II.2. SQUASHES, PUMPKINS, ZUCCHINIS, GOURDS (CURCURBITA SPECIES) – 107
Some work on self-incompatibility and inbreeding depression in the genus has been
performed. Some authors have seen little evidence of inbreeding depression in members
of the genus (Borghi, 1976). Others, however, have observed indications of inbreeding
depression. Mahzabin, Parvez and Alam (2008) indicate that Cucurbita maxima shows
abrupt inbreeding depression after two generations of selfing. Cardoso (2004) observed
inbreeding depression affecting certain traits in a Cucurbita moschata variety after
four successive self-pollination generations. Cardoso (2004) and Hayes, Winsor and
Stephenson (2005a, 2005b) have studied inbreeding depression after four successive
generations of self-pollination in Cucurbita pepo var. texana. In general, inbreeding
depression seems to be intense, which suggests a level of genetic variation at least as
recessive deleterious genes. The selfing rate showed a range from 0.16 to 0.54, but this
might vary among characters, years and conditions (Hayes, Winsor and Stephenson,
2005a, 2005b). Whitaker and Robinson (1986) suggest these different observations might
represent the response of different species or varieties of Cucurbita to inbreeding.
Molecular markers
A number of different genetic markers have been used to analyse the phylogeny of
the genus. Wilson, Doebley and Duvall (1992) studied 15 species using chloroplast
restricted fragment length polymorphism analysis and analysed the relationships between
different cultivars of C. pepo. The cultivated species, with the exception of C. ficifolia,
form a cluster, and the relationships among C. moschata, C. argyrosperma ssp.
argyrosperma and C. argyrosperma ssp. sororia were not resolved. Jobst, King and
Hemleben (1998) analysed the internal transcribed spacers of ribosomal DNA for
11 species of the genus, but the alleles between species are shared, possibly because
they are species of very recent origin and/or because there is gene flow between species.
Using ISSRs (inter-simple sequence repeats), Katzir et al. (2000) analysed C. pepo and
discovered a clear differentiation between C. pepo ssp. pepo and C. pepo ssp. ovifera (as
found in other studies) with C. fraterna clustering with C. pepo ssp. ovifera.
King, Jobst and Hemleben (1995) studied nuclear microsatellites in nine species of
the genus. They suggested that Cucurbita (like most of the Cucurbitaceae) contains a
large amount of satellite DNA. In particular, species of the genus Cucurbita contain
interesting specific satellite DNA with individual variations among some species.
Within Cucurbita, the genes coding for the ribosomal 18S, 5.8S and 25S rRNA are
present in high copy numbers and appear highly methylated (Hemleben et al., 1988;
King et al., 1993; Torres-Ruíz and Hemleben, 1994). Mitochondrial DNA is also
specially structured and larger than in other angiosperms. Sanjur et al. (2002) analysed
65 individuals from 14 taxa (8 species of Cucurbita and 2 outgroups) with a
mitochondrial nad1 gene. They found 16 haplotypes. Four groups can be defined in the
phylogenetic analysis: a basal group including C. ficifolia; a group composed of
C. foetidissima; a group formed by C. maxima, C. andreana and C. ecuadorensis; and a
large group with C. okeechobeensis ssp. martinezii at the base, including C. pepo,
C. argyrosperma ssp. sororia, C. argyropserma ssp. argyrosperma and C. moschata.
Six independent origins of domestication can be inferred based on this phylogeny.
Other phylogenetic studies have been carried out, both at wider levels, analysing for
instance the relationships of Cucurbita with other plant groups (an example within the
family is offered by Chung, Decker-Walters and Staub, 2003), and within a given species
(as within C. pepo as in Paris et al., 2003).