9430 - Laserfiche WebLink

Home Browse Search

1 SS Conserntion Genetics of Desert Fishes Introduction Fishes in the Sonoran, Mohave, Great Basin, and Chihua- huan deserts of North America aze declining at an alarttl- ing rate. At least 20 taxa have gone extinct in the last several decades (Meffe, in press), and many more are in immediate danger of similaz fates (Williams et al. 1985). This major conservation problem has prompted govern- mental agencies and academic scientists to conduct re- search on basic ecologies and life histories, develop propagation programs for species perpetuation and re- introduction, and restore degraded habitats. We address here a different aspect of conservation of this highly endangered fauna, the application of population genet- ics to management and long-term perpetuation of desert fishes. Desert Aquatic Systems North American deserts aze extremely arid, with high geographic relief, heavily dissected drainages, and widely dispersed surface waters. Natural aquatic sys- tems in these regions aze fragmented on both broad geographic and local scales. In addition to being iso- lated, desert aquatic systems can experience enormous physico-chemical fluctuations, as in streams and rivers (Hinckley & Meffe 1987), or constant but hazsh condi- tions, as in some springs (Deacon & Hinckley 1974). The extreme isolation and severe abiotic conditions have led to a high degree of phenotypic divergence of fishes, often with small populations living in small hab- itats (Miller 1958). Many taxa aze relicts from better- watered times that have been trapped in isolated springs and streams during the last 10,000- to 12,000-yeaz post- pluvial period. This has resulted in a high degree of endemism (Williams et al. 1985 ). Several genetic and demographic consequences of such an isolated, fragmented distribution are likely, in- cluding (1) local divergence via natural selection or genetic drift; (2) little or no gene IIow among isolated demes that might otherwise moderate losses of genetic variability after population crashes; and (3) little or no recolonization of isolated habitats after local extinction. Southwestern desert fishes aze thus a vulnerable, "~- tinction-prone" goup due to the limited geographic range and isolation of many species. Added to this natural fragility aze major anthropo- genic disturbances of two types: habitat destruction (impoundment and diversion of rivers, groundwater pumping, and drying of surface springs) and the intro- duction of a huge diversity of exotic fishes, many of whom prey on, compete with, or hybridize with native species (Naiman & Soltz 1981). Management of such a Mege and Viijenhoek fauna over the long term must therefore consider the genetic consequences both of natural fragmentation with small population sizes and of artificial disruption by man. General and philosophical aspects of genetic man- agement of fishes have been outlined elsewhere (Meffe 1986, 1987; Ryman & Utter 1987). Here we consider more specific approaches of genetic studies and exper- imental manipulations of raze and endangered south- western fishes. Divergence, Migration, and Gene Flow Two demographic factors, genetically effective popula- tion size and migration rate, affect the degree of diver- gence among local colonies of fishes. Colonies diverge from one another as a consequence of local selection pressures, mutation, and random genetic drift; drift alone can cause considerable divergence among small colonies. The rate of divergence due to drift in turn depends on the genetically effective size of the local populations (Ne). Ne represents the number of breeding adults in each generation, and is affected by sex ratio, pattern of mating, and variance in reproductive output among individuals (Craw & Kimura 1970). Gene flow via migration (m, or the proportion of in- dividuals exchanged between colonies per generation) maintains genetic variation within colonies and retazds divergence among colonies. Divergence occurs as a product of Ne and m. if Ne is small, colonies tend to diverge rapidly, as a result of random processes. High rates of gene IIow (m) aze needed to maintain genetic homogeneity among them. As long as N~rn > 1, local colonies will tend not to diverge significantly with re- gazd to the types of alleles present (Allendorf 1983)• Thus, a pair of large populations with mean Ne > 10,000 that exchange individuals at a rate of m = 0.001 will not diverge significantly by chance alone, since Nenn > 10. However, a pair of small populations with Ne < 1000 and with a similaz rate of gene exchange would diverge, sitlce Ne < 1. This scenario is based on the assumption that colonies are dispersed geographically as if they were islands and gene flow can occur equally among all the islands. As we discuss below, this scenario may be an appropriate de- scription for some aquatic systems (the Death Valley model), but for most freshwater fishes, riverine habitats impose a more complex population structure (the Stream Hierarchy model). Zoogeographic Models of Gene Flow We envision two models of genetical population struc- ture in desert fishes. The existence of these different Conservation Biology Volume 2, No. 2, Junc-1988