2E. Environmental variation and local adaptation


The habitats of the White Sands pupfish vary considerably in regard to salinity levels. Salinity is relatively high at Lost River (13.5-60ppt), moderate at Salt Creek (10.5-21ppt), and relatively low at Mound (1.5-4ppt) and spatially variable at Malpais Springs (3.5-21.5ppt)(Stockwell and Mulvey 1998).
This difference in salinity appears to provide the conditions for local adaptation; indeed an allozyme locus appears to be under selection due to differences in environmental salinity in these habitats (Stockwell & Mulvey 1998). Stockwell and Mulvey (1998) reported rapid evolution at an allozyme locus (Phosphogluconate dehydrogenase; Pgdh) in recently established populations of White Sands pupfish (C. tularosa). The allele frequencies in two recently established populations (Lost River and Mound Spring) varied significantly from allele frequencies in the native parental population (Salt Creek). This shift and additional evidence suggested that Pgdh might be under selection by salinity (Stockwell & Mulvey 1998).

Further, salinity has indirect effects on pupfish, because complex life cycle parasites can be excluded from the sites with high salinity (Rogowski and Stockwell 2006). For instance, a white grub (Family: Diplostomatidae) occurred at high prevalence and intensity in fish from Malpais Spring and Mound Spring, but was absent from fish examined from Salt Creek and Lost River (Stockwell and Collyer unpublished data). The life cycle of white grubs includes snails (Physa sp.), fish, and the definitive host is typically herons (Hoffman 1967). The absence of the white grub at the two saline sites is apparently due to the absence of snails (Physa sp.) at these sites. Recent experimental work has shown that Physid snails can not tolerate salinity above 7ppt (Stockwell unpublished data). Environmental differences also affect the distribution of other parasites. For instance, two additional parasite species occurred only in Salt Creek pupfish; Heterophyidae species 1 and 2 (Rogowski and Stockwell 2006) (Stockwell and Collyer unpublished data).

A subsequent habitat survey revealed the presumptive host of these parasites, a recently described springsnail species (Juturnia tularosa; Hershler et al. 2002) that occurs at high abundance in Salt Creek. Whether this springsnail species is actually part of the life cycle of the two trematode species is yet to be determined. These parasites may be costly to their hosts. Indeed, a population decline have observed at Mound Spring in 1995 coincided with parasite outbreaks (Pittenger 1996). Three lines of evidence suggest that parasitism by Diplostomatids are costly. First, seasonal differences in parasite intensities suggest that heavy parasite burdens reduce over-winter survivorship at Malpais Spring and Mound Spring (Collyer and Stockwell 2004). Second, parasite load was inversely correlated with fat storage (Stockwell unpublished data). Third, recent experimental work conducted at North Dakota State University has shown that this parasite is costly in terms of mortality risk, fat storage and growth rates (Collyer and Stockwell 2004). These data are consistent with the idea that parasite load may regulate pupfish populations as indicated by the population crashes at Mound Spring.

Flow regime also varies among the sites and may influence fish body shape. Recent work revealed that the Salt Creek and Lost River fish are streamlined, whereas Malpais Spring and Mound Spring fish are deep bodied (Collyer 2003; Collyer et al. 2005). These body shapes correspond with the flow conditions in the local habitas. Salt Creek and Lost River have periodic high flows, whereas Malpais Spring and Mound Spring have no appreciable flow. A common garden experiment showed body shape to be heritable, providing evidence for a case of contemporary evolution (sensu Stockwell et al. 2003) for the Mound Spring population (Collyer 2003; Collyer et al. 2005).

These environmental data show that the ecological characteristics vary considerably between the two native sites as well as between the two sites with introduced populations of pupfish. Thus, translocations may result in altering the evolutionary trajectory of the targeted taxum.
Translocation efforts are typically made to assure that the "refuge" habitat is representative of the targeted taxa's native habitat. For this strategy to be successful, extensive ecological information will be required from both the native and targeted "refuge" habitats. Even under the "best case" scenarios, one or more important ecological factors are likely to differ between the native site and the refuge site, which could potentially lead to rapid evolution in the refuge site (Stockwell and Weeks 1999; Stockwell et al. 2003).

Thus, rapid evolution may ultimately lead to a refuge population that is not adapted to its native environment. This issue is of particular concern, because the establishment of at least one additional refuge population of White Sands pupfish is a possible conservation strategy (Stockwell et al. 1998).


2A. Relationships among populations

2B. Introduction History

2C. Fragmentation and gene flow

2D. Monitoring habitats and populations

2E. Environmental variation and local adaptation

2F. Potential impacts of exotic species

2G. Biblography