ABSTRACT
In general, structural connectivity measures may be calculated simply by examining a map. One of the simplest measures of patch-level structural connectivity is nearestneighbor distance, which is the Euclidean distance to the nearest patch.[2] This measure was shown to be a useful predictor of the incidence of a frugivorous beetle in forest fragments.[38] This simple measure can be infused with more ecological information by measuring distance to the nearest occupied patch rather than distance to any patch,[2] but the more comprehensive measure is not always an improvement on the simpler one.[38] Another common way to measure patch-level structural connectivity is to examine the amount of habitat in a buffer around the focal patch. This measure was found to be much better than nearestneighbor distance for predicting colonization events in two butterfl y species but was very sensitive to buffer size.[32] At the landscape level, structural connectivity can be measured, for example, based on percolation or lacunarity. In these approaches, the landscape is considered as a twodimensional grid where each cell in the grid is classifi ed as either habitat or non-habitat. Landscape connectivity is then measured as the physical connection of habitat, in the case of percolation,[39] or as the variability and size of interpatch distances, in the case of lacunarity.[40] The lacunarity index was shown to be a good predictor of dispersal success for simulated organisms in fractal landscapes.[41]
Because functional connectivity is specifi c to the species or process of interest, it is necessary to know something about species’ movement behavior to measure functional connectivity. Species may perceive and respond to landscape pattern differently according to their dispersal characteristics, their preferred habitat type, or other life history traits.[26] For example, seed-dispersing birds differ in their presence in remnant trees within the matrix based on their frugivory levels. Fruiting trees could represent stepping stones across the matrix for birds with a completely fruit-based
La nd
— M
ar sh
es
Actual connectivity measures are based on empirical, often spatially explicit, information about the movement of a particular organism. At the patch level, actual connectivity can be measured as the number of immigrants into a patch.[48] At the landscape level, actual connectivity may be measured as the number of patches visited by an organism or movement rates across the landscape.[49,50] However, these approaches can present a paradox in that animals may move more frequently between patches in lower-quality habitat, counterintuitively resulting in higher connectivity indices in these less-desirable environments.[51] Another approach to measuring actual connectivity draws heavily from behavioral ecology. For instance, Bélisle[51] suggests several different experimental methods for determining the motivation underlying movement of individuals through the landscape, including translocations, playback experiments, and measuring giving-up densities. Playback and homing experiments have been used to study the effect of roads[52] and other barriers[53] on animal movement. More recently, playback techniques have been used to parameterize graph theory models (described below) to explain the occurrence pattern of an Atlantic rainforest bird.[54]
Functional connectivity can also be inferred from genetic information. Dispersal infl uences gene fl ow between subpopulations,[55,56] which results in genetic differences among organisms occupying different parts of the landscape. Landscape genetics is a rapidly growing fi eld and will likely continue to make large contributions to measuring and understanding the consequences of landscape connectivity.[57]
Graph theory, also called network analysis, is a fl exible method for measuring landscape connectivity that has gained traction over the last decade. A graph is a set of nodes connected by links, where a link between nodes indicates a connection between them. In landscape ecology, the nodes typically represent habitat patches, and links indicate dispersal between patches.[58] Commonly used
diet, but may not for birds with mixed diets.[42] Functional connectivity measures are usually considered to be superior to structural connectivity measures.
