Ecology of the European hare in a farmland area of Northern Italy

Since the mid 1960s, the number of European hares Lepus europaeus in Italy declined, as well as in the whole Europe, possibly because of habitat changes, agriculture intensification, urbanization, and diseases [1]. The present work describes several aspects of the ecology of a wild population of European hare in a farmland area of Northern Italy: population dynamics, habitat selection, survival and mortality. The study was carried out in the province of Piacenza, in a protected area (“Borgonovo Val Tidone”, 909 hectares) dominated by arable lands (about 85% of the total surface). Hare density at the study site has estimated since 2012 by spotlight counts, carried out each year in late winter, when the detectability of hares is highest, and by Distance Sampling [2]. Radio telemetry has used to follow the fate of 39 hares (35 adults and 4 juveniles) since January 2015. A kernel analysis was used to produce home-range estimates. We analyzed habitat selection of Johnson's second and third order [3] using permutation based combination of sign tests by the R package "phuassess" [4, 5]. Moreover we analyzed habitat selection for resting places comparing micro-habitat characteristics of forms to those of an equal number of random points selected in the same habitat types. We tested for the differences between forms and controls by The Wilcoxon test and by a Conditional Logistic Regression Analysis (CLRA). Survival was calculated using the Kaplan-Meier function, and the Mantel-Cox test was used to detect the existence of differences between sexes. Cox regression was used to individuate the factors affecting the death probability of hares. The best detection function was the hazard rate (AIC=4137.98, ESW=97.2, CV=4.3); pre-reproductive density declined over the study period, halving from 2012 (37 hares per Km2, SE=0.5) to 2016 (17 hares per Km2, SE=0.5). At the second order of selection the habitat use was significantly different from a random one (P=0.0001); in particular alfalfa, and unpaved roads were selected while fallows and ploughed fields were avoided. Also at the third order of selection habitat use was significantly different from a random one (P=0.0003); urban areas and unpaved roads were significantly avoided, while the other habitats were used as available. Hares showed a preference for cover at their resting places throughout the year, selecting fields in the good season, and hedgerows in the limiting one, when open fields are characterized by little or no vegetation, and do not offer an adequate cover. The forms in respect to the control points were characterized by a greater cover of the scrub layer, higher herbaceous layer, and higher vegetation. The CLRA showed a positive and significant effect of the height of the dominant herbaceous species on the form probability and a negative one of the brightness (Table 1). Average survival of radio-tagged hares was of over 300 days, which is among the highest values recorded for the species. We did not detect differences between males and females (Mantel-Cox test, P = 0.72) (Fig. 1). Eighteen hares survived over 250 days. Among mortality causes, we found both poaching and predation, but, in the majority of cases, the advanced state of decomposition of carcasses did not allow us to determine the cause of death. The Cox regression showed positive effects on death probability of the percentage of alfalfa and maize in the home-ranges and a negative effect of the home-range size (Table 2). The low number of juveniles captured, together with the high survival rate of adults, lead us to hypothesize that the decline of this population may be due to a high juvenile mortality, which determines a low recruitment rate. The loss of cover habitats could further contribute to the decline of populations living in intensively cultivated areas.