9430 - Laserfiche WebLink

Home Browse Search

Meffe and V'rijenhoek zoogeographic models must be recognized in manage- mentprograms aimed at habitat restoration and restock- ing of fish populations. The Death Valley Model Death Valley and similaz intermontane basins of the Southwest contain a series of extremely isolated spring- heads and small streams that were connected in piuviai times when the basins were filled by inland lakes and seas (Miller 1948; Hubbs & Miller 1948). Asa conse- quence of natural post-Pleistocene drying wends, these habitats now contain remnant fish populations that were formerly part of a larger interbreeding population. Pres- ently, geographical isolation among colonies is com- plete (m = 0; Fig. 1), and divergence is accelerated by small colony sizes and possibly strong, local selection pressures. Total genetic diversity of a species in this mode! (Ht) consists of diversity within colonies (H~) plus differ- ences among colonies within the total range (D~t) (Nei 1975). If a species is endemic to a single locality, then the latter term drops out and total genetic diversity con- sists of heterorygosity within the single site. As colonies lose variation and become more homozygous via ge- netic drift, random differences among colonies also ac- cumulate through time. Since we envision no migration (m = 0) to counteract these processes, both the de- cline of H~ and the growth of Dot are simple functions of Ne within colonies. Depending on its mode of action, natural selection DEATH VALLEY MODEL 0 _®_ 0 0 ~~- `~~ 0 ~ ~~ D 0~ IO \ ~ ~~ ~~ 0\~ U------~~ 0 0 H t = He + Dc t Figure 1. Graphic representation of the Death Valley model of population structure Isolated springs and streams (represented by Awl) have no chance of nat- ural gene flow among them (m = 0). Conseraton Generics olDesert Fishes 159 will alter the effects of genetic drift. Directional selec- tion may accelerate divergence for locally adaptive traits, but divergence will be slowed for characters that aze under some form of stabilizing or balancing selec- tion throughout the range of the species. This model is exemplified by pupfishes (genus Cyprinodon) and kil- lifishes (genus Empetrtchthys), which aze naturally iso- lated in many remnant springs and streams in Death Valley and adjacent basins in California and Nevada (Miller 1948). Considerable evidence exists for morphological, physiological, behavioral, and life historical differences among populations of desert pupfishes (Naiman & Soltz 1981), supporting this model. Turner (1974) was the first investigator to study genetic population structure in these fishes. His electrophoretic studies of pupfishes in and near Death Valley revealed little overall allelic vaziation. However, most vaziation occurred among, rather than within, these remnant populations. For ex- ample, Cyprinodon nevadensu contains three recog- nized subspecies that occur in isolated habitats of the Amargosa basin. Most of the thirty gene loci he exam- ined were homozygous within populations, but four were polymorphic in one population. Cyprinodon di- aboluand G Salinas; which apparently derived from G nevadensis since the drying of Death Valley, presently occur in single isolated habitats, and each species ap- pears to be homorygous at most loci for alleles that occur in G nevadensts. However, each species also con- tains some unique alleles that may have arisen via mu- tation since isolation from the presumed ancestor. Un- fortunately, one cannot compute genetic diversity statistics from the data presented in Turner's pioneering study, but in general the data aze consistent with our model: very little polymorphism occurs within, and con- siderable divergence occurs among, populations. In managing a genetic population structure repre- sented by the Death Valley model, we need not be con- cernedwith human interruption of gene flow. Such gene flow has historically not existed, at least not since an- cestral lakes have desiccated. On the contrary, we should take precautions to avoid creating gene flow among populations that are naturally isolated and may therefore be in the process of adaptive radiation. Artifi- cial gene flow would tend to reduce or eliminate ge- netic divergence among localities (D«), which may rep- resent asubstantial proportion of the genetic diversity in the species. The major concern for fishes in the Death Valley model is maintenance of high Ne within localities. Crashes due to man's intervention have occurred, jeop- azdizing several species' existence. In August 1969, the Owens pupfish (Cyprinodon radiosus) nearly went ex- tinct when its only known habitat completely dried (Miller & Pinter 1971). Fortunately, some 800 fish were Conservation Biology Volume 2, No. 2, June 1988