Spatio‐temporal variation in European starling reproductive success at multiple small spatial scales. Issue 16 (22nd July 2015)

• Record Type: Journal Article
• Authors: Brickhill, Daisy; Evans, Peter G.H.; Reid, Jane M.
• Is Part Of: Ecology and evolution. Volume 5:Issue 16(2015:Sep.)
• Page Start: 3364
• Page End: 3377
• Publication Date: 2015-07-22
• Subjects: Ecology -- Periodicals; Evolution -- Periodicals; 577.05
• DOI: 10.1002/ece3.1615
• Languages: English
• ISSNs: 2045-7758

Abstract

<abstract abstract-type="main" id="ece31615-abs-0001"> <title>Abstract</title> <p>Understanding population dynamics requires spatio‐temporal variation in demography to be measured across appropriate spatial and temporal scales. However, the most appropriate spatial scale(s) may not be obvious, few datasets cover sufficient time periods, and key demographic rates are often incompletely measured. Consequently, it is often assumed that demography will be spatially homogeneous within populations that lack obvious subdivision. Here, we quantify small‐scale spatial and temporal variation in a key demographic rate, reproductive success (RS), within an apparently contiguous population of European starlings. We used hierarchical cluster analysis to define spatial clusters of nest sites at multiple small spatial scales and long‐term data to test the hypothesis that small‐scale spatio‐temporal variation in RS occurred. RS was measured as the number of chicks alive ca. 12 days posthatch either per first brood or per nest site per breeding season (thereby incorporating multiple breeding attempts). First brood RS varied substantially among spatial clusters and years. Furthermore, the pattern of spatial variation was stable across years; some nest clusters consistently produced more chicks than others. Total seasonal RS also varied substantially among spatial clusters and years. However, the magnitude of variation was much larger and the pattern of spatial variation was no longer temporally consistent. Furthermore, the estimated magnitude of spatial variation in RS was greater at smaller spatial scales. We thereby demonstrate substantial spatial, temporal, and spatio‐temporal variation in RS occurring at very small spatial scales. We show that the estimated magnitude of this variation depended on spatial scale and that spatio‐temporal variation would not have been detected if season‐long RS had not been measured. Such small‐scale spatio‐temporal variation should be incorporated into empirical and theoretical treatments of population dynamics.</p> </abstract>