Study area (Gravina in Puglia and Alta Murgia National Park; Apulia, Italy), nests and roosts. The study area corresponds to the SPA (Special Protection Area) “Murgia Alta” IT9120007 and is included within the IBA (Important Bird Area) “Murge”. 

Context in source publication

Context 1

... provision of suitable nesting sites near the higher quality areas; d) the optimal timing for harvesting. Our study is the first contribution to the assessment of the foraging strategies and the necessary conservation policies of the Lesser Kestrel in Southern Italy. Keywords Alta Murgia National Park; data-loggers; foraging movements; hatching period; sexual divergences; special protection area. Resource selection studies are commonly conducted because it is generally assumed that if animals select habitat disproportionately to their availability, that habitat improves their fi tness, reproduction and survival (Stephens and Krebs, 1986). From foraging theory strategy, it’s also known that if land cover change causes impoverishment and/or loss of preferred hunting habitats, a species would obtain a lesser hunting yields, with direct implications for its conservation. Changes in land-use, both with concurrent aspects of global change, have a strong impact on the structure of biological communities (Gil-Tena et al., 2009). Many species of conservation interest in Europe are considered associated with traditional farm landscapes and the semi-natural habitats they produce and maintain (Tucker and Evans, 1997). Land abandonment has been an important land-use change in recent decades (Ostermann, 1998). The decrease in farming mainly affected the least productive agricultural land, and activated the recovery of semi-natural vegetation (Sirami et al., 2008). In most of the Mediterranean region, land abandonment has occurred during the last century, leading to the naturalization and vegetation closure of many areas, thus favouring the spread of forests (Debussche et al., 1999). This caused a decrease of open grassland-like habitats and an increase in shrubland and, on the long-term, woodland cover (Romero- Calcerrada and Perry, 2004), thus determining a decline of species tied to open habitats (Suárez-Seoane et al., 2002; Sirami et al., 2007), in particular migrant species associated with open farmland habitats (Sirami et al., 2008). On the other hand, agricultural intensi fi cation and abandonment of traditional farming had dramatic impacts on farmland birds (Donald et al., 2001), in particular on the quality of foraging patches and food availability (Donázar et al., 1993), thus affecting species’ fi tness components such as the number of offspring that parents are able to raise (Tella et al., 1998). In the past, the reduction in quality and extent of foraging habitats has been the primary cause of decline for Lesser Kestrel (Negro, 1997; Peet and Gallo-Orsi, 2000). Extensive cereal fields, fallows, pasturelands and field margins in agricultural areas are usually considered the main habitats used by this species for foraging (Cramp and Simmons, 1980; Donázar et al., 1993; Tella et al., 1998). Arthropod abundance is usually higher in these types of land-use (Martínez, 1994; Moreira, 1999). On the other hand, for hunters such as the Lesser Kestrel, access to prey is also affected by vegetation structure (Shrubb, 1980; Toland, 1987), in particular by land cover offering shelter to prey, and height which obstructs hunting manoeuvres. This may explain why they usually avoid hunting in habitat patches with taller vegetation cover, such as abandoned crop fields or shrublands (Tella et al., 1998). In addition, the species has declined markedly in the last decades also because of agricultural intensification and pesticide use, which affected their foraging habitats and food availability (Parr et al., 1995; Bustamante, 1997; Tella et al., 1998; BirdLife International, 2004). Despite the urgent need for the conservation of this species, at present little is known about foraging habitat selection of Lesser Kestrels in Italy (Sarà 2010). Due to this reason, in this paper we investigate the patterns of land-cover type selection of Lesser Kestrels in the main colony in Italy (Alta Murgia National Park, Gravina in Puglia and the surrounding rural areas) during the hatching period. The goals of our work were to individuate preferred foraging habitats within and outside the Alta Murgia National Park, and explore the hypothesis of potential foraging divergences with regard to sex. In fact, sexual differences in foraging habitat selection can be hypothesized to arise as a consequence of two necessities for females during the hatching period, i.e. spending as much time as possible in parental care and limiting energy requirements for foraging movements. No studies focus on this topic for Lesser Kestrels yet, but sexual divergences in foraging selection might have important consequences on conservation strategies for this species. We also explored if the relationship between the Lesser Kestrel’s foraging movements and the spatial arrangement of habitats may influence the foraging habitat selection. Based on our results, several management policies are proposed for the conservation of this important species in Italy. The study area (Fig. 1) corresponds to the Alta Murgia National Park and the SPA (Special Protection Area) “Murgia Alta” IT9120007 (Apulia, Southern Italy) and is included within the IBA (Important Bird Area) “Murge” (Heath and Evans, 2000). It comprehends the main colony of Lesser Kestrels in Italy (Bux et al. 2008, Gustin et al., 2014), i.e. the town of Gravina in Puglia and the surrounding rural areas. Lesser Kestrel is a migratory, colonial, small (body length 29–32 cm, wingspan 58–72 cm) falcon breeding mainly in holes and crevices in large historic buildings within towns and villages, or often in abandoned farm houses scattered across the countryside (Negro, 1997). The Lesser Kestrel is primarily insectivorous, feeding mainly on beetles, myriapods and grasshoppers (Franco and Andrada, 1977; Kok et al., 2000). It inhabits steppe-like ecosystems around the Mediterranean and central Asia. In Western Europe it is mainly a summer visitor, migrating to Africa in winter (Rodríguez et al., 2009). Today Lesser Kestrel is considered a “least concern” species (BirdLife International, 2013; Gustin et al., 2014). Nine individuals (4 males and 5 females) were surveyed in a period of 20 days from June 20 to July 9 2012 in the colony of Gravina in Puglia. Surveys were conducted using TechnoSmart GiPSy-4 data-loggers (backpack harness; 23x15x6 mm; total weight: 1.8 g plus 3.2 g battery), that provided for each GPS point information about date (dd/mm/yyyy), local time (hh:mm:ss), latitude (degrees-minutes-seconds), longitude (degrees-minutes-seconds), altitude (meters above mean sea level) and instantaneous speed (km/h). Data acquisition occurred every 5 minutes during two time periods: day (08:00-19:00 H local time) and night (02:00-06:00 H local time). In situ surveys allowed us to locate nests and roosts used by the observed individuals. Birds were captured at their nest boxes when they were delivering food to their nestlings and fitted with data-loggers. To download the data from the data-loggers, birds were recaptured at their nest boxes after batteries were exhausted three days later. GPS data were imported into the GIS GRASS (Neteler and Mitasova, 2008). Layers used for the subsequent analyses were: a) boundaries of the Alta Murgia National Park, b) land cover at 1:10,000 scale (provided by the Apulia Region), c) digital terrain model (DTM) of the study area (digitized at 1:10,000 scale by the authors from available topographic maps of Apulia Region), d) nest and roost locations. We estimated home-ranges using fi xed kernel estimators (Worton, 1989) at 95% isopleth, which were calculated with least-squares cross-validation and adjusted to extreme locations (Worton, 1989). The 95% isopleth (HR95 from now on) is most widely used in the literature and represented the full home range. Foraging points (FP) have been individuated using two steps. First, for each GPS point we achieved flight height above ground level (a.g.l. hereafter) by subtracting terrain elevation (indicated by DTM) from altitude a.s.l. (provided by data-loggers). Second, among the GPS points having flight height a.g.l. equal to 0, we chose only those ones having an instantaneous speed (provided by GPS) equal to 0. We privileged this conservative approach rather than using also GPS points with low instantaneous speed (e.g., less than 1 or 2 km/h) because we preferred to miss some FP rather than being at risk of including also non-foraging points. These two steps allowed us to detect locations of the study area where Lesser Kestrels remained motionless at ground level (i.e., instantaneous speed and flight height a.g.l. equal to 0). Detected FP thence represented strike attempts (i.e., strikes in which the bird landed on the ground), not necessarily successful captures. For the purposes of this work we considered that strike attempts were a type of foraging habitat selection. Foraging habitat selection was investigated at the following levels: a) FP as compared to habitat availability in HR95; b) male FP as compared to habitat availability in HR95; c) female FP as compared to habitat availability in HR95. Thomas and Taylor (1990) distinguished three types of use-availability design used in the studies of habitat and resource selection. In design I studies, the animals are not identi fi ed; the habitat use and availability are measured at the scale of the population. In design II studies, the animals are identi fi ed and the use is measured for each one, however, the availability is measured at the scale of population. In design III studies, the animals are identi fi ed and both the use and the availability are measured for each one. The choice of the proper use- availability design can be evaluated only in reference to a specific data set and a specified model (Hurlbert, 1984). In order to decide which type of design to use, we applied the pairwise test of multiple associations (Janson and Vegelius, 1981; Ludwig and Reynolds, 1988). The pairwise test is based ...