The decline and rise of the Kransberg Cape Vulture colony over 35 years has implications for composite population indices and survey frequency

Abstract

Long‐term monitoring is key for detecting population declines. Composite indices allow researchers to combine trends from disparate monitoring programmes into a single estimate of population change. Inferences from composite indices, however, are limited to the time periods and areas studied. We show that the number of breeding pairs in a colony of Cape Vultures Gyps coprotheres in South Africa declined between the years 1983 and 2003 but increased in both the number of breeding pairs and fledglings per pair from 2004 to 2017. We performed a retrospective power analysis determining the minimum annual frequency with which the colony could have been monitored without sacrificing inference into population trends. This power analysis revealed the post‐2003 population increase would not have been apparent if we skipped more than 2 years between surveys. We incorporated our estimates into a previously published composite index for Cape Vultures that considered only data collected pre‐2000 and demonstrate that the inference is unchanged if the trend from the decline period or the entire study is incorporated, yet if only the trend during the period of population increase is used, there is no longer a statistically significant decline across the species’ range. Our results demonstrate the utility of long‐term monitoring because if our study had concluded before 2003, there would be little evidence of the current population increase at Kransberg.