Measuring connectivity in habitat mosaics: the equivalence of two existing network indices and progress beyond them

Network-based methods are being actively developed to respond to the needs for operational assessments of the degree of landscape connectivity and of the impact of landscape changes in ecological flows and related ecological processes. Among these, a recent paper by Matisziw and Murray (2009) presented the C index as an adequate and advantageous way of ranking habitat patches by their importance for the maintenance of landscape connectivity. We show that this index is equivalent and conveys the same information in undirected graphs as a previously described index, the landscape coincidence probability (LCP), which can be readily computed in any landscape network through the Conefor Sensinode software package. We slightly generalize the LCP definition for cases involving asymmetric dispersal, which makes LCP compatible with C and maintains the equivalency between both indices in directed graphs. We place LCP and C in a broader context of other existing indices and ongoing developments and describe how some of these may be better suited for the analysis of the connectivity in landscape networks and their changes. We conclude by highlighting the need (1) to go beyond the identification of unobstructed movement paths or habitat components (sets of interconnected patches) when pursuing the most appropriate landscape connectivity indices and (2) for increased efforts in assessing and reporting the potential overlaps, coincidences and synergies between the available approaches in order to guide the final user and facilitate index selection in a densely populated metric space