Genetic variation in the endangered Astragalus jaegerianus (Fabaceae, Papilionoideae): a geographically restricted species.

Abstract.--Knowledge of genetic variation and population structure is critically important in the conservation of endangered species. The level and partitioning of genetic variation in the narrow endemic Astragalus jaegerianus was investigated using DNA sequence data and AFLP markers. The DNA sequence data for the cpDNA trnL-F and nrDNA ITS regions were monomorphic for A. jaegerianus but polymorphic for two congeners, which suggest A. jaegerianus is genetically depauperate. On the other hand, the genome-wide survey using AFLP markers revealed substantial gene diversity (0.2660) and significant population structure (global [F.sub.ST] = 0.133, p

**********

Genetic variation is recognized as an important raw material of evolutionary and ecological processes in natural populations (Wright 1931; Karron et al. 1988; Clegg 1990; Epperson 1990; Frankham et al. 2002; Conner and Hartl 2004; but see Ackerly 2003). Effective conservation strategies for geographically restricted species require information about the level of genetic variation present and whether genetically distinct subpopulations exist. This information can aid management efforts by suggesting number of individuals and populations needed for the conservation of the species (Mace and Lande 1991; Lynch 1996; Nunney 2000; Frankham et al. 2002). The federally endangered Astragalus jaegerianus Munz (Fabaceae) is a narrow endemic of the Mojave Desert (southern California; USA) with small populations of relatively few individuals (U.S. Fish and Wildlife Service 2001). It is restricted within a range of 300 [km.sup.2] but only occupies approximately 85 [km.sup.2] of that habitat. In an intensive survey conducted in 2001, this species had an estimated population size of 5,700 plants (U.S. Fish and Wildlife Service 2001; Charis Professional Services Corporation [Charis] 2002; M. Hessing, Charis, unpublished data). To assist conservation efforts for Astragalus jaegerianus, we investigated the level of genetic variation and its partitioning within the range of this species.

Population genetic theory predicts that geographically widespread plant species with large populations should typically possess higher levels of genetic variation than do geographically restricted species with small populations (reviewed in Hamrick et al. 1979; Hamrick and Godt 1989; Karron 1991; Soltis and Soltis 1991; Godt et al. 1996; Frankham et al. 2002; Avise 2004). This is predicated on the premise that species with small populations and few individuals are subject to genetic drift that reduces genetic variation (Barrett and Kohn 1991; Ellstrand and Elam 1993). Other factors affecting low genetic variation in restricted species may include strong directional selection and increased inbreeding (Karron 1991). While such reviews provide valuable comparisons of levels of genetic variation among widespread and restricted species, the effects of rarity on genetic variability may be confounded by other variables, such as historic demographic changes in population size, the lack of shared evolutionary histories, and sampling at dissimilar spatial scales (Karron 1987b; Gitzendanner and Soltis 2000; Cole 2003). For instance, in their review of the literature on levels of genetic variability between congeners, Gitzendanner and Soltis (2000) did find a pattern of higher levels of genetic variation for geographically widespread species than their geographically restricted congeners. But, they also note that there is a wide variance in the range of genetic variation among geographically restricted species, such that some restricted species exhibit as high or higher levels of genetic variation as their widespread congeners, while some widespread species exhibit low levels of genetic variation similar to that found in their restricted congeners (see also Karron 1987b; Karron et al. 1988; Soltis and Soltis 1991; Frankham 1995; Avise 2004).

Concerning historic demographic changes in populations, Karron (1987b; Karron et al. 1988) compared levels of genetic variation (allozyme variants) in geographically restricted and widespread vascular plant species across eleven genera that included Astragalus and found that while geographically restricted species tend to have lower levels of genetic variation than widespread species, there are exceptions. Additionally, Travis et al. (1996) examined AFLP variation in populations of the endangered sentry milk vetch (Astragalus cremnophylax Barneby var. cremnophylax) on the north and south rims of the Grand Canyon, and found that north rim populations possess a level of genetic variation consistent with historical levels, while the south rim populations possess almost no genetic variation. In both of these studies, the authors hypothesized that such heterogeneity in levels of genetic variation among geographically restricted species may be partially attributed to factors affecting reduced range size during the recent past. However, in the case of A. jaegerianus, there is no evidence of a historic range reduction (Munz 1941; Barneby 1964; U.S. Fish and Wildlife Service 2001; Charis 2002).

Population genetic theory also predicts that narrow endemic species with small populations and few individuals should exhibit strong population structure due to a higher proportion of genetic variation partitioned among populations (Hamrick and Godt 1989; Hamrick et al. 1991; Frankham et al. 2002; Avise 2004). However, Hamrick and Godt's (1989) review of the allozyme literature revealed that, while some geographically widespread species typically have a greater proportion of genetic variation partitioned within populations, there is no significant difference between geographically widespread and restricted species in how genetic variation is partitioned. In contrast, theory predicts that species with restricted ranges and small populations are expected to have a greater proportion of genetic variation partitioned among populations (Karron 1991; Frankham et al. 2002; Avise 2004).

In this study, we tested the hypotheses that levels of genetic variation in species of restricted ranges and few individuals should be low, with a significant proportion of this variation partitioned among populations. We also examined whether the genetic structure of Astragalus jaegerianus correlates with its population geographic distribution. Such an estimate of genetic variation and population structure provides a basis for sustainable management of Astragalus jaegerianus.