Research

1. Comparative landscape & cityscape genomics

Land cover (top) and resistance surface (bottom) maps used to model dispersal and gene flow across the landscape – New Haven, Conn. in this example.

One primary goal of conservation is to understand what aspects of modern landscapes impede or promote movement of wildlife – and sometimes to mitigate the barriers. To this end, I use genomic sequencing (ddRADseq) and landscape resistance modeling to identify these potential barriers among populations of mammals and amphibians. One project involves comparing five mammal species found across urban to rural gradients in southern New England. In another project we are comparing patterns of gene flow for rats in four major cities that serve as “replicate” landscapes (Salvador, Brazil, New York City, Vancouver and New Orleans). The sampling design and statistical analyses of this work allow me to directly compare how landscapes affect species in similar or distinct ways (Richardson et al. 2016). That way we can provide generalities that apply to most landscapes, or highlight impacts unique to a particular landscape or species. For example, wood frogs and spotted salamanders are very similar ecologically, yet are impacted by different parts of the same landscape: roads impede wood frog movement, while rivers hinder spotted salamander gene flow in New England (Richardson 2012). This work highlights the fact that landscapes do not have homogonous effects on species, and the importance of investigating multiple species to assess landscape effects on dispersal.

2. Urban ecology & genetic insights into public health in cities

Left: Pau da Lima favela in Salvador, Brasil. Right: density and landscape resistance maps for rat movement.

Left: Pau da Lima favela in Salvador, Brasil. Right: density and landscape resistance maps for rat movement.

Salvador, Brasil is a large city of three million residents that has experienced a 280% increase in its population since 1970. Much of this added population is concentrated in favelas that are characterized by refuse piles, open sewage and overgrown vegetation. These conditions promote Norway rat infestations – a reservoir host for the zoonotic pathogen that causes leptospirosis. Leptospirosis outbreaks occur each year in favelas and millions of dollars are spent annually on public health campaigns to eradicate these rats. In order to target interventions, epidemiologists and public health officials have requested information on the areas where rats are moving between suitable slum habitats. We are using genomic data to evaluate movement patterns for >700 Norway rats trapped over 5 years so that public health authorities can target areas of high movement for increased eradication resources (Richardson et al. 2017)

3. Conservation genomics

Conservation efforts focus on the long-term persistence of populations, including demographic trends and genetic viability over time. Spring-breeding amphibian aggregations provide a good system to answer questions that lie at the intersection of population demography and population genetics. In many species, all adults arrive at temporary ponds to breed shortly after ponds have thawed. As a result, it is straightforward to census these populations during this contracted breeding period. However, detecting trends can take many years. To compare the use of survey data with efficiently obtained genetic estimates of population size, I am comparing long-term (15 years) demographic estimates of population size (Nc) with genetic estimates of effective population size (Ne) that can be obtained within weeks. Encouragingly, the genetic and demographic estimates correspond closely, indicating that Ne may be a suitable tool for pressing conservation assessments of amphibian populations.

4. Ecological Physiology

Marbled salamander larvae must overwinter in challenging pond conditions

All physiological processes occur within an ecological context. Within this environmental context, we are examining the physiological adaptations that allow the marbled salamander (Ambystoma opacum) to persist in the northern part of its range. Their larvae must overwinter in ponds, and as a result, their range is restricted to areas with mean winter temperatures near or above freezing. Marbled salamander occurrence may be associated with dissolved oxygen (DO) levels, which decrease rapidly without air exchange once ponds freeze. To explore the physiological mechanisms behind this pattern we are A) monitoring DO levels in ponds throughout their northern range, B) using experiments to test physiological tolerance to freezing and reduced DO, and C) exploring potential adaptations in gill structure and oxygen consumption across a latitudinal gradient in temperature. This project has direct implications for climate change, as the marbled salamander has been expanding northward as winter temperature increases. In separate ongoing work, I am also exploring variation in amphibian metabolism using data on foraging behavior, assimilation efficiency and elemental nutrient composition (i.e. carbon:nitrogen:hydrogen ratios).

5. Evolutionary ecology across spatial scales

We use experiments in natural habitats (top) and simulations (bottom) to estimate the speed and spatial scale of evolutionary divergence among populations.

We use experiments in natural habitats (top) and simulations (bottom) to estimate the speed and spatial scale of evolutionary divergence among populations.

Evolutionary divergence between populations depends on the interplay between natural selection and gene flow between habitats. Given assumptions about high gene flow, very little research has explored this divergence at fine spatial scales. Using transplant experiments between ponds, novel genetic analyses and controlled lab experiments, we have found that adaptive divergence of amphibian populations can occur at fine spatial scales in response to strong natural selection (Richardson & Urban 2013). I am currently extending this work by (A) manipulating natural selection in the field and tracking the community response, and (B) using transcriptomics (RNAseq) to measure differences in gene expression in response to selection.

Relative size difference between the spotted (left) and predatory marbled (right) salamander larvae by June in Connecticut vernal ponds. Right: shared zooplankton food resource.

We are exploring these eco-evolutionary dynamics in a set of natural pond communities that experience heterogeneous predation risks (and hence natural selection) across the landscape. We have found that predation risk across years leads to evolved responses in spotted salamander foraging behavior and growth rates (Urban & Richardson 2015). We are currently conducting whole-pond and enclosure experiments to manipulate natural selection from aquatic predators, and then tracking both the evolutionary response of their salamander prey and the community-level resource effects across trophic levels (i.e. primary productivity, zooplankton diversity and abundance, macroinvertebrate assemblages).

Comments are closed, but trackbacks and pingbacks are open.