Genetic diversity and population structure in the tomato-like nightshades Solanum lycopersicoides and S. sitiens

Abstract

Background and aims: Two closely related, wild tomato-like nightshade species, Solanum lycopersicoides and Solanum sitiens, inhabit a small area within the Atacama Desert region of Peru and Chile. Each species possesses unique traits, including abiotic and biotic stress tolerances, and can be hybridized with cultivated tomato. Conservation and utilization of these tomato relatives would benefit from an understanding of genetic diversity and relationships within and between populations.

Methods: Levels of genetic diversity and population genetic structure were investigated by genotyping representative accessions of each species with a set of simple sequence repeat (SSR) and allozyme markers.

Key results: As expected for self-incompatible species, populations of S. lycopersicoides and S. sitiens were relatively diverse, but contained less diversity than the wild tomato Solanum chilense, a related allogamous species native to this region. Populations of S. lycopersicoides were slightly more diverse than populations of S. sitiens according to SSRs, but the opposite trend was found with allozymes. A higher coefficient of inbreeding was noted in S. sitiens. A pattern of isolation by distance was evident in both species, consistent with the highly fragmented nature of the populations in situ. The populations of each taxon showed strong geographical structure, with evidence for three major groups, corresponding to the northern, central and southern elements of their respective distributions.

Conclusions: This information should be useful for optimizing regeneration strategies, for sampling of the populations for genes of interest, and for guiding future in situ conservation efforts.

Fig. 1.

Map showing the geographical locations of S. lycopersicoides and S. sitiens populations used in this study. Each population is named after the collection site (see Table 1).

Fig. 2.

Comparison of population genetic diversity in S. lycopersicoides (L), S. sitiens (S) and S. chilense (C). Values are based on a common set of markers (Table 2) applied to 12–13 S. lycopersicoides populations (one additional population was genotyped with SSRs only), 6–7 S. sitiens populations and one S. chilense population (SSRs only). Bars represent allozyme data and microsatellite data, as indicated. Values are the averages (+s.d.) of all tested populations within each species.

Fig. 3.

Microsatellite phylogeny of S. lycopersicoides accessions based on Cavalli-Sforza's chord distance. Numbers at nodes represent the frequency (expressed as a percentage) at which a node occurred among 10 000 replications. Geographical groups are circled. A = northern, B = central and C = southern groups.

Fig. 4.

Microsatellite phylogeny of S. sitiens accessions based on Nei's genetic distance. Numbers at nodes represent the frequency (expressed as a percentage) at which a node occurred among 10 000 replications. Geographical groups are circled. A = northern, B = central and C = southern groups.

Fig. 5.

STRUCTURE analysis of S. lycopersicoides populations based on SSR data. Bars represent the membership coefficients (Q) of individual plants (298 total in 14 populations), based on SSR allele frequencies, using K values for two, three or 14 groups. Numbers on the horizontal axes correspond to the population numbers in Table 1.

Fig. 6.

STRUCTURE analysis of S. sitiens populations based on SSR data. Bars represent the membership coefficients (Q) of individual plants (155 total in seven populations), based on SSR allele frequencies, using K values for two, three or seven groups. Numbers on the horizontal axes correspond to the population numbers in Table 1.

Fig. 7.

Levels of genetic diversity within population clusters of S. lycopersicoides. Values represent the averages (+s.d.) of populations within each group, based on allozymes (stippled bars) or SSRs (striped). N = northern, C = central and S = southern cluster.

Fig. 8.

Levels of genetic diversity within population clusters of S. sitiens. Values represent the averages (+s.d.) of populations within each group, based on allozymes (stippled bars) or SSRs (striped). N = northern, C = central and S = southern cluster.