U.S. Department of Energy, Office of Science

Program for Ecosystem Research

Research Project   Predicting the impact of climatic change on animal distributions: the importance of local adaptation and species' traits

Principal investigator:   Jessica J. Hellmann

Butterfly study locations

Sites constituting the periphery and the core of the species' ranges are circled. Filled and unfilled symbols indicate the northern half of these butterflies' coastal distribution and define the study region. Filled dots are locations where genetic samples have been taken; unfilled dots are projected sampling locales. Laboratory warming experiments are also being conducted on the University of Notre Dame campus.

Project goal

To test the generality of geographic range shifts under climatic change by evaluating if populations of contrasting species are adapted to climatic conditions at the edge of their range.

Ecosystem being studied

Peripheral versus core populations of two co-occurring butterfly species, Erynnis propertius and Papilio zelicaon, are studied in oak-grasslands of western North America. The study species differ in body size (vagility) and resource specialization, providing an opportunity to detect differences in adaptation and range dynamics if they occur. These species are indicators of Mediterranean-type ecosystems in North America, systems that harbor a number of organisms of national significance.


Species with traits that promote gene flow should be well-adapted to the central portion of their range and increase at their poleward range edge when warming occurs. In contrast, species with traits that reduce gene flow may show adaptation to local conditions and not increase at the poleward (cool) range edge when warming occurs. The initial prediction was that P. zelicaon would conform to the former and E. propertius would conform to the latter. Results to date provide mixed support for this prediction (see a-c vs. d below). Current studies are aimed to rectifying these mixed findings.

Findings to date include:

(a) Erynnis propertius increases in density with increasing latitude while P. zelicaon declines. This is evidence that edge populations of E. propertius are well-adapted to the edge of their range, perhaps more so than to core conditions.

(b) The two species differ in their responsiveness to climatic variation at the edge of their ranges. Erynnis propertius show equal fitness when translocated among sites near the range edge, while the fitness of P. zelicaon increased when placed in warmer locations. This indicates that P. zelicaon may respond favorably to local warming at the range edge and thus have the capacity for a poleward range shift, perhaps more so than E. propertius.

(c) Both species show evidence of genetic differentiation (based on microsatellite and mtDNA markers) at the edge of their range, indicating that populations may be locally adapted to peripheral conditions. This divergence between core and peripheral populations was more pronounced in E. propertius.

(d) Both species show improvements in fitness when reared under conditions characteristic of the center of their range, though P. zelicaon is sensitive to events of extreme heat stress. This result was independent of source population, indicating that populations at the range edge of both species could increase if conditions there became more like the range core (i.e., warmer).

Why this is important

Understanding the responses of biodiversity to climatic change is vital to setting conservation goals and to managing the delivery of ecosystem goods and services. Reducing scientific uncertainty about the genetic basis of species ranges, and whether ranges of organisms and ecosystems can be expected to move in response to future climatic changes, will be essential to improving forecasts of future organism and ecosystem ranges and to forecasting whether species extinctions are likely to occur with future climatic change.

Butterfly population structure

Population structure in E. propertius (top) and P. zelicaon (bottom) inferred from microsatellite markers. These bar plots indicate Bayesian assignment probabilities in STRUCTURE analysis for K population clusters. Each vertical bar indicates one individual. Colors indicate different population clusters such that individuals of the same color are likely to be related to one another. Both species are genetically distinct at the northern portion of their range with population and regional differences being more pronounced in E. propertius. (From Zakharov & Hellmann 2008)


The project examines gene expression of populations in different locations within the two species' ranges.

During 2008, field-collected caterpillars of both species will be studied with 454 DNA sequencing to generate a large number of sequences expressed during the larval stage. Butterfly larvae are particularly sensitive to climate and all experiments in this project focus on that life stage. Transcript data will be used to create a microarray for each species.

During 2009, field-collected caterpillars will be exposed to climatic conditions representing the edge and the core of the species' distributions (i.e., a warming experiment for peripheral populations), and extracted mRNA (cDNA) will be hybridized to the chip. These methods will enable detection of differences in gene expression among populations and species under varying climatic conditions, revealing the number and identity of traits affecting climate-related fitness.


Jessica J. Hellmann, University of Notre Dame

Neil Lobo, University of Notre Dame

Jason Dzurisin, University of Notre Dame

Shannon Pelini, University of Notre Dame


Zakharov EV, Hellmann JJ (2008) Genetic differentiation across a latitudinal gradient in two co-occurring butterfly species: revealing population differences in a context of climate change. Molecular Ecology 17:189-208

Funding period:   March 2005 to present