Contributed Papers A Quantitative Conservation Approach for the Endangered Butterfly Maculinea alcon.

For endangered species with limited dispersal capacity, the preservation of individual patches or sites is extremely important. The quality and size of habitat patches and their isolation from other patches are the main parameters for an assessment of population persistence, but translating persistence probabilities into practical measures is still a weak link in conservation management. I provide a quantitative conservation approach for the endangered myrmecophilous butterfly Maculinea alcon in the Netherlands. All 127 colonies known on heathland since 1990 were investigated for patch quality, size, and isolation. I assessed habitat quality in three 10 × 10 m plots for most colonies. Site occupancy in 1998–1999 was only 56%. Occupancy was best explained by a logistic regression including patch area, host ant presence, host plant abundance, overall heathland area, and connectivity between sites ( 2= 0.410, < 0.0001); it correctly classified the occupied or vacant status for 82% of the sites. Connectivity contributed only 3.6% to the total explained variation of site occupancy, indicating that habitat characteristics were more important than isolation in determining population persistence at the examined scale level (>500 m). Grazing and sod cutting had a beneficial impact, but in combination these practices proved detrimental. Hydrological measures to prevent drainage were also associated with lowered occupancy. I used the different components in the logistic regression to formulate objective management recommendations. These consisted of sod cutting, reduction of management intensity, enlargement of habitat, or combinations of these recommendations. The results highlight the importance of careful management when site quality is determined by multiple factors. The quantitative conservation approach followed here can be fruitfully extended to other endangered species, provided enough is known about their ecological requirements and how management actions affect them. [ABSTRACT FROM AUTHOR]