Article 8:

(Essay) Detection at Multiple Levels: Edith's Checkerspot Butterfly and Climate Change

By Camille Parmesan, University of Texas, Austin

The year was 1992. Global warming was headline news. Were cars and air-conditioning driving the planet into a new climate regime? As climate scientists debated, biologists worried. NASA responded to this concern by expanding its Office of Mission to Planet Earth to include biological responses to climate change. Could the effects of the very slight warming recorded by thermometers (only about one half a degree Celsuius at that time) be detected in wild species?

To any researcher working in the field with Edith’s checkerspot butterfly (Euphydryas editha), it was clear that this species was very sensitive to yearly variation of climate. This intuition was backed up by 40 years of publications documenting responses of wild populations to the vagaries of weather and a very seasonal climate. With years of studying the ecology and behavior of E. editha in the wild, a bit of NASA backing, and a four-wheel drive Subaru, I set out on what would be a nearly five-year adventure in search of climate change impacts.

One of the simplest and most intuitive expected responses of species to global warming is a shift of species distributions towards the poles and up the mountains. This tracking of an envelope of climate in which a species can survive can be seen clearly in the fossil records of many species, from beetles to shrews to spruce. It’s not surprising, then, that this is one of the first signs of response to twentieth century warming looked for by scientists. My first realization in my adventure was that for this particular question, simple though it is, the skills of a detective are as essential as the skills of a scientist.

In order to discover if a species has moved, one must first know where it has been. Although field guides contained approximate range maps for North American butterflies, these were not accurate enough to detect subtle range shifts. The first task, then, was to gather as many historical records as possible. I spent one year visiting natural history museums in the United States, Canada, Britain, and France, then another six months visiting private collectors.

Make a note to add training in sociology and diplomacy to the list of skills needed. Private collectors are wary of exposing their secluded sites to the general public, with good reason—the extinction of the large copper (Lycaena dispar) in Britain is attributed to over-enthusiastic butterfly collectors. Trust is an essential component.

Out of thousands of specimens, there were nearly 1400 distinct locations recorded, most of which had only been visited once. Fortunately, early lepidopterists thought each ecotype of E. editha was a separate species, therefore historical records were abundant and sites were evenly spread across the current range of E. editha. It was clear that a change in distribution could not be detected without a recensusing of these historically recorded populations. About 600 sites were labeled with enough detail to potentially find again. After five field seasons (4.5 years) and 40,000 miles, my many helpers and I had managed to visit 292 of these.

Ultimately, this unintended experimental design was essential for interpreting the data. By having a current assessment of each site, I was able to classify the sites as to whether the habitat was still suitable for the butterfly. If the habitat was degraded (because of over-grazing or logging), then that site was classified as unsuitable and not included in the final analysis. Impacts of land use on wild populations is a major concern for conservation, but for this particular study, the goal was to tease apart those types of human impacts from the effects of long-term climate change. Therefore, only sites that still appeared to contain suitable habitat (defined as an abundance of healthy host plants and local nectar sources) were included in the climate change analysis. Once habitat was deemed to be suitable, a census was conducted to determine if the butterfly population still existed or had gone extinct.

Local extinctions are a natural part of E. editha biology. Just as a single drought cannot be said to be caused by global warming, a single population extinction caused by that drought cannot be linked to global warming. The link to climate change is easier to detect by looking for particular patterns of populations extinctions over time and over many hundreds of populations across the entire species range.

The present-day censuses of these historical sites did indeed reveal an asymmetrical pattern of population extinctions on a continental scale. While climate change could have caused subtle responses, or responses that were idiosyncratic for each habitat type, the actual pattern was quite simple.

Further, the altitudinal cline in frequency of population extinction had a breakpoint at 2400 m (fewer extinctions at the highest elevations). Over the same time period, winter snowpack showed a similar pattern of change, also with a breakpoint at 2400 m. Below 2440 m, snowpack had become 14% lighter and melt date had advanced by one week during the twentieth century (p<0.05 for both). In contrast, snowpack had become 8% heavier and melt date had not changed above 2440 m (Johnson et al. 1999).

Separate analyses showed that other factors (such as proximity to large urban areas) were not associated with the observed extinction patterns (Parmesan, unpublished). Since the only strong associations were between the extinction patterns and various climatic trends, regional climate warming was, by default, the most likely cause of the observed distributional shift.

Fortunately, correlating patterns are not the only evidence. Prior empirical studies had pinpointed specific cause and effect links between various aspects of climate and weather events and E. editha behavior, physiology, individual fitness, and population dynamics. This empirical evidence provides strong reinforcement of the causal link between climate change and the observed range shift.

Effects of Weather and Climate on Population Dynamics

Many population extinctions have been associated with particular climatic events (Singer and Ehrlich 1979; Ehrlich et al. 1980; Singer and Thomas 1996; McLaughlin et al. 2002). The 1975–1977 severe drought over California caused the extinction of 5 of 21 surveyed populations (Singer and Ehrlich 1979; Ehrlich et al. 1980). Wet years have had opposite effects in different habitat types, causing population crashes in the San Francisco Bay area (E. editha bayensis; Dobkin et al. 1987) and population booms in Mexico (E. editha quino; Murphy and White 1984). The wet year in Mexico was associated with apparently long-distance dispersal, making it likely that many colonizations took place during this rare event.

In a large metapopulation (Rabbit Meadow) that sits just under the 2400 m breakpoint in the Sierra Nevada Mountains (California), all local populations using the host plant Collinsia torreyi went extinct in response to three distinct extreme climate events (Singer and Thomas 1996; Thomas et al. 1996). Between 1989 and 1992, two winters of exceptionally light snowpack caused early flight seasons, each of which led to local population declines of an order of magnitude. These “false springs,” which were followed by severe weather (snowstorms that are normal for the time of year), affected the population at a vulnerable stage (adult butterflies). The final event that caused extinction of the already depleted populations was a severe post flight season freeze (all pre-diapause larval offspring were killed).

In the mountains, trends towards lighter winter snowpack in lower elevation populations (Johnson et al. 1999), such as at Rabbit Meadow, have likely caused an increase in detrimental “false spring” events. This would be consistent with increased population extinctions below 2400 m where snowpack has declined over the twentieth century.

The Importance of Butterfly–Host Plant Synchrony

Empirical studies on E. editha within a population have elucidated many subtle and complex interactions between climatic variability and individuals. Warmer temperatures cause faster development. Local topographic diversity (e.g., north and south facing hillsides) and variation among habitats in vegetation cover can give as much variation of temperature as is found across years, from 2°C to 4°C difference. This has allowed researchers to study responses to fine-scale temperature variations. A 2°C warmer microclimate raises larval temperature by up to 3°C which speeds up larval growth rate by several days (Weiss et al. 1988; Boughton 1999).

However, a reoccurring theme from decades of experiments is that synchrony is more important than absolute speed. The relationship between climate and survival of E. editha is typically mediated not by direct effects of temperature or precipitation on the insect, but by their indirect effects on timing of the butterfly’s life cycle relative to that of their host and nectar plants. In general, weather conditions that speed the plant life cycle relative to that of the insect (such as hot, cloudy, or dry conditions) cause increased larval mortality. The reverse is also true—conditions that slow the plants relative to the insects increase insect fitness. (Singer 1972, 1983; Singer and Ehrlich 1979; Weiss et al. 1987, 1988; Boughton 1999). The gradual warming and drying trend in southern California (Karl et al. 1996) has likely led to a steady shortening of the window of time in which the host is edible, causing increased larval mortality in these southernmost populations.

The phase relationships between the butterflies and their host plants represents a perpetual balance problem for individuals in their attempts to maintain synchrony with their host plants while maximizing mating success or fecundity. Because the phase relationship is strongly affected by temperature and precipitation, any systematic trends in climate will affect this synchrony. Thus, the ultimate population response to systematic climatic trends depends on the interplay between host plant distribution across the micro- and macro-topographic landscape, larval and adult dispersal, and female choice of oviposition sites (Singer 1971,1972; Weiss et al. 1988; Boughton 1999).

Unlike many other insects, E. editha has evolved a life history strategy that renders them “at the limits of their ecological tolerance,” making them susceptible to extreme climate years (Singer 1971). In the short term, the high extinction rates of southern E. editha populations support the view that this species is tightly adapted to mean local conditions, with little evidence for flexibility in its behavior or life history strategies (Singer 1994). However, E. editha has demonstrated a remarkable ability for rapid evolution (Singer et al. 1993), which makes long-term response to climate change more difficult to predict.

The Quino Checkerspot: Endangered Species and Climate Change

The Quino checkerspot (E. editha quino) is a federally listed endangered subspecies of Edith’s checkerspot. Although habitat destruction is the primary cause of its decline, climate change poses problems for its recovery. Quino checkerspot populations along the southernmost boundary (in Mexico) are at the greatest risk from continuing warming and drying climate trends. Unfortunately, these are also the best remaining habitats with the lowest degree of threat from development. By contrast, most Quino habitat has been destroyed by development in the Los Angeles–San Diego corridor. The case of the Quino checkerspot has resulted in the first habitat recovery plan to list climate change not only as a current threat, but also as a factor that should be considered in reserve design and recovery management (Anderson et al. 2001).

Literature Cited

Anderson, A., E. Allen, M. Dodero, T. Longcore, D. D. Murphy, C. Parmesan, G. Pratt, M. C. Singer. 2001. Quino Checkerspot Butterfly (Euphydryas editha quino) Recovery Plan. U.S. Fish and Wildlife Service, Portland, OR.

Boughton, D. A. 1999. Empirical evidence for source-sink dynamics in a butterfly: Temporal barriers and alternative states. Ecology 80: 2727–2739.

Dobkin, D. S., I. Olivieri, and P. R. Ehrlich. 1987. Rainfall and the interaction of microclimate with larval resources in the population dynamics of checkerspot butterflies (Euphydryas editha) inhabiting serpentine grassland. Oecologia 71:161–166.

Ehrlich, P. R., D. D. Murphy, M. C. Singer, C. B. Sherwood, R.R. White and I. L. Brown. 1980. Extinction, reduction, stability and increase: The responses of checkerspot butterfly populations to the California drought. Oecologia 46:101–105.

Johnson, T., J. Dozier, and J. Michaelsen. 1999. Climate change and Sierra Nevada snowpack. IAHS Publication 256:63–70.

Karl, T. R., R. W. Knight, D. R. Easterling, and R. G. Quayle. 1996. Indices of climate change for the United States. B. Am. Meteorol. Soc. 77:279–292.

McLaughlin, J. F., J. J. Hellmann, C. L. Boggs and P. R. Ehrlich 2002. Climate change hastens population extinctions. Proc. Nat. Acad. Sci. USA 99:6070–6074.

Murphy, D. D. and R. R. White. 1984. Rainfall, resources, and dispersal in southern populations of Euphydryas editha (Lepidoptera: Nymphalidae). Pan-Pac. Entomol. 60:350–354.

Singer, M. C. 1971. Ecological studies on the butterfly, Euphydryas editha. Ph.D. dissertation, Stanford University, CA.

Singer, M. C. 1972. Complex components of habitat suitability within a butterfly colony. Science 173:75–77.

Singer, M. C. 1983. Determinants of multiple host use by a phytophagous insect population. Evolution 37:389–403.

Singer, M. C. 1994. Behavioral constraints on the evolutionary expansion of insect diet; A case history from checkerspot butterflies. In L. Real. (ed.), Behavioral Mechanisms in Evolutionary Ecology, pp. 279–296. University of Chicago Press, Chicago.

Singer, M. C. and P. R. Ehrlich. 1979. Population dynamics of the checkerspot butterfly Euphydryas editha. Forts. Zool. 25:53–60.

Singer, M. C., C. D. Thomas, and C. Parmesan. 1993. Rapid human-induced evolution of insect diet. Nature 366:681–683.

Singer, M. C. and C. D. Thomas. 1996. Evolutionary responses of a butterfly metapopulation to human and climate-caused environmental variation. Am. Nat. 148:S9–S39.

Thomas, C. D., M. C. Singer, and D. Boughton. 1996. Catastrophic extinction of population sources in a butterfly metapopulation. Am. Nat. 148:957–975.

Weiss, S. B., D. D. Murphy, and R. R. White. 1988. Sun, slope and butterflies: topographic determinants of habitat quality for Euphydryas editha. Ecology 69:1486–1496.

Weiss, S. B., R. R. White, D. D. Murphy and P. R. Ehrlich. 1987. Growth and dispersal of larvae of the checkerspot butterfly Euphydryas editha. Oikos 50:161–166.