Home

General Research Interests

My research examines how ecological and evolutionary forces acting within stream networks shape the population biology and community interactions of resident species. In addition to providing basic insight, I hope this work will lead to empirical and theoretical tools for landscape-scale conservation of stream biodiversity, and thereby bring the scale of stream research in line with the scale of human impact to stream systems (i.e., entire streams and drainages of multiple streams). By opening up stream networks to research questions derived from spatially explicit models of species demography, community ecology, and evolution, I expect this research to provide the challenges and understanding that make for exciting and useful science.

I pursue these research interests using a combination of

experimental, descriptive, molecular genetic, and modeling

techniques, and by remaining open to perspectives and

methods from across the sub-disciplines of ecology and

evolutionary biology. I find that this approach leads to

compelling research questions, elucidates mechanisms

and causality, and lets me explore conservation and

Matt Ayres

management applications of my results.

Current projects

Spatial population dynamics in streams

Genetic isolation by slope and distance

in a headwater stream network

Effects of timber harvest on stream

salamander populations

The genetics of post-glacial recolonization

In the spring salamander Gyrinophilus

porphyriticus (Plethodontidae)

Spatial population dynamics in streams

Few studies directly quantify the contribution of dispersal to local population dynamics. There is also minimal empirical support for the view that persistence of resident stream organisms requires a balance between downstream drift and upstream dispersal. I am conducting an intensive study of movement and population dynamics in the headwater stream salamander Gyrinophilus porphyriticus (Plethodontidae) with the objectives of (1) generating a quantitative description of the frequency, directionality, and temporal pattern of dispersal, and (2) assessing the effects of movement along the stream corridor on the local population dynamics of this species. G. porphyriticus dispersal displays a consistent upstream bias, overcompensating for downstream drift. In a focal stream, net emigration from downstream reaches offset reduced recruitment in the upper reaches. These surprising findings have led me to explore the ecology and evolution of directionally-biased dispersal.

To identify the practical implications of these results, I constructed a metapopulation model to simulate the effects of landscape-scale perturbation on stream organisms, incorporating interpopulation dispersal. Results of these simulations showed the importance of identifying and preserving source populations and dispersal routes for stream species in human-impacted networks. They also highlighted the vulnerability of headwater specialists to anthropogenic perturbation and the strong positive effect of stream restoration when recolonization is possible. Return to top of page

Genetic isolation by slope and distance in a headwater stream network

There is broad support for the negative relationship between interpopulation gene flow and distance, where distance is derived from two-dimensional geographic coordinates. Although populations are distributed across three-dimensional landscapes, the general relationship between gene flow and slope is not well understood. Using the amplified fragment length polymorphism (AFLP) technique, we tested the hypothesis that gene flow between populations of the headwater stream salamander Gyrinophilus porphyriticus (Plethodontidae) is independently affected by distance, measured along the stream corridor, and slope, represented by change in elevation along the stream corridor. Because G. porphyriticus exhibits upstream-biased dispersal, we predicted that increased slope would act primarily to limit gene flow between populations. With data from 18 G. porphyriticus populations located throughout the 31.6 km² Hubbard Brook Watershed in New Hampshire, we found that pairwise genetic and geographic distances were positively correlated. Genetic distances between downstream and upstream populations in nine study streams were also positively related to change in elevation over a standardized 1-km distance. These patterns of isolation by slope and distance are consistent with phenotypic variation observed among G. porphyriticus populations. More generally, our results suggest that constraints on gene flow resulting from an association with montane and headwater habitats promote genetic differentiation in G. porphyriticus, and thus may help to explain why these habitats are global hotspots of plethodontid diversity. These results also elucidate management requirements for population connectivity in stream networks, and indicate that population distribution along both distance and elevation gradients may be important predictors of gene flow and dispersal.

Collaborators: Gene E. Likens, Mark A. McPeek

Return to top of page

Effects of timber harvest on stream salamander populations

Timber harvest is an important perturbation to headwater streams throughout the United States, yet its effects on resident species and the mechanisms underlying these effects are poorly understood. A general lack of understanding of the conservation biology and community ecology of small, 1^st and 2^nd-order streams has led to insufficient protection of these streams under state and national land-use regulations. We investigated the effects of logging, aquatic community composition, and population connectivity on the stream salamander Gyrinophilus porphyriticus, a headwater specialist. There were strong negative effects of logging-associated sedimentation and brook trout on G. porphyriticus abundance. Abundance of salamander larvae was negatively related to brook trout abundance, while adult salamander abundance was negatively related to sedimentation. These findings underscore the value of information on species life history and community ecology in assessing sensitivity to anthropogenic disturbance, suggesting that larval resistance to sedimentation can buffer populations from extinction in fishless streams impacted by logging. This work also led to the first evidence that large-scale population connectivity can increase the likelihood of species persistence in stream networks exposed to perturbation. The presence of a confluent occupied stream, a potential immigrant source, mitigated the negative effects of timber harvest on G. porphyriticus populations.

Collaborators: Doug T. Bolger, Keith H. Nislow

Return to top of page

The genetics of post-glacial recolonization in the spring salamander Gyrinophilus porphyriticus (Plethodontidae)

Molecular genetic techniques can provide valuable insight on dispersal patterns at both historic and contemporary time scales, and can elucidate evolutionary constraints on species ecology and behavior. Gyrinophilus porphyriticus is one of only three stream salamander species to recolonize the northern Appalachians following the Wisconsinan Glaciation, approximately 12,000 years ago. We assessed large-scale spatial variability in DNA haplotype diversity among G. porphyriticus populations south and north of the glacial boundary. Two reciprocally monophyletic clades were recovered within the northern range of this species: one from the coastal plain of southeastern Pennsylvania, south of the Wisconsinan glacial boundary, and the other from non-glaciated and glaciated areas within and west of the Appalachian uplift (i.e., western PA, NY, NH, and VT). This suggests that the Appalachians may be an old and major phylogeographic boundary in this species, and a historic dispersal barrier. Relative to these northern samples, the number of distinct haplotypes recovered from streams within a 3-km² area in Virginia was surprisingly high. These data illustrated the difference in age and associated level of genetic heterogeneity within and among G. porphyriticus populations in the southern and northern portions of the species’ range. They also documented a striking lack of correlation between the topology of the stream network where the Virginia samples were collected and the topology of the resulting phylogenetic trees. This difference may be a consequence of contemporary gene flow.

Collaborators: H. Brad Shaffer

Home