Data from: Isolating the role of the matrix at patch and landscape scales

Global increases in habitat loss and fragmentation have resulted in
non-habitat land cover, or the matrix, becoming an increasingly prominent
feature of landscapes. The matrix can influence the population dynamics of
species in fragments by modifying processes operating locally on
individual patches (e.g., edge effects on survival) or at landscape scales
(e.g., inter-patch dispersal). However, the relative magnitude of patch-
vs. landscape-scale matrix effects on the populations found in patches
remains unclear. We established 12 experimental landscapes in which we
controlled for habitat amount and fragmentation while manipulating the
quality of the matrix around (i) individual habitat patches and (ii)
across the entire landscape in a factorial design. We then compared the
magnitude of local- and landscape-scale matrix effects on a specialist
herbivore, Chelinidea vittiger (Hemiptera: Coreidae). Population size in
fragments was influenced by both patch- and landscape-scale treatments:
population size increased in patches surrounded by high-quality matrix,
but only in landscapes dominated by low-quality matrix, due in part to
decreased inter-patch movements in these landscapes. In contrast, the
effects on both survival and reproductive output were solely at the
patch-scale, with both lower in patches surrounded by low-quality matrix.
Our results underscore the outsized importance of matrix habitat
immediately adjacent to fragment edges -- even though patch-scale
manipulations affected only a fraction (3%) of the area that
landscape-scale manipulations did, patch-scale effects were more common.
The relationship between dispersal, population size, and scale-dependent
effects of matrix quality emphasizes the need to explicitly consider the
spatial scale at which different processes operate when predicting
responses to habitat fragmentation. Our results also suggest that the
matrix immediately adjacent to habitat remnants is of particular
importance when considering alternative strategies for landscape
conservation or restoration.