Road Ecology Center

The barrier effect imposed by roads and railways on vertebrate populations has aroused both scientific and social concern and has led to the construction of crossing structures for such fauna in new infrastructures. Good practice demands that investment in such mitigation measures should be followed by systematic monitoring of their effectiveness, in order to improve the design of further works. These monitoring schemes need standardized protocols in order to deliver scientifically sound results at an affordable cost. In this context, the present contribution analyzes the suitability of monitoring schemes aimed at determining which vertebrate species use crossing structures in relation to the number of days spent monitoring each crossing structure. The analysis considers data on vertebrates using 22 structures crossing a motorway in northwest Spain, which were monitored for 15-26 consecutive days. Species accumulation curves were fitted by non-linear estimation procedures to the species accumulation pattern detected at each crossing structure in order to estimate the asymptotic number of species using each one of them. Modelling was carried out using 11 functions applied in ecological studies to analyze species accumulation curves in relation to sampling intensity. The results show that species accumulation curves for crossing structures have a rapid increase phase followed by a long tail of slow accumulation. Thus, 25 or more monitoring days may be needed to detect over 80 percent of the species using a crossing structure, but 60 percent of them are detected by day 10, and 70 percent, by day 16. The statistical fit obtained for different function types allows the Clench model to be recommended for evaluating the results obtained in monitoring programs intended to determine the number of species using each crossing structure. This model yielded the highest mean explanatory power (mean r2=0.905) using only two parameters; it provided neither a systematic overestimate nor an underestimate of richness, and offered a low degree of uncertainty (2.3% non-significant parameters). In short, 10 to 15 days of monitoring may be enough to provide a basic knowledge of the animal species using crossing structures at a particular time, although the monitoring period could be somewhat shorter or longer according to the requirements of particular cases.