Fish metacommunity dynamics and threatened species ecology in a desert river system

Habitat fragmentation is one of the major causes of local and regional species extinctions in freshwater ecosystems. To predict future trends in community composition, and the potential sequence of extinctions due to fragmentation of the river continuum, it is important to understand how local and regional processes affect the dynamics of species’ immigration and extinction rates, and patterns of abundance and occupancy. In Chapter II, I examined fish immigration and extinction rates, and abundance and occupancy patterns, in relation to habitat isolation, size, and variability in the Rio Grande and its tributaries in the Trans-Pecos region of Texas, USA. The results indicated that as habitat isolation increased in tributaries, fish immigration rate and riverine species abundances decreased. Fish assemblages showed significant nested subset patterns across study sites, and the influence of habitat size and isolation on the nested subset patterns varied with spatial scale. Certain non-native fish taxa were idiosyncratic in their distribution, reducing the strength of the nested subset pattern. In Chapter III, I examined species-habitat volume relationships, nested subset patterns, and distance-decay relationships at the stream-reach scale for fish assemblages in 42 pools from three Rio Grande system stream reaches with different fragmentation levels. The slopes of the species-habitat volume curves did not significantly differ among the stream reaches, suggesting that within stream-reach, rescue effects were not apparent. In contrast, the intercepts of the species-volume curves considerably varied among the stream reaches, suggesting that rescue effects from the mainstem Rio Grande may play a role in determining species richness in stream reaches. The resulting nested subset patterns showed that fish assemblage structure at my study sites was largely extinction-driven, and large, deep pools are important aquatic refugia that can prevent the local extinction of certain fish taxa. I found a significant distance-decay relationship for fish assemblages in the most fragmented stream segment, possibly because of predation by non-native species on native species in disconnected pool habitats. The results of my research suggest that increased habitat fragmentation by human activities can accelerate the local and regional extinction of certain native fish taxa and contribute to the dominance of ecologically tolerant species (including non-native species), leading to a decline in regional diversity. The Rio Grande and its tributaries in the Trans-Pecos region of Texas have been impacted by a variety of anthropogenic activities such as dewatering and the introduction of non-native species. These environmental manipulations have negatively affected native fish populations leading to extirpations and population declines throughout the region. Campostoma ornatum and Notropis chihuahua inhabit Rio Grande tributary streams in the Trans-Pecos region and are Texas state-listed threatened species. Little is known about their status and ecological requirements in the region. In Chapter IV, I analyzed variation in abundance and incidence of C. ornatum and N. chihuahua from Alamito, Terlingua and Tornillo creeks, as well as the mainstem Rio Grande, in relation to local environmental factors (e.g., stream size and water quality), abundance of nonnative species, season, and distance from the Rio Grande. The abundance of N. chihuahua was positively related to maximum pool depth, and both incidence and abundance of C. ornatum were positively related to the percentage of gravel substrate. These results suggest that decreases in habitat size and gravel substrate will negatively affect the persistence and maintenance of these threatened species in the tributaries. In Chapter V, I examined variation in the seasonal fish assemblages in Tornillo Creek within and between two time periods (1967–1970 vs. 2009–2011) to test whether native and nonnative invasive species have increased their relative abundances and persistence in the creek and replaced the threatened species. I also examined contemporary fish assemblage–environment associations in order to understand what environmental factors were associated with seasonal fish abundance and occupancy patterns. The results of this study indicated that fish assemblages were very different between the two time periods. This was largely due to the declining persistence and relative abundance of C. ornatum and the increasing persistence and relative abundance of four other fish species (Cyprinella lutrensis, Gambusia affinis, Cyprinus carpio, and Notropis braytoni). Contemporary seasonal fish assemblage patterns in Tornillo Creek were less variable than historical assemblages and were linked to several environmental factors including water temperature, stream depth, and current velocity. I suggest that the maintenance of stream flow and connectivity are important for the immigration and emigration of certain riverine fishes in Tornillo Creek.