Landscape genetics of stream fishes and genetic indicators of environmental condition

It is well understood that the diversity and distribution of freshwater fishes reflect past climate variation, drainage evolution and contemporary land use. It is becoming clear that these forces also may act on the genetic diversity and genetic structure of freshwater fishes. We are working with colleagues at the US EPA and several universities to examine the influence of land use and habitat degradation on the genetic diversity and genetic structure of stream fishes, to better understand how contemporary factors exacerbate historical conditions to compromise population or species persistence. This work is also intended to promote the development and use of genetic indicators of aquatic environmental condition.

Much of this research has so far focused on common and abundant species in eastern North American (ENA) stream fish assemblages. The studies that have been completed on stonerollers (Cyprinidae: Campostoma) exemplify how information on historical and contemporary factors can enhance conservation and management of stream fishes. A molecular phylogenetic analysis of the genus demonstrated the presence of cryptic evolutionary diversity (Blum et al., 2008). The recovered phylogeny argues for revising the genus to recognize at least nine Campostoma lineages as species. Extensive comparisons of within and between lineage variation suggest that evolutionary lineages of stonerollers inhabiting once-glaciated basins have significantly lower genetic diversity (e.g. heterozygosity) compared to stonerollers in unglaciated, southern basins (Blum et al., 2012). Additional research on the variance of within-lineage genetic variation across mid Atlantic basins indicates that regional patterns of genetic diversity strongly reflect historical conditions and contemporary environmental heterogeneity (Blum and Bagley, in prep.). Campostoma anomalum pullum in New York and Pennsylvania, for example, exhibit naturally low levels of genetic diversity, but diversity is even lower in streams with high pH. On the other hand, the naturally low levels of genetic diversity within C. anomalum anomalum in Pennsylvania are not further depressed in streams with high pH. Rather, C. anomalum anomalum are more sensitive to reduced discharge. Similar findings for other species, including creek chubs (Semotilus atromaculatus) and white suckers (Catostomus commersonii), reinforce that genetic information on biogeography and contemporary ecology can help identify and conserve at-risk populations of stream fishes (Blum et al., 2012).  

This work motivated further research on basin and watershed scale patterns of genetic variation. Focusing on central stonerollers (Campostoma anomalum anomalum), mottled sculpin (Cottus bairdii) and rainbow darters (Etheostoma caeruleum) in Indiana and Ohio, these studies  have shown that basin-level patterns of genetic diversity correspond to drainage area, which likely reflects variation in habitat carrying capacity (Blum and Bagley, in prep). Patterns of genetic variation at finer spatial scales strongly reflect land use and habitat characteristics (e.g. stream depth, substrate diversity) that influence reach-level carrying capacity (Waits et al., 2008; Blum et al., 2012).  At an even finer spatial, a study of central stonerollers examining effective population size and migration in the urbanized Mill Creek basin (Ohio) found evidence of asymmetric source-sink dynamics between minimally impacted headwaters and lower mainstem reaches in the Cincinnati metro area (Waits et al., 2008). Since migration is often low among river basins (Waits et al., 2008; Darling et al., in prep.), particularly among species exhibiting relatively low vagility (e.g. mottled sculpin; Lamphere and Blum, 2012), this suggests that restoration of fish populations in urban watersheds requires the security of intact tributaries or headwaters as well as remediation of impacted sites. By demonstrating that drainage area, location and environmental heterogeneity can explain fine scale patterns of genetic variation in stream fishes, these studies also provide a conceptual framework for the use of genetic indicators of aquatic environmental condition (Blum et al., 2012).  

Building from this work, we recently completed an interdisciplinary study focusing on the development and validation of genetic assessment protocols for oceanic island stream ecosystems. Current assessment protocols for oceanic island streams use measures of native fish species diversity and the presence of tolerant native fish species to rate levels of impairment. These protocols are likely to underestimate impairment because fish assemblages of insular streams are naturally depauperate. Similarly, since nearly all native stream fishes on oceanic islands are amphidromous (a life history involving both freshwater and marine stages), responses to environmental stressors may be misunderstood if consideration is not given to the processes that sustain populations in degraded waterways relative to near-shore habitats. By assessing individual- and population level responses to environmental stressors, genetic analysis of native stream biota can provide more sensitive bioassays of anthropogenic impacts than traditional assemblage-level analyses. However, use of genetic assessment protocols for oceanic islands requires identifying geographic and environmental factors responsible for variation in recruitment and dispersal of native amphidromous fishes. To better understand conditions contributing to genetic variation, we undertook genetic comparisons of immigrating juveniles to resident adults, and assessed migration using otolith microchemistry. We also examined how measures of genetic variation, field census estimates of population size, and movement rates derived from otolith microchemistry vary according to in-stream and watershed conditions.  Completion of this work was intended to increase basic understanding of insular stream ecology and responses of native amphidromous fishes to environmental stressors, and to identify the spatial scales over which watershed management and restoration efforts are most beneficial to native fishes on oceanic islands. The project final report is available here: http://www.serdp-estcp.org/Program-Areas/Resource-Conservation-and-Climate-Change/Natural-Resources/Watershed-Processes-and-Management/RC-1646.

Control and mitigation of aquatic invasive species in Pacific Island streams

Conservation and management of at-risk native species on Pacific islands requires stewardship of ecosystems altered by non-native species. Because eradication is not always feasible, consideration must be given to alternative management approaches, including restoration of key ecological processes that enable at-risk native species to persist in the presence of non-native species. Though management of aquatic invasive species (AIS) are consistently identified as a primary threat to biodiversity in Pacific island streams, AIS control is poorly understood and even less is known about mitigative approaches, such as stream flow restoration, that can suppress AIS and sustain at-risk native species. The objective of this project is to characterize genes-to-ecosystem outcomes of experimental AIS removals across a stream flow gradient and two flow restoration scenarios as well as to build model-based decision tools to determine the extent to which AIS control, mitigation, or some combination thereof fosters recovery of key ecological processes and TER-S in Pacific island streams.

 

This project is being conducted in watersheds across the Koolau Range on the Hawaiian island of Oahu. Following a before-after-control-impact study design, genes-to-ecosystem outcomes of removing non-native fish will be assessed in 12 watersheds selected to represent stream flow variation on Oahu. After 6 months of pre-manipulation assessment, non-native fish will be removed from one 100-m stream reach in each watershed, with efforts targeting drift-feeding poeciliids (i.e., guppies) and algae-eating loricariids (i.e., catfish). Responses to removals will be tracked for a period of two years, with adjacent upstream and downstream reaches serving as references.

 Following a paired catchment study design, genetic variation, life history, and demography of Awaous stamineus, as well as community and ecosystem responses to AIS removal also will be characterized under conditions of diverted continuous low flow, undiverted continuous high flow, and pulsed high flow matching diurnal biotic rhythms. Comparisons will be made within and among watersheds to assess outcomes of flow modification with and without AIS removal.

 

Data from the two removal studies will be used to parameterize a coupled biophysical individual-based model (IBM) to evaluate the sensitivity of native species to AIS removal and flow restoration. The data also will be used to build systematic conservation planning (SCP) models to evaluate cost and benefit functions for selecting sites and watersheds to maximize return on investments in AIS management. 

 

Development of actionable information and innovative approaches for managing AIS can substantively improve stewardship of stream ecosystems that cross DoD lands, especially on islands where installations harbor native species under federal or state protection. Native fishes and invertebrates are becoming increasingly imperiled in Pacific island streams. This project will demonstrate a novel approach to AIS management, advance understanding of the response of native species and ecological processes to AIS control, and determine whether outcomes of AIS control differ according to hydrological conditions. This project also will demonstrate whether flow restoration can be an effective tool for AIS management, and whether mitigation augments at-risk native species responses to AIS control in Pacific island streams. Finally, the project will deliver model-based prioritization tools that will provide defensible and transparent analytical frameworks for adaptive watershed management and aquatic species conservation on Pacific islands.

Environmental determinants of hybridization among native and invasive freshwater fishes

Hybridization between non-native and endemic fish species can lead to the loss of native biodiversity. Nowhere is this more of a concern than in southeastern US river basins that harbor the greatest number of endemic fish species in North America. In collaboration with graduate students and colleagues at the USGS, Duquesne University, and the University of Georgia, we are examining hybridization among non-native and native Cyprinella minnows across the upper Coosa River basin (Alabama, Georgia, Tennessee). This research aims to characterize the formation of a hybrid swarm between introduced red shiners (C. lutrensis) and endemic blacktail shiners (C. venusta stigmatura) in order to assess the risk of hybridization between C. lutrensis and all native congeners in southeastern rivers, including federally threatened blue shiners (C. caerulea).

Observational studies of hybridization between naturally sympatric C. lutrensis and C. venusta in the Brazos River (Texas) and San Marcos River (Texas) have noted that episodes of hybridization often coincide with conditions of high water turbidity, and that hybridization appears to cease following remediation of turbid conditions. This suggests that reducing turbidity in the Coosa River basin could eliminate hybridization between C. lutrensis and native congeners. GIS analysis of land use across the basin, habitat assessments, and survey-based distributional studies indicate that water turbidity and agricultural land use are strong predictors of the presence and abundance of C. lutrensis x C. venusta hybrids in mainstem reaches and tributaries in the upper Coosa River basin (Walters et al., 2008). Behavioral trials also demonstrated strong sexual isolation between C. lutrensis and C. venusta (Blum et al., 2010, Ward et al., 2012), which supports findings from molecular assays of hybrid populations indicating that F1 hybrids are exceedingly rare. However, the presence of later generation hybrids in the upper Coosa River basin and evidence of heterosis from studies of hybrid fitness also suggest that hybrid progeny are able to reproduce (Walters et al., 2008; Blum et al., 2010; Ward et al., 2012).

Several follow-on studies are now underway. Genetic comparisons of non-native to native populations are being carried out to determine the number of introduction events, the putative source(s) of introduced C. lutrensis, and whether post-introduction admixture has occurred in the Coosa River basin (Glotzbecker et al, in revision). Further comparisons are being made between non-native populations in the Coosa River basin to introduced populations in North Carolina, Utah, Arizona, Nevada and California to determine whether separate invasions exhibit genetic parallels such as loss of genetic diversity or admixture following multiple introductions (Glotzbecker et al, in revision). Mate choice trials are being conducted to experimentally test whether turbidity obscures courtship behaviors that act as reproductive isolating barriers (Glotzbecker et al., 2015; Ward et al., in prep). In addition, ethotoxicological assays are being carried out to test whether chronic or acute exposure to estrogenic mimics during early development and adulthood gives rise to aberrant spawning behavior that weakens sexual isolation (Ward and Blum, 2012; Ward et al., in prep).