiheft 2 [Verzeichnis], 421. 1959. Citrullus 
colocynthoides var. citroides (L.H. Bailey) Millan, Darwiniana 14: 697. 1969. LECTOTYPE 
(designated here): [USA. Colorado.] "Trade or Label Name: Citron," Dreer 536, cultivated 
at Jtliaca, N.Y., [collected, presumably, by L.H. Bailey] 20 Sep 1930 (BH digital image!; Fig. 
4). This label also is annotated at the top with "photo." Packet attached. The protologue has 
only this: "Var. citroides. CITRON OR PRESERVING MELON. Fr. small, with white hard flesh, 
used only for preserving; seeds not marked or marbled (as they usually are in watermelon)." 

The original publication of var. citroides has been generally cited as "Gentes Herb. 2: 
186. 1930," but in that publication Bailey provided neither a description nor type citation, 
noting (p. 186) that "the var. was included in a brief account of Citrullus for Hortus, and [the] 
type lias now been plated and the page released for printing although the present paper will 
be actually published in advance of that book" 

Several sheets at BH are possibly interpreted as original material of var. citroides. 
The label heading of the 1889 specimen is "Garden Herbarium of Cornell University 
Experiment Station;" the others have "Herbarium of L. H. Bailey." 

1. identified as "Citrullus vulgaris Schrad.," Trade or Label Name: Colorado Citron, University- 
Garden: Sep: 15, 1889. 

3. Trade or Label Name: Citron, Dreer 536, Cultivated at Ithaca, NY., [coll.] Sept 20, 
label also is annotated at the top with "photo. " Packet attached. 

4. Identified as "Citrullus vulgaris Schrad. var 
Preserving Citron, Dreer 536, Sown May 25, 

5. Trade or Label Name: Colorado Reserving Citron, Dreer 536, [coll.] Sept 22, 1938, Ithaca. 

Seeds from which collections 2-5 were grow'n apparently were made as one gathering 
either by Henry A. Dreer (1818-1873), who operated a seedhouse and nursery in 
Philadelphia, or his son, William F. Dreer (1849-1918), who ran a large nursery in Riverton. 
New Jersey, and left an annual endowment to the Cornell Horticulture Department. At least 
one other specimen grown from seeds from H.A Dreer is in the BH herbarium (see 
Cucurbita notes below). It is possible that collection 1 was grown from the same lot, as both 
it and collection 5 are noted to be the "Colorado Citron." In any case, collection 4 is the only 
one identified as "var. citroides," but neither it nor collection 5 was made until after 
publication of the name. Presumably collections 2 and 3 were those that Bailey would have 
regarded as type material, and collection 3 is designated here as the lectotype. With his long- 
standing interest in Cucurbitaceae, Bailey himself probably made the herbarium collections; it 
apparently is his handwriting on the labels. 

Citrullus colocynthoides Pangalo, Bull. Applied Bot, Leningrad 1929-30, 23, 3: 66 [Trudy Prikl. Bot. 
23(3): 66]. 1930 [not "Citrullus vulgaris colocynthoides Schweinf. 1883!," as noted by 
Mansfeld 1959, p. 421). TYPE: not seen, presumably at WIR. The Biodiversity Collections 

: Nomenclature in domesticated/wild Cucurbitaceae 7 

Index indicates that specimens of K.I. Pangalo are in the herbarium of the N.I. Vavilov 
Institute of Plant Industry (WIR). 

In a note below the entry for Citrullus vulgaris var. citroides, Bailey (1930, p. 186), 
observed the following: "I now find that the watermelon has been newly studied in Russia 
and the papers are just at hand after my account is ready (Bulletin No. 3 of Applied Botany 
and New Cultures, Leningrad, 1930). Perhaps the species, Citrullus colocynthoides, there 
proposed by K.I. Pangalo, includes the garden preserving citron here named Citrullus 
vulgaris var. citroides; that new species includes "citron forage watermelons," growing of 
their own accord in the southern part of Soviet Russia something like a weed, nobody 
cultivating them, recently utilized as forage plants. The illustration of the fruits of these 
spontaneous melons shows some of them to be externally like the American garden citron 
melon and others very different from anything I have seen. It will be interesting to determine, 
if possible, whether the preserving citrons originated from Russian or Asian stocks. The 
name colocynthoides is proposed as an independent species of Citrullus, not as a var. of 
Citrullus vulgaris, and therefore it and var. citroides are not in nomenclatorial conflict, even 
though we may find (as I suspect) that the American and Russian plants are equivalent for 
purposes of systematic diagnosis." 

Citron melons have not been commonly documented as naturalized in the USA but they 

appear to be scattered as weeds in pinelands, sandy fields, cotton fields, and orange groves, and along 
roadbanks, railroad banks and tracks, and roadsides and other disturbed sites; known from collections 
and literature from Arizona. California, Florida, New Mexico, Nevada, and Texas (Fulks et al, 1979; 
Grichar et al. 2001; Stephens 2003; Hall et al. 2004). In addition to citron melon, vernacular names 
for Citrullus caffer in English-speaking countries are red-seeded citron, preserving melon, jam melon, 
and stock melon. 

Literature Cited (introductory comments and Citrullus) 

Alefeld, F.G.C. 1866. Landwirthscliaftliche flora oder Die nutzbaren kultivirten Garten und 

Feldgewachse Mitteleuropa's in alien ihren wilden und Kulturvarietaten fur Landwirthe, 

Gartner, Gartenfreunde und Botaniker insbesondere fur landw r irthschaft!iche Lehranstalten. 

Wiegandt & Hempel, Berlin. 
Bailey, L.H. 1930. Three discussions in Cucurbitaceae. Gentes Herb. 2: 175-186. 
Bailey, L.H and E.Z. Bailey. 1930. Hortus: A Concise Dictionary of Gardening, General Horticulture 

and Cultivated Plants in North America. Macmillan, New York. 
Crawford, D. J. 2010. Progenitor-derivative species pairs and plant speciation. Taxon 59: 1413- 

Dane F. and P. Lang. 2004. Sequence variation at cpDNA regions of watermelon and related wild 

species: implications for the evolution of Citrullus hapiotypes. Amer. J. Bot. 91: 1922-1929. 
Dane F., P. Lang, and R. Bakhtiyarova. 2004. Comparative analysis of chloroplast DN A variability 

in wild and cultivated Citrullus species. Theor. Appl. Genet. 108: 958-966. 
Dane F. and J. Liu. 2007. Diversity and origin of cultivated and citron type watermelon {Citrullus 

lanatus). Genet. Resour. Crop Evol. 54: 1255-1265. 
Fosberg, F.R. and M.-H Sachet. 1980. Systematic studies of Micronesian plants. Smithsonian Contr. 

Bot. 45: 1^10. 
Fulks, B.K., J.C. Scheerens, and W.P. Bemis. 1979. Natural hybridization of two Citrullus species, 

J. Hered. 70: 214-215. 
Fursa, T.B. 1972, K sistematike roda Citrullus Sclirad, [On the taxonomy of genus Citrullus Sclirad. J, 

Bot. Zhurn. 57:3L41. 
Grebenscikov, I. 1953. Die entwicklung der melonsystematik. Kulturpflanze 1: 121-138. 

: Nomenclature in domesticated/wild Cucurbitaceae 8 

Grichar W.J., B.A. Besler, and K.D. Brewer. 2001. Citronmelon (Citrullus lanatus var. citroides) 

control in Texas peanut (Arachis hypogaea) using postemergence herbicides. Weed Technol 

15: 81^84. 
Hall, D.W., V.V. Vandiver, and J. A Ferrel. 2004. Citron (Citron Melon). Citrullus lanatus (Thunb.) 

Mats. & Nakai. In Weeds in Florida, SP 37, Florida Cooperative Extension Service, Inst, of 

Food and Agric. Sci., Univ. of Florida. 
Herrington, M.E., P.J Brown, and A.R. Carr. 1986. Introgression as a source of bitterness in 

watermelon. HortScience 21: 1237-1238. 
Jarret, R.L. and M. Newman. 2000. Phylogenetic relationships among species of Citrullus and the 

placement of C. rehmii De Winter as determined by interna] transcribed spacer (ITS) 

sequence heterogeneity. Genet. Resour. Crop Evol. 47: 215-222. 
Jarvis, C. 2008. Order Out of Chaos: Linnean Plant Names and their Types. Linnaean Society of 

London in association with the Natural History Museum, London. 
Jeffrey C. 1967. Cucurbitaceae. Pp. 1-156, in E. Mine-Redhead and RM. Polhill (eds.). Flora of 

Tropical East Africa, Vol. 47. 
Jeffrey, C. 1968. Systematic categories for cultivated plants. Taxon 17: 109-114. 
Jeffrey C. 1980. Further notes on Cucurbitaceae: V. The Cucurbitaceae of the Indian subcontinent. 

Kew Bull. 34: 789-809. 
Levi, A, C.E. Thomas, AP. Keinath, and T.C. Wehner. 2001. Genetic diversity among watermelon 

(Citrullus lanatus and Citrullus colocynthis) accessions. Genet. Resour. Crop Evol. 48: 559 - 

Levi, A, C.E. Thomas, M. Newman, O.U.K. Reddy, and X. Zhang. 2004. ISSR and AFLP markers 

sufficiently differ among American watermelon cultivars with limited genetic diversity. J 

Amer. Soc. Hort. Sci. 129: 553-558. 
Levi, A and C.E. Thomas. 2005. Polymorphisms among chloroplast and mitochondrial genomes of 

Citrullus species and subspecies. Genet. Resour. Crop Evol. 52: 609-617. 
Maggs-Kolling G., S. Madsen, and J.L. Christiansen. 2000. A phenetic analysis of morphological 

variation in Citrullus lanatus in Namibia. Genet. Resour. Crop Evol. 47: 385-393. 
Mujaju, C, J. Sehic, G. Weiiemark, L. Garkava-Gustavsson, M. Fatih, and H. Nybom. 2010. 

Genetic diversity in watermelon (Citrullus lanatus) landraces from Zimbabwe revealed by 

RAPD and SSR markers. Hereditas 147: 142-153. 
Navot N. and D. Zamir. 1987. Isozyme and seed protein phylogeny of the genus Citrullus 

(Cucurbitaceae), Plant Syst. Evol. 156: 61-67. 
Nimmakayala, P., Y.R. Tomason, J. Jeong, S.K. Ponniah, A Karunathilake, A. Levi, R. Perumal, and 

U.K. Reddy, 2010. Genetic reticulation and interrelationships among Citrullus species as 

revealed by joint analysis of shared AFLPs and species-specific SSR alleles. PI. Genet. 

Resour. 8: 16-25. 
Singh, AK. 1978. Cytogenetics of semi-arid plants. III. A natural interspecific hybrid of 

Cucurbitaceae (Citrullus colocynthis Schrad x C vulgaris Schrad). Cytologia 43: 569-576. 
Stephens, J.M. 2003. Citron — Citrullus lanatus (Thunb.) Mansf. var. citroides (Bailey) Mansf. 

Publ. HS585, Ser. Hortic. Sci. Dept, Florida Cooperative Extension Service, Inst, of Food 

and Agric. Sci.. Univ. of Florida, <> 


Only two formal infraspecific taxa within Cucumis melo were recognized by Kirkbride 
(1993), following an earlier informal proposal by Jeffrey (1980a) — subsp. melo and subsp. agrestis. 
Grebenscikov (1953) recognized C. melo and C. agrestis at specific rank and arranged many other 
infraspecific taxa in a hierarchical system of horticultural names, Kirkbride also suggested that other 
variants described within C. melo should be treated with horticultural names, and classification of 
melons into two major lineages has been generally supported by molecular phytogenies (e.g., 
Stepansky et al. 1999; Decker- Walters et al. 2002). Subsp. melo comprises the large-fruited, sweet 

: Nomenclature in domesticated/wild Cucurbitaceae 9 

"dessert" melons of commerce (canteloupe, honeydew, and muskmelon, and the snake melon/ 
cucumber melon) — var. melo (including var. cantalupd), var. inodorus, and var. flexuosus — 
originating mostly in western Asia and Europe, all known strictly as cultivars. Molecular resolution 
between var. cantalupo and var. inodorus is slight (Staub et al. 1997; Silberstein et al. 1999; 
Stepansky et al. 1999), despite significant differences between them in morphology and physiology. 
Subsp. agrestis comprises a group of cultivars and landraces, as well as free-living wild forms and 
feral forms, essentially the generally smaller, mostly non-sweet, and morphologically more variable 
types, which originated mostly in India and southeast Asia. 

The two taxa are distinguished by these contrasts. 

to lanate with spreading hairs; stems glabrous sparsely 
ti cultivated forms commonly fragrant and sweet 

Cucumis melo subsp. melo 

1. Hypanthium and youngest fruits retrorsely or antrorsely sericeous with short appressed hairs; stems 
hispid with retrorse hairs; fruit 2.5-5 cm in diam., mostly not fragrant and not sweet, but with 
numerous exceptions Cucumis melo subsp. agrestis 

Statements that wild forms exist in both subspecies (e.g., Jeffrey 1980b; Zohary & Hopf 
1983; Stepansky et al. 1999; Pitrat et al. 2000) appear to have been based on taxonomic arrangements 
that regard the Himalayan entities Cucumis trigonus Boiss. and C. callosus (Roettl.) Cogn. & Harms 
as synonyms of C melo subsp. melo, Parthasarathy & Sambandam (1980) found that crosses 
between C. melo (presumably subsp. melo) and plants identifiable as C. callosus resulted in abundant 
and fully fertile seeds and FiS with normal meiosis; they concluded that C. callosus (with C trigonus 
as a synonym) does not warrant separate species status and is "nothing but a progenitor of C. melo." 

Jeffrey (1980b) noted that Cucumis callosus is a synonym (a "tropical wild variant") of 
Cucumis melo, while Chakravarthy (1982) treated C. callosus as a distinct species. Verma and Pant 
(1985) treated C trigonus as a synonym of C. callosus, but Matthew (1983) regarded C. trigonus as a 
distinct species of peninsular and western India. Earlier, Clarke (1879) treated C melo var. agrestis 
as a synonym of C. trigonus and noted that C. melo is "perhaps the cultivated form of C. trigonus." 
Nazirnuddin and Naqvi ( 1984) regarded C. callosus and C. trigonus both as synonyms of C. melo var. 
agrestis. Diversity and ambiguity of interpretation are widespread. 

Recent study by Sebastian et al. (2010) indicates that both Cucumis trigonus and C. callosus 
are morphologically distinct species (morphology not discussed in their report) but with DNA 
sequences "nearly identical to those of C. melo." They observed (p. 1472) that C. trigonus and C. 
callosus "likely represent the wild progenitor of cultivated melon." Their cladogram, however, 
appears to show samples of subsp. agrestis and subsp. melo in a sister relationship, more closely 
related to each other than to C trigonus and C. callosus. Perhaps a more accurate summary of their 
portrayal would be that the wild progenitor of cultivated melon (C. melo subsp. melo) is equally as 
likely to be represented by C trigonus and C. callosus as by free-living forms of subsp. agrestis. 

With acceptance of Cucumis trigonus and C. callosus as distinct species, patterns of diversity 
in landraces of subsp. agrestis need to be reexamined in terms of taxonomic recognition and rank. 
Landraces of Cucumis melo in south and eastern Via 'mduJ s ng India) mio a high level of variation 
isozymes, DNA and morphology (Sujatha el al. 1991; Akashi el al. 2002; Dhillon et al. 2007; Tanaka 
et al. 2007; Dwivedi et al. 2010) and some of them might ultimately be recognized at higher rank. A 
best I can infer, if C, trigonus and C, callosus are accepted as distinct species, it appears that no wild 
forms are represented within subsp. melo. Wild forms and apparently primitive expressions exist in 
several lineages of subsp. agrestis. 

Nesom: Nomenclature in domesticated/wild Cucurbitaceae 10 

Taxonomic rank of subsp. agrestis. 

If recognized at specific rank in order to emphasize an evolutionary and taxonomic status 
coordinate with that of Cucumis trigonus and C. callosus, the correct name of subsp. agrestis would 
be Cucumis chidcam L. (see below). It seems likely, however, that cultivars of subsp. melo have been 
derived from one expression or another of subsp. agrestis, and treatment of the two as a conspecific 
subspecies parallels that of wild progenitor/domesticated derivative pairs in Cucurbita. 

On the other hand, the molecular identity (or near-identity) of subsp. agrestis and subsp. melo 
apparently is matched in various other pairs/groups of accepted wild species within. Cucumis 
(Ghebretinsae et al. 2007; Schuman et al. 2007; Renner et al. 2007), thus the biological rationale for 
assigning rank is not exactly analogous to Cucurbita. Further, sterility barriers in Cucumis exist only 
between species groups rather than species — Jeffrey (1980a) formally divided Cucumis into five 
cross -sterile species groups within two subgenera. 

The occurrence of sweet-fruited genotypes at least in vars. agrestis and conomon of subsp. 
agrestis indicates that multiple domestications have occurred in parallel with domestication in subsp. 
melo (Jeffrey 1980a; Stepansky et al. (1999); Pitrat et al. 2000; Sebastian et al. 2010). Stepansky et 
al. (1999) suggested that occasional occurrence of sweet agrestis fruits also may have resulted from 
hybridization between wild and feral genotypes with sweet or vegetable landraces. Parallel trends 
among entities in the two subspecies also exist in vestiture, fruit morphology, duration, and sexuality. 
Cucumis melo var. flexuosus is variable in hypanthmm ^e^titui^ but moLualai data plau. it within 
subsp. melo (Silberstein et al. 1999; Stepansky et al. 1999; Lopez-Sese et al. 2003). Further sampling 
will be necessary to assess the potential pattern of independent domestications. 

Status of Cucumis nwh> var. texanus. 

A landrace of Cucumis melo subsp. agrestis — C melo var. texanus Naud. — is widespread 
and relatively abundant in the southeastern USA (Nesom 2011 and references therein). 
Morphological and molecular data (Decker-Walters et al. 2002) indicate that var. texanus has 
differentiated there in situ but shows the greatest genetic affinities to var. chito and to cultivars from 
Eastern Asia., including var. conomon. Stepansky et al. (1999) found var. conomon to be strongly 
differentiated from other varieties within subsp. agrestis (vars. agrestis, chito. dudaim, and 
momordica). Relationships among the latter four taxa are less strongly resolved on the basis of 
morphological and molecular data. 

Among possible scenarios for the origin of var. texanus in the Western Hemisphere (see 
Decker- Walters et al. 2002) is that its progenitor was brought in by humans intentionally (perhaps by 
Asian immigrants) or unintentionally (as seeds mixed with those of other introduced crops). Given 
the distinctive genetic differentiation of var. texanus, the time of its introduction likely was pre- 
Columbian. Decker-W 7 alters et al. also noted (p. 194) that "The relatively uniform morphological and 
genetic character of wild populations in North America supports a single origin of introduction to that 
continent." Unlike the situation in Cucurbita melopepo (see below) of the USA, there appears to be 
little introgression from cultivars of Cucumis melo into the wild populations of var. texanus. 

Plants identified as Cucumis melo var. dudaim, a modern and independent introduction to 
North America, are encountered outside of cultivation in Arizona and California along waterways and 
irrigation canals, fields, and roadsides. 

Decker- Walters et al. (2002) compared wild North American populations (var. texanus) of 
Cucumis melo with a range of samples of var. dudaim, var. chito, small-fruited Old World 
populations, and a small set of other varieties and recorded detailed measurements of morphological 

: Nomenclature in domesticated/wild Cucurbitaceae 

features of vars. texanus, dudaim, and chito. The key below draws from their observations; 
measurements are given, as an approximate average ± one standard deviation 

1. Plants monoecious; fruit 40 ± 5 mm wide, 49 ± 11 mm long, rind yellow, striped; flesh bitter; seeds 4.6 ± 
4 mm long; hypanthial hairs 0. 8 ± 0. 1 mm long Cucumis melo var. texanus 

1. Plants andromonoecious; fruit 61 ± 9 mm wide, 67 ± 17 mm long, rind yellow or orange to red, striped, 
streaked, or speckled; flesh commonly sweet; seeds 7.2 ± 0.7 mm long; hypanthial hairs 1.7 ± 0.4 mm 

2. Rind with red to orange or brown stripes on a yellow to orange background .. Cucumis melo var. disclaim 
2. Rind usually with stripes, but sometimes streaked or speckled, on a yellow background 

Cucumis melo var. chito 

Decker- Walters et al. also noted that "Overall, plant parts of NA populations [var. texanus] were 
generally smaller than those of Chito and Dudaim cultivars. Where size differences existed between 
Chito and Dudaim, the Dudaim plant parts were usually larger. Interestingly, flowers of NA 
populations are in the upper-normal size range typical for C. melo, whereas Dudaim flowers are 
relatively large for the species. 1 ' And "although ripe fruits of the cultivars always have a sweet 
external aroma, those of NA populations vary from scentless to sweet." Stepansky et al. (1999) found 
variability in sexuality in var. dudaim and var. chito — most were andromonoecious but with some 
accessions of both apparently monoecious. 

Commonly recognized infraspecific taxa/cultivar groups of Cucumis melo. 

Wide extremes of variation and horticultural selections exist within Cucumis melo, especially 
as based on fruit characters (e.g., size, shape, surface features, color, texture, taste, composition). The 
species includes feral, wild, and cultivated forms, including "dessert" melons, as well as non-sweet 
forms that are consumed raw, pickled, or cooked. This has led to a proliferation of names for the 
variants, and various systems of infraspecific classification have been proposed. 

A widely used system proposed by Naudin (1859), dividing Cucumis melo into a single wild 
variety — C. melo var. agrestis, and six cultivated ones, cantalupensis, inodorus, conomon, dudaim, 
flexuosus and momordica — has been variously modified and simplified (e.g., Munger and Robinson 
1991) as well as extended into a detailed hierarchical system (Grebenscikov 1953; see comments by 
Hammer et al. 1986). An excellent overview of infraspecific nomenclature in C. melo was provided 
by Pitrat et al. (2000; largely repeated by Burger et al. 2010 but "adapted to the International Code of 
Nomenclature for Cultivated Plants"), who proposed d snnilat alternative, with 5 cultivar groups in 
subsp. agrestis, 11 cultivar groups in subsp. melo, but molecular data have not supported their 
apportionment of the groups among the two subspecies. Many of these names may be encountered 
for plants potentially found outside of cultivation, and a simplified system more similar to the earlier 
ones is summarized here, followed below by formal nomenclatural summaries. 

3 commonly recognized var. cantalupo. 

a. Cucumis melo var. cantalupo. Asia, Africa; fruits aromatic, large to medium- large, round slightly 
ovoid or depressed-globose, moderately to strongly ribbed, rind smooth or warty, whitish to yellow, 
orange, greenish, or nearly black, flesh orange or green, sweet, dessert melons; usually 
andromonoecious. Examples: Charentais, Prescott Fond Blanc. 

b. Cucumis melo var. reticulatus. Europe, western Asia, North America, South America, Japan; fruits 
aromatic, large to medium-large, round to slightly ovoid, ribbed or unribbed, rind reticulate to finely 
netted, variable in colour, flesh orange or green, sweet, dessert melons; andromonoecious. Pitrat et al. 

: Nomenclature in domesticated/wild Cucurbitaceae 12 

(2000) equaled var. reticulatus with cultivars "mainly developed tins century in the U.S. A" (as in the 
brief description above and as cited in cultivar names following) but the neotype or lectotype 
presumably would be chosen from among heirloom races in cultivation in Europe before 1828 (see 
comments below). In the original description by Seringe, "Melon des Cannes" was noted as an 
example and the currently available Noir des Cannes apparently is similar. Examples of USA 
cultivars: Rocky Ford, Top Mark, Netted Gem, Bender, Hearts of Gold, Hales Best. 

c. Cucumis melo var. inodorus. Asia, Spain; fruits non-aromatic, large, round to ellipsoid, non- 
climacteric and long-storing, rind thick, smooth, wrinkled, or warty, white to yellow or green, flesh 
white, dessert melons; usually andromonoecious. Examples: honeydew, winter melon, muskmelon, 

d. Cucumis melo var. flexuosus. Middle East, Asia, northern Africa; fruits very elongated, rind light 
green to green-striped, ribbed or wrinkled, flesh white, non-sweet, eaten immature as cucumbers or 
pickled; usually monoecious. Examples: snake cucumber, snake melon, cucumber melon, Armenian 

Subspecies agrestis. 

e. Cucumis melo var. agrestis. Africa and Asia as free-living forms; fruits very small (less than 5 
cm), inedible with very thin mesocarp and small seeds; monoecious. 

f. Cucumis melo var. chito. Asia, or reportedly of American feral origin; fruits aromatic, small, plum- 
size, rind yellow, flesh white, used as pickles: monoecious. Combined with var. dudaim by some 
(e.g., Munger & Robinson 1991; Robinson & Decker- Walters 1997) but found to be distinct by Pitrat 
et al. (2000), Decker- Walters (2002), and others. "This type seems to occur mainly in Central 
America as a feral and is not really cultivated. It was probably introduced by African slaves to that 
region" (Pitrat et al. 2000, p. 35). Examples: mango melon, vine peach, glass melon. 

g. Cucumis melo var. conomon. Eastern Asia cultivars; fruits smooth, white-fleshed, with thin rinds 
and often eaten as pickles; andromonoecious. Sometimes said to include var. acidulus Naud., which 
has orange rind. Examples: oriental pickling melon, Freeman's cucumber, Shiro-Uri. 

free-living populations; fruits usually not aromatic, 

i. Cucumis melo var. dudaim. Persia, grown as ornamental (for the aroma), sometimes for edible 
fruits; fruits aromatic, small, red to orange or brown striped; andromonoecious. Examples: dudaim 
melon, Queen Anne's pocket melon, smellmelon, pomegranate melon, plum granny. 

j. Cucumis melo var. chate. Northern Africa, western and central Asia; fruits medium-size, round to 
elongate, ribbed, light to dark green, flesh white to light orange, eaten raw or pickled; monoecious or 
sometimes andromonoecious. Example: Carosello. 

k. Cucumis melo var. momordica. India; fruits not aromatic, large, non-sweet, with thin rind splitting 
at maturity, flesh mealy-white, bland: monoecious. Examples: Indian snapmelon, phoot. 

Formal nomenclature for commonly recognized infraspecific taxa of Cucumis melo. 

Nomenclatural summaries are provided here for a selection of the most generally used names 
relating to the Cucumis melo complex. Details of nomenclature for other names, which number more 
than 500, can be found in Kirkbride's monograph (1993). 

: Nomenclature in domesticated/wild Cucurbitaceae 13 

CUCUMIS MELO L., Sp. PI. 2: 1011. 1753. Protologue: "CUCUMIS foliorum angulis rotundatis, 
pomis torulosis. Hort. cliff. 451. Hort. ups. 292. Mat. med. 444. Roy. Lugdb. 263. Melo 

vulgaris. Bauh. pin. 310. Melo. Bauh. hist. 2. p. 242. Habitat O." LECTOTYPE 

(Meeuse 1962, p. 61; see comments in Kirkbride 1993, p. 81): "HU 4," without data, 
cultivated at Uppsala, Herb. LINN 1152.8 (LINN digital image!). 

The lectotype has no fruit but the expression of Cucumis melo that Linnaeus had in 
mind can be inferred. Bauhin and Cherler's Historia Plantarum Universalis (vol, 2, 1651), 
cited by Linnaeus in the protologue, illustrates a plant (p. 242; shown here in Fig. 6) with 
similar fruits. The accompanying description refers to fruits that are 'torose' (apparently the 
"torulose" = warty condition described by Linnaeus), channeled, reticulate, and green- 
colored. Entries in Hortus Cliffortianus (1737) and van Royen's Florae Leydensis (1740) 
both refer to Robert Morisorfs Plantarum Historiae Universalis (1680, Vol. 2), which 
includes a drawing (tab. 6, fig. 4; shown here in Fig. 7). The illustrated fruits are channeled 
and show what appears to be a shallowly warty rind. They are very similar to fruits of plants 
later described by Seringe as Cucumis melo var. cantalupo (see notes above and below), 
especially modern domesticates such as Charentais and Noir des Carmes. 

a. Subspecies melo 

Cucumis melo vm. flexuosus (L.) Naud, Ann. Sci. Nat, Bot, ser. 4, 11: 63. 1859. Cucumis flexuosus 
L., Sp. PI., ed. 2. 1437. 1763. Melo flexuosus (L.) Pangalo, Bot. Zhurn. (Moscow & 
Leningrad) 35: 577. 1950. Cucumis melo subsp. flexuosus (L.) Grebens., Kulturpfl. 1: 135. 
1953. Cucumis melo convar. flexuosus (L.) Grebens.. Kuiturpfl., Beiheft 2 [Verzeichnis], 
426. 1959. LECTOTYPE (Kirkbride 1993, p. 104): "Habitat in India," without data, Bauhin & 
Cherler, Hist. PI. Univ. 2: 248 [icon]. 1651. 

Cucumis melo var. cantalupo Ser. in DC, Prodr. 3: 300. 1828. Cucumis cantalupo Haberle ex 
Reichenb., Fi, Germ. Excurs., 295. 1831. Cucumis cantalupensis Haberle ex M. Roem, Fam. 
Nat. Syn. Monogr. 2: 69. 1846 ["C. cantalupo Rchb." and "C. melo B cantalupo Ser." cited 
in synonymy]. Cucumis melo var. cantalupensis Naud., Ann. Sci. Nat, Bot., ser. 4, 11: 47. 
1859 [Cucumis cantalupensis Haberle ex M. Roem. cited in synonymy]. Cucumis eumelo 
Pangalo subsp. cantalupo (Ser.) Pangalo, Bot. Zhurn. (Moscow & Leningrad) 35: 576. 1950. 
Cucumis melo convar. cantalupo (Ser.) Grebens., Kulturpfl. 1: 135. 1953, TYPE: Apparently 
not designated. No specimen was cited in the Prodromus entry by Seringe. "Fructu magno 
late costato verrucoso." "Cantaloup orange" (fructo parvo) and three cultivars (including two 
kinds of 'Prescott') with depressed fruits were included among examples. 

None of the nomenclatural proposals listed above cited a specimen, but ail probably 
referred to the same entity and it seems likely that all of them past 1828 had Seringe's 
description in mind. Cucumis cantalupo Haberle ex Reichenb. of 1831 (without an 
attribution to Seringe) preceded C. cantalupensis Haberle ex M. Roem. of 1846, which was 
superfluous. At varietal rank within C. melo, Naudin's proposal of 1859 was superfluous. 

Cucumis melo var. reticulatus Ser. in DC, Prodr. 3: 300. 1828. TYPE: Apparently not designated. 
No specimen was cited in the Prodromus entry by Seringe. "Fructu rotundato vel oblongo, 
cortice reticulata griseo." Fruits were described as large to medium in size, and "M.[elon] des 
Carmes" was included among examples. 

Var. reticulatus is probably correctly considered a formal synonym of Cucumis melo 
sensu stricto (var. melo). 

: Nomenclature in domesticated/wild Cucurbitaceae 14 

Cucumis melo var. inodorus Naud, Ann. Sci. Nat, Bot, ser. 4, 11: 56. 1859. TYPE: Apparently not 
designated; no specimen cited by Naudin (or Jacquin). 

An earlier proposal by Jacquin (Cucumis melo inodorus Jacq., Monogr. melon. 173. 
1832) provided an infraspecific name for a "group" of melons composed of several 
"varieties," thus the epithet was proposed without a rank and the valid varietal name was only 
later established by Naudin. Protologue of Jacquin: "Nous avons era devoir donner a ce 
groupe le nom de melons inodores, parce que toutes les varietes qui le composent sont 
absolument depourvues d'arome; excepte quelques-unes qui, plus anciennement cultivees 
sous de limate de l"Europe, exhalent une legere odoeur, mais qui est ainsi que nous l'avons 
dit, une modification resultant de notre temperature et des precedes que nous employons pour 
les cultiver." A later listing of Cucumis melo var. inodorus by CO. Harz (Landw. Samenk. 2: 
783. 1885) was superfluous if it indeed was intended as a new nomenclatural proposal. 

b. Subspecies agrestis (Naud.) Pangalo 

Cucumis melo var. agrestis Naud., Ann. Sci. Nat. Bot, ser. 4, 11: 73. 1859. Cucumis melo subsp. 
agrestis (Naud.) Pangalo, Cucurbitacees, in Zhukovsky, La Turquie Agricole [Zemledelich. 
Turtsiya], 534. 1933. Melo agrestis (Naud.) Pangalo. Bot. Zhum. (Moscow & Leningrad) 35: 
580. 1950. Cucumis agrestis (Naud.) Grebens., Kulturpfl. 1: 134. 1953. Cucumis melo forma 
agrestis (Naud.) de Wilde & Duyfies, Sandakania 17: 55. 2008. LECTOTYPE (Kirkbride 
1993): cultivated at Paris from seeds from India, C Naudin s.n. (P; isolectotype: P). 

Cucumis melo var. dudaim (L.) Naud., Ann. Sci. Nat. Bot, ser. 4, 1 1: 69. 1859. Cucumis dudaim L., 
Sp. PI. 2: 1011. 1753. Cucumis melo subsp. dudaim (L.) Grebens., Kulturpfl. 1: 134. 1953. 
Melo dudaim (L.) Sageret, Ann. Sci. Nat. 8: 313. 1826. LECTOTYPE (Jeffrey 1980c, p. 20): 
Dillenius, Hort Eltham. 2: 223, t. 177, f 218. 1732. "Semina Marocco primum delata 
dicuntur. ... P. Collinson, a Nobil. Dn. Petremislumfructum & ser 

Cucumis melo var. chate (Hasselq.) Sageret, Mem. Agric. Soc. Roy. Centr. Agric. 58: 488. 1825. 
Cucumis chate Hasselq., Iter Palaest, 491. 1757. Cucumis melo subsp. chate (Hasselq.) 
Hassib, Cucurbit. Egypt 133. 1938. NEOTYPE: (Jeffrey 1980c, fide Kirkbride 1993): Plant 
cultivated at Uppsala. Sweden, presumably from Egypt. Herb. LINN 1152.11, digital 
image!). Kirbride (1993) noted that Jeffrey cited the Species Plantarum number (1152/5) 
rather than the LINN number (1 152. 1 1). The protologue: "Terra praegnans post inundationem 
Nili circa Cairum, nee in ullo alio loco universi ^Egypti colitur, neque aliud sert solum." 

Cucumis pictus Jacq., Hort. Vindob. 3: 17, t. 27. 1776. LECTOTYPE (Kirkbride 1993, p. 115): Hort. 
Vindob. 3: t 27. 1776. Cultivated from seeds from India. Protologue: "Ex India orientali 
semina adlata suerunt Per tres annos in horto perstitit Quarto anno frucus maturos non 
dedit. Ex dissecto fractu guttae gummosae sudant.'" A synonym of Cucumis melo var. 
dudaim, fide Grebenseikov (1959). 

\is melo var. conomon (Thunb.) Makino, Bot. Mag. Tokyo 16: 16. 1902. Cucumis > 
Thunb., Nov. Acta Soc. Sci. Upsal. 3: 208. 1780 [Fl. Jap.. 324. 1784]. Melo , 
(Thunb.) Pangalo, Bot. Zhurn. (Moscow & Leningrad) 35: 580. 1950. Cucumis melo convar. 
conomon (Thunb.) Grebens., Kulturpfl. 1: 135. 1953. Cucumis melo subsp. conomon 
(Thunb.) Grebens., Kulturpfl. 1: 135. 1953. TYPE: Japan. "Rwa, Furi uri, Sjiro uri, Tske uri 
et Tsuke uri, item communissime Konomon. Kaempf. Am. exot. Fasc, V, p. 81 1" (as in Fl. 
Japon., p. 324) (UPS-THUNBERG fiche 22794!). 

: Nomenclature in domesticated/wild Cucurbitaceae 15 

Cucumis pubescens Willi, Sp. PI. 4: 614. 1805. TYPE: Cultivated at Berlin, C.L. Willdenow s.n. 
(holotype: B-W, fiche 18048!). Provenance unknown, as explicitly indicated by Willdenow 
in the protologue; the sheet has "Hort. bot. Berol. W." A synonym of Cucumis melo var. 
agrestis, fide Grebenseikov (1953). 

Cucumis melo var. chito (Morren) Naud, Ann. Sci. Nat, Bot., ser. 4. 11: 67. 1859. Cucumis chito 
Morren, Ann. Soc. Roy. Agric. Gand 5: 351, t. 278. 1849; J. Hort. Sci. Access. 5: 341. 1849. 
LECTOTYPE (Kirkbride 1993, p. 103): Ann. Soc. Roy. Agric. Gand 5: t. 278. 1849. 

Cucumis melo var. acidulus Naud., Ann. Sci. Nat,, Bot., ser. 4, 11: 66. 1859. LECTOTYPE (Kirkbride 
1993, p. 104): Cultivated at the Museum d'Histoire Naturelle, Paris, from seeds sent by Jules 
Lepeire from Pondicheri, India, 1858, C. V. Naudin s.n. (P; isoiectotype: P). 

Cucumis melo var. momordica (Roxb.) Duthie & Fuller, Field & Garden Crops 2: 50, t. 49. 1883. 
Cucumis momordica Roxb., Fl. Ind. 3: 720. 1832. NEOTYPE (Kirkbride 1993, p. 114): 
Roxburgh, Icones Roxburghianae, t. 456. 1964. India; described from cultivation from 
Tanjore country and southern parts of Karnatik. 

Cucumis melo var. texanus Naud.. Ann. Sci. Nat, Bot., ser. 4, 16: 160. 1862. LECTOTYPE (Kirkbride 
1993, p. 113): Cultivated at the Museum d'Histoire Naturelle, Paris, from seeds sent by F.lias 
Durand from Philadelphia, 1861, C V. Naudin s.n. (P; isoiectotype: P). 

The wild progenitor of Cucumis sativus 1 ! 

Features and properties of the free-living Cucumis hardwickii Royle have been reported in 
numerous studies (e.g., Horst & Lower 1978; Puchalski et al. 1978; Kupper & Staub 1988; Bisht et al. 
2004; Choudhary & Singh 2010). It is obviously closely related to C. sativus L, the two — alone in 
the genus — sharing the derived chromosome number of In = 14, and C. hardwickii sometimes has 
been posited as the ancestor of C. sativus (see Sebastian et al. 2010). The geographic range of C. 
hardwickii is the northwestern Himalayas southward into the Eastern and Western Ghats and the 
central Plateau region of India. 

In a series of reciprocal crosses among cucumber, muskmelon, and 19 wild species of 
Cucumis, Robinson and Kowalewski (1978) found that the only species that successfully crossed with 
C. sativus was C. hardwickii — producing fully fertile FiS — and they regarded the two as 
conspecific entities (they suggested the rank of subspecies but did not make a formal nomenclatural 
proposal). Puchalski and Robinson (1990) and others also have treated the tw r o taxa as conspecific 

Among plants identified as Cucumis hardwickii, Bisht et al. (2004) found a high diversity for 
morphological as well as RAPD markers, and gene flow between C. hardwickii and C sativus was 
indicated by the presence of segregating populations of apparent natural hybrids in several regions. 
Morphological variants have been recognized within var. hardwickii (Schuman et al. 1985; Staub & 
Kupper 1986). 

Kirkbride (1993) and Jeffrey (1980b, 2001) observed that Cucumis hardwickii can be 
identified and recognized in its typical form but that morphological intergradation argues against its 
recognition as a distinct evolutionary entity with formal taxonomic recognition. Whitaker and Davis 
(1962) seconded earlier suggestions that C hardwickii is an expression of some feral form of C. 
salmis, rather than an ancestor. A study by de Wilde and Duyfjes (2010) agreed that C. hardwickii is 
not sharply demarcated from feral forms of C sativus and reduced its formal rank within C sativus to 

: Nomenclature in domesticated/wild Cucurbitaceae 16 

On the other hand. Cucumis hardwickii is reported to be isozymically distinct from C. sativus. 
Knerr & Staub (1991) found that C. hardwickii possesses alleles for Per-4 and Idh which were not 
present in the remainder of their samples of C. sativus. Meglic et al. (1996) also found that C. 
hardwickii is isozymically distinct from a large number of C. sativus samples. 

Without consistent morphological differentiation between Cucumis hardwickii and the highly 
variable domesticated and feral forms of C. sativus sensu stricto, but with an apparent molecular 
distinction, their recognition at varietal rank seems appropriate. Rank of forma (as interpreted here) 
would imply that "hardwickii" is a populational variant. At least the relationship is not clearly one of 

CUCUMIS SATIVUS L., Sp. PI. 2: 1012. 1753. LECTOTYPE (Ten Pas et al. 1985, p. 290): Herb. 
Burser vol. 17, no. 97 (UPS). From cultivation (as cited by de Wilde & Duyfjes 2007). 

a. Cucumis sativus var. sativus 

b. Cucumis sativus var. hardwickii (Royle) Gabaev, Cucumbers, 47, 1932. Cucumis hardwickii 

Royle, 111. Bot. Himal. Mts. 1: 220; 2: t. 47. 1835. Cucumis sativus unranked hardwickii 
(Royle) Alef, Landw. Fl., 196. 1866. Cucumis sativus forma hardwickii (Royle) W.J. de 
Wilde & Duyfjes, Sandakania 17: 58. 2008. TYPE (as cited by de Wilde & Duyfjes 2007): 
India, J.F. Royle s.n. (LIV; isotype: K). 

Another distinctive, formally named expression of Cucumis sativus has recently been 
recognized in the Bhutan flora (Grierson 1991): C. sativus var. sikkimensis Hook, f Grierson 
distinguished it from the common cucumber by the following contrasts. 

1. Leaves 3-5-lobed; ovary with 3 placentae; fruit greenish var. sativus 

1. Leaves 7-9-lobed; ovary with 5 placentae; fruit mottled yeliow r and brown var, sikkimensis 

Kirkbride (1993) placed var. sikkimensis in the synonymy of Cucumis sativus, but I have not 
encountered the name among samples evaluated in taxonomic studies of C. sativus. Presumably var. 
sikkimensis is derived from within the C. sativus lineage but this appears to be unconfirmed. 

Comments by Hooker in the protologue are at least vaguely suggestive that he thought var. 
sikkimensis might be derived from Cucumis hardwickii. "The origin of the common Cucumber, 
which is supposed to be unknown, is in all probability the C. Hardwickii, Royle, of the Himalaya 
Mountains, which inhabits the sub-tropical region of the range from Kumaon to Sikkim, This opinion, 
founded on specimens gathered by myself in the latter country, is also adopted by M. Naudin, upon 
the same materials (.Am. Sc. Nat., I.e., p, 30). The flowers and leaves of the two plants are almost 
identical, but the fruit of C. Hardwickii is small, smooth, and very bitter; it is, however, striped with 
white and green, a very usual character with the Sikkim cultivated Cucumbers. ... [The Sikkim 
cucumber] is grow r n in all parts of the Sikkim and in the Nepal Himalaya, up to 5000 feet elevation, in 
prodigious quantities," 

c. Cucumis sativus var. sikkimensis Hook, f, Bot. Mag. 102 (ser. 3, vol. 32): t. 6206. 1876. TYPE: 
Sikkim. Cultivated in the Tropical Economic House, Royal Botanic Gardens, Kew, s.n., 
from seed provided by Major Trevor Clarke, Aug 1875 (holotype: K; isotype: K). The 
mature fruit is beautifully illustrated by Walter Hood Fitch in the protologue. Nepalese 
cucumber, Sikkim cucumber, brown-netted cucumber, red-netted cucumber, Khiva cucumber 

: Nomenclature in domesticated/wild Cucurbitaceae 17 

Literature Cited (Cucumis) 

Akashi, Y., N. Fukuda, T. Wako, M. Masuda, and K. Kato. 2002. Genetic variation and phylogenetic 

relationships in East and South Asian melons, Cucumis melo L., based on the analysis of five 

isozymes. Euphytica 125: 385-396. 
Bisht, I.S., K.V. Bhat, S.P.S. Tanwar, D.C. Bhandari, K. Joshi, and A.K. Sharma. 2004. Distribution 

and genetic diversity of Cucumis sativus var. hardwickii (Royle) Alef in India. J. Hort. Sci. 

Biotech. 79: 783-791. 
Burger, Y., H.S. Paris, R. Cohen, N. Katzir, Y Tadmor, E. Lewinsohn, and AA Schaffer. 2010. 

Genetic diversity of Cucumis melo. Hotticuit. Rev. 36: 165-198. 
Chakravarty H.L. 1982. Cucurbitaceae. Fascicles of the Flora of India 11: 1-136. 
Choudhary, H. and D.K. Singh. 2010. Breeding potential of Indian germplasm of Cucumis sativus 

var. hardwickii for cucumber improvement. Acta Hort. (ISHS) 871: 409-416. 
Chung, S.M., J.E. Staub, and J.F. Chen. 2006. Molecular phylogeny of Cucumis species as revealed 

by consensus chloroplast SSR marker length and sequence variation. Genome 49: 219-229. 
Clarke, C.B. 1879. Cucurbitaceae. Pp. 604-635, in ID. Hooker. 

Sabiaceae to Cornaceae. 
Cogniaux C.A 1881. Cucurbitaceae. Pp. 325-951, in de Candolle A. and C. de Candolle, 

Monographiae Phanerogamarum Prodromi 3. Masson et cie, Paris. 
Decker-Walters, D.S., S.-M. Chung, J.E. Staub, H.D. Quemada, and A.I. Lopez-Sese. 2002. The 

origin and genetic affinities of wild populations of melon {Cucumis melo, Cucurbitaceae) in 

North America. PI. Syst. Evol. 233: 183-197. 
Dhillon, N.P.S., R Ranjana, K. Singh, I. Eduardo, AJ. Monforte, M, Pitrat, N.K. Dhillon and P.P. 

Singh. 2007. Diversity among landraces of Indian snapmelon (Cucumis melo var. 

momordica). Genet. Resour. Crop Evol. 54: 1267-1283. 
Dwivedi, N.K., O.P. Dhariwal, S.G. Krishnan, and D.C. Bhandari. 2010. Distribution and extent of 

diversity in Cucumis species in the Aravalli ranges of India. Genet. Resour. Crop Evol. 57: 

Ghebretinsae, A.G., M. Thulin, and J.C, Barber. 2007. Relationships of cucumbers and melons 

unraveled: molecular phylogenetics of Cucumis and related genera (Benincaseae, 

Cucurbitaceae). Amer. J. Bot. 94: 1256-1266. 
Grebenscikov, I. 1953. Die entwicklung der melonsystematik. Kulturpflanze 1: 121-138. 
Grebenscikov, I. 1959. Cucumis L. Pp. 423^126, in R Mansfeld. Vorlaufiges Verzeichnis 

landwirtschaftlich oder gartnerisch kultivierter Pflanzenarten. Kulturpflanze, Beih. 2. 

Akademie-Verlag, Berlin. 
Grierson, AJ.C. 1991. Cucurbitaceae. Pp. 246-270, in Flora of Bhutan, Vol. 2, part 1. Royal 

Botanic Garden, Edinburgh. 
Hammer, K., P. Hanelt, and P. Perrino. 1986. Carosello and the taxonomy of Cucumis melo L. 

especially of its vegetable races. Kultapflanze 34: 249-259. 
Horst, E.K. and R.L. Lower. 1978. Cucumis hardwickii: A source of germplasm for the cucumber 

breeder. Cucurbit Genet. Coop. Rep. 1: 5. 
Jeffrey, C. 1980a. A review of the Cucurbitaceae. Bot. J. Linn. Soc. 81: 233-247. 
Jeffrey, C. 1980b. Further notes on Cucurbitaceae V. The Cucurbitaceae of the Indian subcontinent. 

Kew Bull 34: 789-809. 
Jeffrey, C. 1980c. The Cucurbitaceae of Eastern Asia. Royal Botanical Gardens, Kew. 
Jeffrey, C. 2001. Cucurbitaceae. In P. Hanelt (ed.), Mansfeld's Encyclopedia of Agricultural and 

horticultural crops 3: 1510-1557. Springer -Ver lag, Berlin, Heidelberg. 
Kirkbride, J.H. 1993. Biosystematic monograph of the genus Cucumis (Cucurbitaceae). Parkway 

Publishers, Boone, North Carolina. 
Knerr, L.D. and J.E. Staub. 1991. A multivariate re-evaluation of biochemical genetic diversity in 

Cucumis sativus L. Cucurbit Genet. Coop. Rep. 14: 25-28. 

: Nomenclature in domesticated/wild Cucurbitaceae 18 

Kupper, R.S. and J.E. Staub. 1988. Combining ability between lines of Cucumis sativus L. and 

Cucumis sativus var. hardwickii (R.) Met Euphytica 38: 197-120. 
Lopez-Sese, A.I., J.E. Staub, and M.L. Gomez-Guillamon. 2003. Genetic analysis of Spanish melon 

{Cucumis melo L.) germplasm usijig a standardized molecular-marker array and 

geographically diverse reference accessions. Theor. Appl. Genet. 108:41-52. 
Matthew, K.M. 1983. The Flora of the Tamilnadu Carnatic. Vol. 3. Diocesan Press, Madras. 
Meeuse, ADJ. 1962. The Cucurbitaceae of southern Africa. Bothalia 8: 1-111. 
Meglic, V., F. Serquen, and J.E. Staub. 1996. Genetic diversity in cucumber (Cucumis sativus L.): I. 

A reevaluation of Hie U.S. germplasm collection. Genetic Resour. Crop. Evol. 43: 533-546. 
Monforte, A J., J. Garcia-Mas, and P. Arus. 2008. Genetic variability in melon based on microsatellite 

variation. PI. Breeding 122: 153-157. 
Munger, H.M. and R.W. Robinson. 1991. Nomenclature of Cucumis melo L. Cucurbit Genet. Coop. 

Reports 14: 43^14. 
Nakata, E., J.E. Staub, A.I. Lopez-Sese, and N. Katzir. 2005. Genetic diversity of Japanese melon 

cultivars (Cucumis melo L.) as assessed by random amplified polymorphic DNA and simple 

sequence repeat markers. Genet. Resour. Crop Evol. 52: 405^119. 
Naudin, C.V. 1859. Essais d'une monographie des especes et des varietes du genre Cucumis. Ann. 

Sci. Nat. Bot. ser. 4, 11: 5-87. 
Nesom, G.L. 2011, New state records for Citrullus, Cucumis, and Cucurbita (Cucurbitaceae) outside 

of cultivation in the USA Phytoneuron 2011-1: 1-7. 
Parthasarathy, V. A and C.N. Sambandam. 1980. Taxonomy of Cucumis callosus (Rottl.) Cogn. — 

the wild melon of India. Cucurbit Genet. Coop. Rep. 3: 66-67. 
Pitrat M., P. Hanelt, K. Hammer, N, Katzir, and H.S. Paris. 2000. Some comments on infraspecific 

classification of cultivars of melon. Proc. Cucurbitaceae 2000. Acta Hortic. 510: 29-36. 
Puchalski, J.T., R.W. Robinson, and J.W. Shall. 1978. Comparative electrophoresis of isoenzymes 

of Cucumis species. Cucurbit Genet. Coop. Rep. 1: 39. 
Puchalski, J.T. and R.W. Robinson. 1990. Electrophoretic analysis of isozymes in Cucurbita and 

Cucumis and its application for phylogenetic studies. Pp. 60-76, in D.M. Bates et al. (eds.), 

Biology and Utilization of the Cucurbitaceae. Comstock Publishing Associates, Ithaca and 

Renner S.S., H. Schaefer, and A. Kocyan. 2007. Phylogenetics of Cucumis (Cucurbitaceae): 

Cucumber (C sativus) belongs in an Asian/Australian ciade far from melon (C melo), BMC 

Evol. Biol. 7: 58. 
Robinson, R.W. and D.S. Decker- Walters. 1997. Cucurbits. Crop Production Science in Horticulture 

6. CAB International, New York. 
Robinson, R.W. and E. Kowalewski. 1978. Interspecific hybridization of Cucumis. Cucurbit Genet. 

Coop. Rep. 1:40. 
Schaefer, H. 2007. Cucumis (Cucurbitaceae) must include Cucumella, Dicoelospermum, Mukia, 

Myrmecosicyos, and Oreosyce: A recircumscription based on nuclear and plastid DNA data. 

Blumea 52: 165-177. 
Schuman, D. A, J.E. Staub, and B.E. Struckineyer. 1985. Morphological and anatomical comparisons 

between two Cucumis sativus botanical varieties: hardwickii and sativus. Cucurbit Genet. 

Coop. Rep. 8: 15-18. 
Sebastian, P., H Schaefer, I.R H. Telford, and S.S. Renner. 2010. Cucumber (Cucumis sativus and 

melon (C melo) have numerous wild relatives in Asia and Australia, and the sister species of 

melon is from Australia. Proc. National Acad. Sci. U.S.A. 107: 14269-14273. 
Silberstein, L., I. Kovalski, R Huang, K. Anagnostou, M.M.K. Jahn, and R. Perl-Treves. 1999. 

Molecular variation in melon (Cucumis melo L.) as revealed by RFLP and RAPD markers. 

Scientia Hortic. 79: 101-111. 
Staub, J.E. and RS. Kupper. 1985. Use of Cucumis sativus var. hardwickii germplasm in backcrosses 

with Cucumis sativus var. sativus. HortSci. 20: 436-438. 

: Nomenclature in domesticated/wild Cucurbitaceae 19 

Stepansky, A, I. Kovalski, and R. Perl-Treves. 1999. Intraspecific classification of melons (Cucumis 

melo L.) in view of their phenotypie and molecular variation. PL Syst. Evol. 217: 313-332. 
Sujatha, V.S., V.S. Seshadri, K.N. Srivastava, and T. A More. 1991. Isozyme variation in muskmelon 

(Cucumis melo L.). Indian J. Genet. PI. Breed. 51: 438^144. 
Tanaka, K., A Nishitani, Y. Akashi, Y. Sakata, H. Nishida, H. Yoshino, and K. Kato. 2007. 

Molecular characterization of South and East Asian melon, Cucumis melo L., and the origin 

of Group Conomon var. makuwa and var, common revealed by RAPD analysis. Euphytica 

Ten Pas, H.N., J.W..P. Schoenaker, E.H. Oost, and C.E. Jarvis. 1985. Re-lectotypification of Cucumis 

sativus L. Taxon 34: 288-293. 
Verma, D.M. andP.C. Pant. 1985. Cucurbitaceae. Pp. 157-160, in Flora of India. Series 3. Flora of 

Raipur, Durg, and Rajnandgaon. Botanical Survey of India. 
Whitaker, T.W. and G.N. Davis. 1962. Cucurbits. Interscience Publishers Inc., New York, 
de Wilde, W.JJ.O. and B.E.E, Duyfjes. 2007. The wild species of Cucumis L. (Cucurbitaceae) in 

South-East Asia. Adansonia, ser. 3, 29: 239-248. 
de Wilde, W.JJ.O. and B.E.E. Duyfjes. 2010, Cucumis sativus L. forma hardwickii (Royle) W.J. de 

Wilde & Duyfjes and feral forma sativus. Thai Forest Bull. (Bot.) 38: 98-107. 
Zohary, D. and M. Hopf. 1993. Domestication of plants in the Old World — The origin and spread 

of cultivated plants in West Asia, Europe, and the Nile Valley, Clarendon Press, Oxford. 


In two instances in Cucurbita, progenitor-domesticate pail's have been treated as conspecific 
subspecies (Whitaker & Bemis 1975; Nee 1990; Merrick 1995; Sanjur et al. 2002). Cucurbita 
andreana Naudin, a wild species native to Argentina and Bolivia is identical in haplotype (no base 
pair differences) to the domesticated C maxima and is recognized as C maxima Duchesne subsp. 
andreana (Naudin) Filov. Similarly, the wild gourd C. sororia Bailey of Mexico and Central 
America is identical in haplotype to the domesticated C. argyrosperma Huber and is treated as C. 
argyrosperma subsp, sororia (Baile) - ) Merrick &. Bates. Ecological, morphological, and isozymic 
evidence and crossing studies also support the close relationships of these paired taxa. 

The domesticated species Cucurbita ficifolia Bouche and C. moschata Duchesne have unique 
haplotypes and no wild progenitor has been proposed for either. Mtochondi'ial DNA data combined 
with other information suggest that the wild ancestor of C. moschata will be found in lowland 
northern South America (Sanjur et al. 2002); Nee (1990) noted that the ancestor, from reports from 
Bolivia and Colombia, may be extant but undescribed. Cucurbita ficifolia has a preference for cool, 
high-elevation habitats and probably originated in South America (Nee 1990; Sanjur et al. 2002), 
from where the only reliable archaeological records of the species have come. 

The situation involving Cucurbita pepo L. sensu lato is more complex. Archaeological, 
morphological, and molecular-genetic research indicate that more than a single lineage exists in the C. 
pepo complex. Subsp. pepo is distinct from the rest, which usually has been identified either as 
subsp. ovifera (e.g., Decker and Wilson 1986, 1987; Decker 1988; Decker- Walters et al. 1993) or as 
subsp. ovifera plus subsp. fraterna (e.g., Andres 1987; Lira-Saade et al. 1995). Isozyme data indicate 
that within the range of C pepo sensu lato, "genetic divergence took place long before domestication 
and over an extensive period of time in at least four disjunct and ecologically distinct regions" 
(Decker- Walters et al. 1993). 

1. Subsp. pepo, a Mexican lineage of domesticates, includes jack-o-lantern pumpkins, zucchini, 

marrow, cocozelle, and Mexican landraces. It differs from subsp. ovifera sensu lato by a derived 
molecular feature (a difference in three adjacent base pairs) that also occurs in Cucurbita moschata 
and the C sororiaiC. argyrosperma group (Sanjur et al. 2002). The wild ancestor of subsp. pepo, 

: Nomenclature in domesticated/wild Cucurbitaceae 20 

which presumably shared this derived feature, has not been identified and may be extinct. Forms of 
subsp. pepo presumably were selected from ancestral populations in southern Mexico. 

Data from mtDNA and RAPD studies (Sanjur et al. 2002; Decker- Walters et al. 2002) and 
earlier isozymic and cpDNA studies (e.g., Decker-Walters et al. 1993) also indicate that Cucurbtta 
pepo subsp. pepo is distinct from the entities of subsp. ovifera sensu lato. The clade comprising 

subsp, pepo is basal to that of subsp. ovifera (Sanjur et al. 2002), 

A landrace of Cucurbita pepo recently recognized by Teppner (2000, 2004) as C. pepo subsp. 
gumala Teppner, comprises a series of domesticates apparently authochthonous in Guatemala and 
adjacent southern Mexico, and they may be similar to the ancestral wild form. These have relatively 
small fruits (13-20 cm in diameter, depressed-globose) and extremely thick rind, ripening orange- 
yellow, and orange flesh. Teppner observed that the fruits of subsp. gumala are similar to ancient 
ones of C. pepo from Guild Naquitz cave in Oaxaca (Smith 1997). 

2. Subsp. ovifera sensu lato includes three, geographically separate wild forms, all of which have 
identical mitochondrial DNA sequences (Sanjur et al. 2002) as well as similarities in isozymes and 
other kinds of DNA (see comments following): var. fraterna of northeastern Mexico, var. texana, 
apparently confined to Texas, and the more recently described var. ozarkana of the east-central USA. 
No domesticates derived from var. texana have been identified, but various domesticates, mostly 
identified as var. ovifera, have been derived from within var. ozarkana — the scallop (pattypan), 
acorn, crookneck, and straightneck squashes, and a number of ornamental gourds (e.g., Crown of 
Thorns, Flat-Striped Striped Pear, Bicolor Spoon, White Egg, Nest Egg, Mniature Ball, and most of 
those with deep orange color, Mandan, Orange Warted, Warty Hardhead). The ancestor of most of 
the domesticates is hypothesized to be var. ozarkana, because var. texana has a isozyme pattern not 
found in the other varieties (Decker- Walters et al. 1993; Decker- Walters et al. 2002). 

The cultivars Orange Ball, Miniature Bail, and Orange Warted showed a close allozymic 
relationship with Cucurbita pepo sensu stricto in several earlier studies (e.g., Decker & Wilson 1987; 
Decker 1988), but the RAPD study of Decker- Walters et al, (2002) clusters only Orange Ball within 
C. pepo sensu stricto. Miniature Ball showed a high affinity for var. fraterna in morphological and 
allozymic analyses of Andres (1987) but the origin of Mniature Ball in the 2002 RAPD study is 
indicated to be from var. ozarkana. Similarly, White Egg clustered with var. fraterna in allozyme 
studies (Decker- Walters et al. 1993) but later with var. ozarkana (Decker- Walters et al. 2002). 

Var. fraterna is endemic to Tamaulipas and Nuevo Leon, Mexico, where it grows in 
seasonably dry, upland thornscrub habitat. It also is reported to be common as a weed in agricultural 
fields (Lira et al. 2009). It was first described at specific rank (Bailey 1943) and subsequently 
included within C. pepo at either varietal or subspecific rank (see below). Var. fraterna appears to be 
discrete both morphologically and geographically and it has a coordinate (sister) position relative to 
the two USA entities of subsp, ovifera in the cluster diagrams of Decker- Walters et al, (1993) and 
Decker- Walters et al. (2002). 

Isozyme and RAPD data indicate that var. texana sensu stricto is limited to south-central 
Texas (Decker- Walters et al. 1993, 2002), mostly in drainage systems of the Brazos, Colorado, 
Guadalupe, Nueces, San Antonio, and Trinity rivers (Cowan & Smith 1993). Its habitats — creek and 
river banks, lake shores, marsh banks, low woods, dunes, and disturbed sandy sites — are distinct 
from the drier ones of var. fraterna. Habitats recorded for var. ozarkana are generally similar to those 
of var. texana, except peilups tending to he moie ruderal — creek and river banks, gravel bars, 
bottomland forests, soybean, corn, and cotton fields, old fields, fencerows. railroad ROWs, roadsides, 
and disturbed sites. Var. ozarkana occurs in the central Mississippi Valley and the Ozark Plateau. 

: Nomenclature in domesticated/wild Cucurbitaceae 21 

Populations in Alabama were included in var. ozarkana by Cowan and Smith (1993) but their 
evolutionary status was considered "uncertain" by Decker- Walters et al. (1993); the Alabama plants 
do not share the specialized isozyme pattern of var. texana but otherwise do not closely cluster with 
samples of var. ozarkana. 

Wild populations of subsp. ovifera (not further identified to variety) were distributed in the 
early Pleistocene as far southeast as Florida (Newsom et al. 1993), Some weedy-habitat populations 
of var. ovifera extant in Illinois and Kentucky may have evolved as cultivar escapes (e.g., Wilson 
1990), and some of these may also have experienced subsequent introgression with other nearby 
cultivated, weedy, or wild populations of var, ozarkana (Kirkpatrick & Wilson 1988; Decker- Walters 
et al. 2002 and included references). 

Molecular differentiation between var. texana and var. ozarkana suggests that they have been 
reproductively isolated for a long period, but Teppner (2004) considered morphological variability as 
too overlapping to allow their unarbitrary separation. Molecular (fata reliably separate them and 
clearly delimit var. fraterna as well, but morphology needs to restudied in detail. Fruits of var. 
ozarkana usually are ivory-white at maturity vs. green-and-white striped in var. texana. but this is not 
completely consistent (Andres 1995; pers, observ.). Fruits of var. fraterna are green-and-white 
striped, turning yellow-orange at maturity. 

Entities generally treated as Cucurbita pepo sensu lato are here apportioned between two 
species : C. pepo L. (known only from domesticated forms) and C. meiopepo L. (known from wild 
forms and domesticates), emphasing their geographic and molecular differences. Cucurbita 
meiopepo (1753) predates C. ovifera L. (1766) when that entity is treated at specific rank and is 
regarded here to include var. fraterna, var. texana, var. ozarkana (the wild forms) and var. meiopepo 
(the domesticated forms). The wild forms of C. meiopepo are geographically and molecularly 
distinct, but they cluster closely among themselves and form an evolutionary lineage separate from C. 
[jepo. It seems likely that the three wild taxa of C. meiopepo represent geographic segregates 
(vicariants) of a widespread ancestor, and perhaps C. pepo sensu stricto evidences an even earlier 
vicariant divergence from an ancestor of the entire complex. 

lin Cucurbita and other genera where wild progenitor/ domesticate pairs 
are identified at subspecific rank, the var. meiopepo domesticates are recognized as subsp. meiopepo, 
coordinate with the three wild varieties, which are treated as C meiopepo subsp. texana. .4s currently 
understood, the domesticates are derived only from var. ozarkana, but it is not known whether they 
had a single common ancestor (populationally) or whether some of them have been independently 
derived. The possibility also remains that several of the domesticates are derived from var. fraterna. 

The wild ancestor of Cucurbita pepo has not been identified but perhaps will prove to be 

subsp. gumala or some other uninvestigated landrace of southern Mexico, 

CUCURBITA PEPO L., Sp. PI. 2: 1010. 1753. Cucumis pepo (L.) Dumort, Fl. Belg., 54. 1827. 
LECTOTYPE (Keraudren-Aymonin in Aubreville & Leroy (eds.), Fl. Cambodge Laos Viet- 
Nam 15: 105. 1975): Locality not indicated. Herb. Linn. No. 1151.4 (LINN). 

"Cucurbita pepo, with Herb. Burser XVII: 103 (UPS) as type, was proposed as 
conserved type of the genus by Jaivis (in Taxon 41: 562, 1992), However, the proposal was 
eventually ruled unnecessary by the General Committee (see Barrie, I.e. 55: 795-796. 2006 
for a review of the history of this and related proposals), [quotation continues] 

Bailey (in Gentes Herb. 2: 79. 1929) reproduced the protologue and various of the 
original elements (including a Dalechamp figure and a LINN sheet) and provided an 

: Nomenclature in domesticated/wild Cucurbitaceae 22 

extensive discussion. Keraudren-Aymonin's (1975) type choice of 1151.4 (LINN) has 
priority over that of the Burser material designated by Jeffrey (in Jarvis, I.e. 41: 562. 1992) 
and becomes the type with the failure of the conservation proposal'' (two paragraphs above 
quoted from the online database of The Linnaean Plant Name Typification Project — 
accessed January 201 1). 

A. Subspecies pepo 

Cucurbita aurantia Willd., Sp. PI. 4: 607. 1805. Cucurbita pepo var. aurantia (Willd.) C. Harz, 
Landw. Samenk. 2: 818. 1885. TYPE: Not designated. Locality not indicated in the 
protologue: "Fructus magnitudine facie et colore simillimus Citro Aurantio. W." Type 
material: B-Willdenow fiche 18031; the sheet says "Bouche W." Similar to the Orange Ball 

Cucurbita pepo var. medullosa Alef, Landw, Fl., 218. 1866, TYPE: Not designated. Protologue: 
"Frucht kleiner und etwas weicher als bei der Vorigen, oboid-oblong, regelmassig, dadurch 
der folgenden Gruppe ahnlich. — In England un Nordamerika sehr beliebt, wie Gurken 
zubereitet, also vor der Reife." Generally cited at varietal rank, but Aiefeld described it 
without rank within his "Var.-Gr. 1" and it is perhaps better interpreted at "subvar." rank. 
Explicitly identified by Aiefeld as "Vegetable marrow." 

B. Subspecies gumala 

Cucurbita pepo subsp. gumala Teppner, Phyton (Horn) 40: 34, figs. 3, 24, 25. 2000. TYPE: 
Guatemala. Cultivated in the Botanical Garden of the Institutes for Botany, Universitat Graz. 
Austria, 8 Aug 1989, H. Teppner s.n., from seeds from Guatemala sent by Scheidt (Gieifen, 
Germany) in Sep 1988 (holotype: GZU; isotype: GZU). 

CUCURBITA MELOPEPO L., Sp. PI. 2: 1010. 1753. Cucurbita polymorpha var. melopepo (L.) 
Duchesne ex Lam., Encycl. 2: 157. 1786. Cucurbita pepo var. melopepo (L.) Alef., Landw. 
Fl., 220. 1866. LECTOTYPE (designated here): The illustration in Bauhirfs Historia (J. 
Bauhin and J.H. Cherler, Historia plantarum universalis 2: 224, 1651), reproduced here as 
Fig. 8. Locality not indicated in the protologue: "Cucurbita foliis lobatis, caule erecto, pornis 
depresso-nodosis. Melopepo clypeiformis. Bauh. pin. 312. Cucurbita clypeiformis f 

Siciliana. Bauh. hist. 2. p. 224. Habitat O. Cirrhi sunt in caule, quamvis non scandat 


Comments of Bailey (1929, p. 84): "This plant according to Linnaeus, is erect, not 
climbing or procumbent although he states that it bears tendrils; the leaves are lobed, the fruit 
depressed or flattened endwise and knotty. It is the 'Melopepo clypeiformis' of Caspar 
Bauhin's Pinax and 'Cucurbita clypeiformis' of Johann Bauhin's Historia; the latter' has a 
picture, which is reproduced in Fig. 40; clypeiformis means shield-shaped or buckler-shaped; 
the plant i^ undoubted]* what we know as Bu*,h Scallop squash In the herbarium of 

Linnaeus there is no specimen of C. melopepo named by him." In the entry for Cucurbita 
melopepo in "Order Out of Chaos," Jarvis (2007) noted "Type not designated." 


a. Cucurbita melopepo L. var. melopepo 

Cucurbita verrucosa L., Sp. PI. 2: 1010. 1753. Pepo verrucosus (L.) Moench, Method. PI. 2: 653. 

1794. Cucurbita pepo var. verrucosa (L.) Beguinot, Fl. Italia 3: 163. 1903. TYPE: Not 
designated. Locality not indicated in the protologue: "Cucurbita foliis lobatis, pomis nodoso- 
verrucosis. Cucurbita verrucosa Bauh. hist. 2. p. 222. Melopepo verrucosus. Tournef. inst. 
106." Fruits subglobose and verrucose, apparently similar to the Orange Waited cultivar. 

: Nomenclature in domesticated/wild Cucurbitaceae 23 

Cucurbita ovifera L., Syst. Natur. (ed. 12) 2: 639. 1766; Mant. PI. 1: 126. 1767. Cucurbita pepo 
subsp. ovifera (L.) Decker, Econ. Bot. 42: 11. 1988. Cucurbita pepo var. ov#era (L.) Alef., 
Landw. Fl., 224. 1866. Cucurbita pepo var. ov;/era (L.) C. Harz, Landw. Samenk. 2: 819. 
1885. LECTOTYPE (Bailey 1929, p. 88, Fig. 41): Russia. Province of Astrakhan. "Habitat 
ad Astrachan. DD. Lerche." Sent by J.J. Lerche from Astrakhan, grown by Linnaeus at the 
Uppsala Botanical Garden, Herb. Linn. No. 1 15 1.2 (LINN). 

"The first validly named ornamental gourd was C. ovifera in 1767. The closest 
extant cuitivar to Linnaeus's description and type specimen appears to be either the 'Striped 
Pear' or 'Egg' gourd. ... The full scientific name for the ornamental gourds today is 
practically always written C. pepo var. ovifera (L.) Alef. Yet Alefeld transferred the epithet 
to the rank of 'Var.-Gr.,' not to 'var.' The first traceable authority to validly publish the name 
C. pepo var. ovifera was Harz in 1885. Other varietal names for the ornamental gourds 
precede this, such as C. pepo var. oviformis Vilm. described in 1863, but the name ovifera has 
priority over all of these" (Andres 1995, pp. 67-68). 

Subsp. ovifera is used in various places in the present discussion, reflecting the 
prevalent use of that name in recent literature, but within Cucurbita pepo, subsp. texana 
(Scheele) Filov (1982) has priority over subsp. ovifera (1988). Few publications have used 
the correct name (e.g., Paris et al. 2002). 

Cucurbita lignosa Mill., Gard. Diet. (ed. 8), Cucurbita sp. no. 5. 1768. TYPE: Not designated. 
Locality not indicated in the protologue. "The fruit of the fifth sort hath a hard shell when 

ripe like the first, which may be dried and preserved many years: these are of very different 
forms and size; some are shaped like a Pear, and are no bigger than a large Catherine Pear; 
some are as large as quart bottles, and almost of the same form; others are round and shaped 
like an Orange, and are of the same size and colour, but these are very variable." Synonym of 
C pepo, fide Andres (1995). If acquired by Miller from along the coast of east-central 
Mexico, it suggests that it might be derivative from C. fraterna or C. pepo sensu stricto. 

Cucurbita subverrucosa Willd., Sp. PI. 4: 609. 1805. Cucurbita pepo var. subverrucosa (Willd.) C. 
Harz, Landw. Samenk. 2: 819. 1885. TYPE: Not designated. Locality not indicated in the 
protologue. "Fructus clavato-ellipticus flavus palmaris verrucis paucis sparsis obsitus. Forte 
hybrida progenies a praecedente [C verrucosa L.] orta, sed fructus longe diversus. W." 
Type material: B-Willdenow fiche 18037!; the sheet says "Hort, bot. Berol. W." Synonym 
of Cucurbita pepo, fide Andres (1995) and apparently C melopepo var. melopepo as 
interpreted here. 

Cucurbita pepo var. condensa L.H. Bailey, Cycl. Amer. Hort. 409. 1900. LECTOTYPE (designated 
here): Origin unknown, cultivated at Cornell University. Trade Name: La. White Scallop 
Squash Bush 41, Stock from: Thornburn, University Garden: 18 Sep 1890, L.H. Bailey s.n. 
(BH 33216 digital image!). The protologue noted only this: "Bush pumpkins. Scallop and 
Summer Crookneck squashes. Plant compact little or not at all running. Of horticultural 

As noted in pencil on the sheets, 8 collections of Cucurbita pepo — most with a photo 
or drawing of the mature fruit — were in a BH folder labeled C pepo var. condensa: 2 of 
"Green-Striped Bush, 2 of Scallop, 2 that approach Crown of Thorns, 1 of "Egg Plant Bush," 
and 1 of Pineapple (a variant of 'turbarf morphology). All of these have labels with a printed 
heading of "Garden Herbarium of the Cornell University Experiment Station" and all were 
collected on September 17 and 18, 1890. Bailey did. not cite a specimen and none of these is 
annotated as "var. condensa." but it seems clear enough that these can be considered original 
material. Bailey (1929) listed var. condensa as a synonym of C. pepo var. melopepo. 

: Nomenclature in domesticated/wild Cucurbitaceae 24 

, CUCURBITA MELOPEPO SUBSPECIES TEXANA (Scheele) G.L. Nesom, comb. nov. Tristemon 

texanus Scheele, Linnaea 21: 586. 1848. 
Cucurbita melopepo L. var. ozarkana (Decker) G.L. Nesom, comb. nov. Cucurbita pepo var. 
ozarkana Decker- Walters, .1. Ethnobiol. 13: 69. 1993. TYPE: USA. Arkansas. Independence 
Co. : 200 m downstream from Batesville, 100 m from edge of White River on first terrace, 6 
Nov 1990, B. Smith & C.W. Cowan 115 (holotype: US digital image!). 

Cucurbita melopepo L. var. texana (Scheele) G.L. Nesom, comb. nov. Cucurbita pepo var. 
texana (Scheele) Decker, Econ. Bot. 42: 12. 1988. Cucurbita pepo subsp. texana (Scheele) 
Filov, Fl. Cult. PI. USSR 21: 177. 1982. Cucurbita texana (Scheele) A. Gray, Boston J. Nat. 
Hist. 6: 193. 1850. Tristemon texanus Scheele, Linnaea 21: 586. 1848. TYPE: USA. Texas. 
Comal Co.: Upper Guadeloupe [River], [margins of thickets,] in moist woods, Sep [1845], 
FJ. Lindheimer Fasc. III. 4000 (holotype: B, presumably; possible isotypes: GH, MO-2 
sheets digital images!). The protologue in Linnaea cites as collection data only this: 
"Niederliegend oder uber niedrigem Gebusch rankend an der Guadeloupe: Lindheimer. Juni." 
One of the MO sheets has a handwritten label noting the number "472'' with a date of "Sept 
1 845," but the printed label has 1846, The other Mo"sheet (3265652) has a handwritten label 
noting the number "360" with a date of "Juni Juli 1846." It is not clear that any of the 
collections at GH and MO are duplicates of the type; those dated September presumably are 

c. Cucurbita melopepo L. var. fraterna (L.H. Bailey) G.L. Nesom, comb. nov. Cucurbita pepo var. 
fraterna (L.H. Bailey) Filov, Fl. Cult. PI. USSR 21: 177. 1982. Cucurbita pepo subsp. 
fraterna (L.H, Bailey) Lira, Andres, & Nee, Estud. Tax. Ecogeograf Cucurbitaceae, 77. 
"1995, Cucurbita fraterna L.H. Bailey, Gentes Herb. 6: 288, Fig. 145. 1943. TYPE: Mexico. 
Tamaulipas. On Mesa de Llera, along roadside, herbaceous vine, corolla orange-yellow, Jul 
1937. CL. Lundell 7289 (holotype: MICH digital image!; isotypes: MO, TEX! ). 

Literature Cited (Cucurbita) 

Andres, T.C. 1987. Cucurbita fraterna.. the closest wild relative and progenitor of C. pepo. Cucurbit 

Genet. Coop. Rep. 10: 69-71. 
Andres, T.C. 1995. Complexities in the infraspecific nomenclature of the Cucurbita pepo complex. 

Acta Horticult. 413 : 65-91. 
Bailey, L.H. 1929. The domesticated Cucurbitas. I. Gentes Herb. 2: 62-115. 
Bailey, L.H. 1943. Species of Cucurbita. Gentes Herb. 6: 267-322. 
Cowan, C.W. and B.D. Smith. 1993. New perspectives on a wild gourd in eastern North America. J. 

Ethnobiol. 13: 17-54. 
Decker, D.S. 1988. Origin(s). evolution, and systematics of Cucurbita pepo (Cucurbitaceae). Econ. 

Bot. 42: 4-15. 
Decker, D.S. and H.D. Wilson. 1987. Allozyme variation in the Cucurbita pepo complex: C pepo 

var. ovifera vs. C. texana. Syst. Bot. 12: 263-273. 
Decker-Walters, D.S., T.W. Walters, C.W. Cowan, and B.D. Smith. 1993. Isozymic characterization 

of wild populations of Cucurbita pepo. J. Ethnobiol. 13: 55-72. 
Decker- Walters, D.S., IE. Staub, S.M. Chung, E. Nakata, and H.D. Quemada. 2002. Diversity in 

free-living populations of Cucurbita pepo (Cucurbitaceae) as assessed by random amplified 

polymorphic DNA Syst. Bot. 27: 19-28. 
Jarvis, C. 2007. Order Out of Chaos: Linnaean Plant Names and Their Types. Linnean Society of 

London in association with the Natural Histoiy Museum, London. 
Kirkpatrick, K.J. and H.D. Wilson. 1988. Interspecific gene flow in Cucurbita: C. texana vs. C pepo. 

Amer. J. Bot. 75:519-527. 

: Nomenclature in domesticated/wild Cucurbitaceae 25 

Lira-Saade, R., M. Nee, and T. Andres. 1995. Cucurbita L. (Chapter 1). Pp. 1-115, in R. Lira- 

Saade, (ed.). Estudios Taxonomicos y Ecogeograficos de las Cucurbitaceae Latinoamericanas 

de Importancia Economica. Systematic and Ecogeographic Studies on Crop Genepools. 9. 

Interntl. PI. Genet. Resources Inst., Rome. 
Merrick, L.C. 1995. Squa'h \ umpl in ind g >uiils u bita (Cucurbitaceae). Pp. 97-105, in J. 

Smartt and N.W. Simmonds (eds.). Evolution of Crop Plants. Longman, Essex, U.K. 
Nee, M. 1990. The domestication of Cucurbita (Cucurbitaceae). Econ. Bot. 44(3, Suppl.): 56-68. 
Newsom, L. A, S.D. Webb, and J.S. Dunbar. 1993. History and geographic distribution of Cucurbita 

pepo gourds in Florida. J. Ethnobiol. 13: 75-97. 
Paris, H.S., N. Yonash, V. Portnoy, N. Mozes-Daube, G. Tzuri. and N. Katzir. 2002. Assessment of 

genetic relationships in Cucurbita pepo (Cucurbitaceae) using DNA markers. Theor. Appl. 

Genet. 106: 971-978. 
Sanjur, O.I., D.R. Piperno, T.C. Andres, and L. Wessel-Beaver. 2002. Phylogenetic relationships 

among domesticated and wild species of Cucurbita (Cucurbitaceae) inferred from a 

mitochondrial gene: Implications for crop plant evolution and areas of origin. Proc. Natl. 

Acad. Sci. U.S.A 99: 535-540. 
Smith, B.D. 1997. The initial domestication of Cucurbita pepo in the Americas 10,000 years ago. 

Science 276: 932-934. 
Teppner, H. 2000, Cucurbita pepo (Cucurbitaceae) — history, seed coat types, thin coated seeds and 

their genetics. Phyton (Horn) 40: 1^2. 
Teppner, H. 2004. Notes on Lagenaria and Cucurbita (Cucurbitaceae) — review and new 

contributions. Phyton (Horn) 44: 245-308. 
Whitaker, T.W. and W. P. Bemis. 1975. Origin and evolution of the cultivated Cucurbita. Bull. 

Torrey Bot. Club 102: 362-368. 


I am grateful to Dr. Jochen Heinrichs, Curator of GOET, for information on authentic 
material of Citrullus studied by Sehrader and loan of the Fischer specimens, to Tracy Wells and 

at the Kew Herbarium Library for providing the relevant page of the 1834 Goettingen seed 
catalogue (see Citrullus coffer), to Tom Wendt arid the TEX staff for arranging the loan from GOET 
and for hospitality while studying at TEX, to Amber Schoneman at TEX for providing the images of 
C. caffer, to Roy Gereau and Peter Goldblatt at MO for information on African seed catalogs and 
place names, to Anna Stalter and Peter Fraissinet at BH for information and digital images of 
Citrullus specimens considered to be original material of L.H. Bailey, and to Nancy Elder, Life 
Sciences Librarian at the University of Texas, and Barney Lipscomb at BRIT for help with literature. 
I also am indebted to Jim Reveal for comments on choice of a type for Cucurbita melopepo, Tom 
Wendt for an interesting and helpful discussion on the typification of Citrullus coffer, and Kanchi 
Gandhi for critical observations toward the correct interpretations of authorship citations for names by 
Sehrader in Citrullus. This study was done in conjunction with preparation of the FNANM t 
of Cucurbitaceae and supported in part by the Flora of North America Association. 

t domesticated/wild 




Figure 1. Lectot>pe of Citruttus cqffra Schrad (GOET). See text. 

t domesticated/wild 

Figure 2. Isolectotype ofCitrullus caffra Schrad. (GOET). See text. 

i: Nomenclatire in domesticated/wild 

Figure 3. Isolectotype of Citrullus cqffra Schrad. (GOET). See text. 


Figure 4. Lectotype ofCitrullus vulgaris var. citroides L.H. Bailey (BH). See text. 

Nesom: Nomenclature in domesticated/wild Cucurbitai 

Figure 5. "Mini seedless" triploid watermelon, Citrullus lanatus, at the end of its remarkable oddysey 
beginning from a small, fertile-diploid, bitter-flesh African ancestor. This one was grown in Mexico, 
according to its label. 


Figure 6. "Melones" in Bauhin and Cherler's HistoriaPlantaUniversalis (1651, vol. 2, p. 242). The 
page with the illustration was cited by Linnaeus in the 1753 protologue of Cucumis melo. Along with 
theprotologueand lectotype, this illustration and the one from Moris on's Plantarum Historiae 
Universalis (in Fig. 5) provide a fuller morphological formulation of Linnaeus's concept of the species 
and of C. melo var. melo in the strict sense. 

a .JSf&f&j Mattfa ■ iter, 

Figure 7 "Melo" in Robert Monsoris Plantarum Historme Umv ersalis (1 680. vol. 2, series 1 , tab. 6, 
fig. 4). This illustration is referred to in Hortus Cliffortianus (Linnaeus 1 737) and Florae Leydensis 
(van Royen 1 740), both cited by Linnaeus in the 1 753 protologue of Cucumis melo. 

n : NcmendatLre in domesticated/wld GjiLrbitai 

Figure 8. Cucurbita clypeiformis from Bauhin and Cher ler, Historia plantarum universalis 2: 224. 
1651. The page with this illustration was cited by Linnaeus in the protologue of Cucurbita melopepo 
L. (1753) and the illustration is designated above as the lectotype of the species.