Dispersal Behavior and Its Implications for Reserve Design in a Rare Oregon Butterfly

Abstract

To identify reserve design strategies for an endangered species, we need to understand the dispersal behavior of the species. For example, constructing corridors for an endangered species makes sense only if that species remains in its primary habitat and rarely strays into adjoining areas. I investigated the dispersal behavior of the Fender's blue butterfly ( Icaricia icarioides fenderi ) to assess possible reserve designs for it. The Fender's blue is a rare butterfly that survives in 13 remnants of upland prairie in the Willamette Valley in Or- egon. I observed butterfly flight patterns in primary lupine habitat, in nearby abandoned fields, and at the boundaries between the habitats. Observations indicate that, on an individual basis, the butterflies weakly bias their movements toward lupine areas; on a population basis, however, more than 95% are found within 10 m of lupine patches. These observations imply that although the Fender's blue is usually found in lupine patches, they stray into nonlupine areas frequently enough to call into question the effectiveness of a corridor. Butterfly flight paths were mapped in lupine and nonlupine areas to determine how their rates of spread de- pended on the habitat in which they were found. Butterflies within lupine areas dispersed at relatively slow rates ( , 3 m 2 /s), whereas those flying outside lupine areas disperse much more rapidly ( . 15 m 2 /s). Observa- tions of daily activity patterns suggest that the butterflies fly about 2.3 hours a day, and a mark-recapture- release experiment indicates that butterflies live about 9.5 days. I used the estimates of diffusion and total flight time to suggest the range Fender's blue butterflies disperse in their lifetime. I estimate that a Fender's blue might disperse about 0.75 km in its lifetime if it remains in lupine habitat but more than 2 km if it dis- perses between lupine patches. Historically, the Willamette Valley was a mosaic of upland and wetland prai- rie, with lupine patches rarely more than 0.5 km apart. At that time, the Fender's blue would have had a high probability of dispersing between patches. Today, however, lupine patches are isolated, with 3-30 km be- tween patches. Thus, it is extremely unlikely that dispersing butterflies find new patches when they leave their natal lupine patch. Dispersal observations suggest that stepping stones between existing patches may be more effective than corridors at facilitating the exchange of individuals.