Introduction
Migratory birds, in particular long-distance migrants, are vulnerable to environmental change in multiple regions (Sanderson et al., Reference Sanderson, Donald, Pain, Burfield and van Bommel2006; Both et al., Reference Both, Van Turnhout, Bijlsma, Siepel, Van Strien and Foppen2010). The ortolan bunting Emberiza hortulana is the only long-distance trans-Saharan migrant among old world buntings (Cramp & Perrins, Reference Cramp and Perrins1994; Glutz von Blotzheim & Bauer, Reference Glutz von Blotzheim, Bauer and Glutz von Blotzheim1997). The species has undergone the second most pronounced decline of any bird species in temperate Western Europe in recent decades, with an estimated 82% population reduction between 1980 and 2008 (Klvanova et al., Reference Klvanova, Skorpilova, Vorisek, Gregory and Burfield2010), although the decline began earlier in some places (Conrads, Reference Conrads1977; Lang et al., Reference Lang, Bandorf, Dornberger, Klein and Mattern1990; Meier-Peithmann, Reference Meier-Peithmann1992; Dale, Reference Dale1997). Ortolan bunting populations have recently crashed across northern Europe and Scandinavia (van Noorden, Reference van Noorden1991, Reference van Noorden1999; Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005; Ottvall et al., Reference Ottvall, Green, Lindström, Svensson, Esseen and Marklund2008) and the species has effectively become extinct as a breeding species within the last decade in Belgium, The Netherlands (van Noorden, Reference van Noorden1991, Reference van Noorden1999; Vieuxtemps & Jacob, Reference Vieuxtemps and Jacob2002; van Dijk et al., Reference van Dijk, Hustings, Koffijberg, van der Weide, Deuzeman, Zoetebier and Plate2005) and Switzerland (Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005; Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b), with mostly unpaired singing males remaining in these populations. The species has apparently remained stable in Eastern Europe (BirdLife International, 2004), and the only notable increase has been in Catalonia, in the Mediterranean (Pons, Reference Pons, Estrada, Pedrocchi, Brotons and Herrando2004; Brotons et al., Reference Brotons, Herrando and Pons2008).
Although the life history of the ortolan bunting is generally well resolved we lack consolidated information about the species’ key ecological requirements and most conservation action for the species is based on expert opinion rather than scientific evidence. Thus, there is uncertainty about the optimal conservation measures to implement. Given the challenges of integrating research programmes across regions and countries clear direction is required for appropriate conservation research for the ortolan bunting. In this review we: (1) synthesize existing knowledge of the biology of the ortolan bunting, (2) discuss the proposed causes of the species’ decline, (3) propose priorities for future research to inform conservation action, and (4) provide preliminary evidence-based management recommendations from the information currently available (Pullin & Knight, 2001).
Literature searches were primarily on the ISI Web of Science, the Ornithological Worldwide Literature database (OWL, 2010), and reference lists from published articles. The review of threatening processes considered articles published after 1950 as this is believed to be the year in which many population declines began (Lang et al., Reference Lang, Bandorf, Dornberger, Klein and Mattern1990; Meier-Peithmann, Reference Meier-Peithmann1992; Dale, Reference Dale1997).
Ecology of the ortolan bunting
Habitat requirements
In Mediterranean and sub-Mediterranean Europe the species breeds primarily in open shrubland and steppe-like habitat, particularly on south-facing slopes (Cramp & Perrins, Reference Cramp and Perrins1994; Glutz von Blotzheim & Bauer, Reference Glutz von Blotzheim, Bauer and Glutz von Blotzheim1997; Fonderflick et al., 2005; Brotons et al., Reference Brotons, Herrando and Pons2008). Here, the species favours areas with shrub and tree cover of c. 20–30% (Kölsch, Reference Kölsch1959; Keusch, Reference Keusch1991; Menz et al., Reference Menz, Brotons and Arlettaz2009a) and rarely occurs where tree cover exceeds 30–50% (Fonderflick et al., Reference Fonderflick, Thévenot and Guillaume2005; Fonderflick, Reference Fonderflick2006). In temperate Europe the species breeds primarily in agricultural habitats, particularly areas of small-scale cultivation, set-asides, short-rotation coppice and shrublands in historically burnt habitats, with these habitats often co-occurring (e.g. Berg & Pärt, 1994; Dale & Hagen, Reference Dale and Hagen1997; Berg, Reference Berg2002; Dale & Olsen, Reference Dale and Olsen2002; Goławski & Dombrowski, Reference Goławski and Dombrowski2002; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005). In farmland the species favours field margins with structural elements such as isolated trees, hedges and nearby forest margins (Meier-Peithmann, Reference Meier-Peithmann1992; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002), a characteristic shared by several farmland bunting species (Brambilla et al., Reference Brambilla, Guidali and Negri2008, Reference Brambilla, Guidali and Negri2009).
Within both natural and agricultural landscapes the ortolan bunting breeds primarily in relatively warm, dry areas, with well-drained soils and an annual rainfall below 600–700 mm (Cramp & Perrins, Reference Cramp and Perrins1994; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002), avoiding wet habitats (e.g. Nævra, Reference Nævra2002; Dale & Manceau, Reference Dale and Manceau2003; Hänel, Reference Hänel2004; Deutsch, Reference Deutsch2007). Exceptions include populations occurring in areas with extremely well-drained soils and steep, sloping topography (Conrads, Reference Conrads1977). Ortolan buntings nest on the ground, typically producing only one brood per season, with exceptional replacement clutches and second broods (Garling, Reference Garling1943; Conrads, Reference Conrads1969; Hänel, Reference Hänel2004).
Ortolan bunting populations typically consist of loose aggregations of breeding pairs (Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2007). Local colonization, extinction, and population fluctuations are often observed (Glitz, Reference Glitz1967; Dale & Steifetten, Reference Dale and Steifetten2011), with areas seemingly isolated from other populations also colonized (van Noorden, Reference van Noorden1991, Reference van Noorden1999; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005). This could indicate the existence of a broadscale metapopulation structure, with areas being settled or abandoned as habitat suitability fluctuates following major disturbance events (Brotons et al., Reference Brotons, Pons and Herrando2005). Short-term population increases have been observed in response to fire (Brotons et al., Reference Brotons, Pons and Herrando2005, Reference Brotons, Herrando and Pons2008), clearing of vegetation by a storm, forestry interventions, or cultivation (Conrads & Kipp, Reference Conrads and Kipp1980; Nævra, Reference Nævra2002). Sparse vegetation and a large proportion of bare ground are the most noticeable common features of these habitats (Nævra, Reference Nævra2002). The species often becomes locally common after disturbance (Sposimo, Reference Sposimo1988; Pons, Reference Pons, Estrada, Pedrocchi, Brotons and Herrando2004; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005) with, for example, populations peaking 3–4 years after fire (Pons & Clavero, Reference Pons and Clavero2010). This relationship is particularly strong in Mediterranean and, to a lesser extent, sub-Mediterranean biomes, where occurrence of fires is still commonplace, and it is also noticeable in temperate Europe where the species also nests on historic burns (e.g. Dale & Olsen, Reference Dale and Olsen2002; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005). This indicates that the ortolan bunting behaves like a pioneer species, typically colonizing the early stages of vegetation succession.
Diet
The ortolan bunting has a varied diet, including both plant (seeds) and animal matter (Cramp & Perrins, Reference Cramp and Perrins1994; Glutz von Blotzheim & Bauer, Reference Glutz von Blotzheim, Bauer and Glutz von Blotzheim1997), although the diet of the chicks is restricted to a few dominant prey orders: Lepidoptera, particularly Tortricidae larvae, and Coleoptera in the north of its range (Conrads, Reference Conrads1968, Reference Conrads1969; Hänel, Reference Hänel2004), and Orthoptera, particularly Tettigoniidae, in the south (Kunz, Reference Kunz1950; Keusch & Mosimann, Reference Keusch and Mosimann1984). In Switzerland, Tettigoniidae made up nearly 70% of the total items provisioned to nestlings, a much higher percentage than in the sympatric rock bunting Emberiza cia, which has a more diverse diet (Keusch & Mosimann, Reference Keusch and Mosimann1984). In the north of the range caterpillars are fed to nestlings in the early stages of development, with diet switching towards larger prey in later developmental stages until post-fledging (Kunze, Reference Kunze1954; Knoblauch, Reference Knoblauch1968; Conrads, Reference Conrads1969; Hänel, Reference Hänel2004).
Foraging ecology
Ortolan buntings forage primarily in patches of bare ground within sparsely vegetated habitats (Stolt, Reference Stolt1974; Gnielka, Reference Gnielka1987; Boitier, Reference Boitier2001; Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b). However, prey such as caterpillars are also collected from fields (Conrads, Reference Conrads1969), or gleaned from tree crown foliage, particularly oaks Quercus spp. (Knoblauch, Reference Knoblauch1968; Conrads, Reference Conrads1969; Stolt, Reference Stolt1974; Gnielka, Reference Gnielka1987), which harbour a relatively high density of caterpillars compared to other tree species (Naef-Daenzer, Reference Naef-Daenzer2000). Adult males sometimes forage in the same oaks used as song posts (Hänel, Reference Hänel2004). Consequently, song post selection may function as a signal of territory quality, as oaks appear to be over-represented as song posts, compared to local availability of other tree species (M.H.M. Menz, pers. obs.).
In Switzerland most Tettigoniidae fed to nestlings are captured on the ground (Keusch & Mosimann, Reference Keusch and Mosimann1984) and in Germany Coleoptera are collected on paths or in cereal fields (Knoblauch, Reference Knoblauch1968). Tettigoniidae are most abundant in relatively dense steppe grass or bushes (Arlettaz et al., Reference Arlettaz, Perrin and Hausser1997). However, ortolan buntings do not necessarily forage in habitats with highest prey abundance but rather in those with a high proportion of bare ground (Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b), as observed in other ground foraging birds (Wilson et al., Reference Wilson, Whittingham and Bradbury2005; Schaub et al., Reference Schaub, Martinez, Tagmann-Ioset, Weisshaupt, Maurer and Reichlin2010). Prey accessibility, therefore, rather than abundance, drives foraging habitat selection (Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b; Schaub et al., Reference Schaub, Martinez, Tagmann-Ioset, Weisshaupt, Maurer and Reichlin2010).
In temperate Europe foraging often takes place in cultivated fields, sometimes a distance away from breeding areas (Dale, Reference Dale2000; Dale & Olsen, Reference Dale and Olsen2002). Cereal fields, particularly oats, are important for replenishing body fat prior to and upon return from migration, when birds feed on seeds and sprouting plants (Keusch, Reference Keusch1991; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002). Oat is probably favoured because of the high energy content of the grain (Glutz von Blotzheim, Reference Glutz von Blotzheim1989; Diaz, Reference Diaz1990).
Threats and reasons for decline
Habitat loss and degradation, and reduction in prey availability
Agricultural intensification has resulted in land-use changes such as homogenization of agricultural landscapes, loss of structural heterogeneity and an increased use of pesticides (Newton, Reference Newton2004). A reduction of crop diversity and the transition in cultivation from summer to winter cereals may have contributed to the decline of the ortolan bunting, as such changes will limit the amount of bare ground in cultivated fields. Conversion of rye, and especially oat, to maize cultures has been reported to affect the species negatively (Maes et al., Reference Maes, Gabriëls, Geuens and Meeus1985; Ikemeyer & von Bülow, Reference Ikemeyer and von Bülow1995; van Noorden, Reference van Noorden1999; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002; Deutsch, Reference Deutsch2007; Berg, Reference Berg2008), although this has not been quantified. Creation of monoculture agricultural habitats by destruction of structural habitat elements such as tree lines and hedges may be detrimental, as these provide song posts and foraging opportunities (Knoblauch, Reference Knoblauch1968; Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005). Urbanization has often encroached into small-scale agricultural areas that had typically been preserved close to villages (van Noorden, Reference van Noorden1999), increasing disturbance near to breeding areas (Steiner & Hüni-Luft, Reference Steiner and Hüni-Luft1971).
One of the principal reasons for the observed decline of the ortolan bunting in temperate Europe is probably a reduction in prey availability/accessibility driven by habitat deterioration on the breeding grounds, primarily via changes in agricultural practices (Claessens, Reference Claessens1992; Kutzenberger, Reference Kutzenberger, Tucker and Heath1994; van Noorden, Reference van Noorden1999; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005; Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005; Deutsch, Reference Deutsch2007). A reduction in patches of bare ground within foraging areas will result in decreased prey availability (Wilson et al., Reference Wilson, Whittingham and Bradbury2005; Schaub et al., Reference Schaub, Martinez, Tagmann-Ioset, Weisshaupt, Maurer and Reichlin2010) in two ways. Firstly, agricultural intensification includes increased application of fertilizers, which closes the vegetation and suppresses patches of bare ground, and the use of pesticides eliminates invertebrate prey. Secondly, areas of bare ground also vanish following vegetation encroachment through natural succession after abandonment of traditional agricultural practices such as extensive grazing and burning of dry grass (Stolt, Reference Stolt1974; Dale, Reference Dale1997; Nævra, Reference Nævra2002; Revaz et al., Reference Revaz, Posse, Gerber, Sierro and Arlettaz2005; Wilson et al., Reference Wilson, Whittingham and Bradbury2005; Sirami et al., Reference Sirami, Brotons and Martin2007; Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b; de Groot et al., Reference de Groot, Kmecl, Figelj, Figelj, Mihelič and Rubinić2010). There is increasing evidence that reduction in structural heterogeneity and bare ground is threatening a number of ground-foraging farmland bird species (Wilson et al., Reference Wilson, Whittingham and Bradbury2005; Schaub et al., Reference Schaub, Martinez, Tagmann-Ioset, Weisshaupt, Maurer and Reichlin2010).
Climate change on the breeding grounds
Climate change has been proposed as a possible cause of the decline of the ortolan bunting (Knoblauch, Reference Knoblauch1954; Helb, Reference Helb1974; Maes et al., Reference Maes, Gabriëls, Geuens and Meeus1985; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002; Vieuxtemps & Jacob, Reference Vieuxtemps and Jacob2002). A low tolerance to cold temperatures (Wallgren, Reference Wallgren1952, Reference Wallgren1954) may increase the risk of physiological stress on breeding birds during inclement weather. Microclimate at the nest site is also likely to have consequences for the growth and survival of the nestlings (Conrads, Reference Conrads1977; Lang et al., Reference Lang, Bandorf, Dornberger, Klein and Mattern1990; Dale, Reference Dale2000; Dale & Olsen, Reference Dale and Olsen2002; Grützmann et al., Reference Grützmann, Moritz, Südbeck and Wendt2002; Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005), as observed in other bird species (Ullrich, Reference Ullrich1971). Poor weather during the breeding season, such as cold, rainy springs, also lowers reproductive success through nestling mortality from food limitation (Ruge et al., Reference Ruge, Pflüger, Hölzinger, Labus, Gatter and Schmidt1970; Fonderflick & Thévenot, Reference Fonderflick and Thévenot2002), something also observed in other species (Arlettaz et al., Reference Arlettaz, Schaad, Reichlin and Schaub2010).
The earlier growing season predicted under some climate change scenarios may lead to reduced foraging opportunities for ortolan buntings returning to their breeding grounds because of vegetation closure (Lang, Reference Lang2007). The relatively short nestling phase (9–14 days; Cramp & Perrins, Reference Cramp and Perrins1994; Glutz von Blotzheim & Bauer, Reference Glutz von Blotzheim, Bauer and Glutz von Blotzheim1997) means chicks require a large amount of invertebrate prey in a short period (Meier-Peithmann, Reference Meier-Peithmann1992). Ortolan buntings may have evolved a reproductive phenology to coincide with peak prey availability, as seen in some other bird and mammal species (Blondel et al., Reference Blondel, Dervieux, Maistre and Perret1991; Arlettaz & Fournier, Reference Arlettaz and Fournier1993; van Noordwijk et al., Reference van Noordwijk, McCleery and Perrins1995; Arlettaz et al., Reference Arlettaz, Perrin and Hausser1997, Reference Arlettaz, Christe, Lugon, Perrin and Vogel2001). Chicks usually hatch in c. mid June, when Tettigoniidae are abundant and at a profitable size (Kunz, Reference Kunz1950; Arlettaz et al., Reference Arlettaz, Christe, Lugon, Perrin and Vogel2001). Conrads (Reference Conrads1968, Reference Conrads1977) noted breeding was synchronous with sprouting of oak leaves and the appearance of large numbers of defoliating caterpillars, although this was not quantified. Given the short breeding season and nestling phase, and their relatively late return from Africa (Claverie, Reference Claverie1955), altered climate regimes may see a mismatch between breeding period and prey availability (Both et al., Reference Both, Van Turnhout, Bijlsma, Siepel, Van Strien and Foppen2010).
Altered population structure and dynamics
Studies from multiple regions have reported 29–60% of singing ortolan bunting males remain unpaired during the breeding season (Conrads, Reference Conrads1968; Dale, Reference Dale2001; Fonderflick & Thévenot, Reference Fonderflick and Thévenot2002; Steifetten & Dale, Reference Steifetten and Dale2006; Berg, Reference Berg2008). However, even in declining and fluctuating populations breeding success appears to remain stable (Maes, Reference Maes1989; Steifetten & Dale, Reference Steifetten and Dale2006). In small and isolated populations unpaired males may be all that remain prior to population extinction (Dale, Reference Dale2001; Vieuxtemps & Jacob, Reference Vieuxtemps and Jacob2002; Donald, Reference Donald2007; Menz et al., Reference Menz, Mosimann-Kampe and Arlettaz2009b). In Norway populations are limited by a drastic reduction in the number of breeding pairs because of females dispersing away from the population (Dale, Reference Dale2001; Steifetten & Dale, Reference Steifetten and Dale2006), resulting in the male-biased sex-ratio seen in declining populations (Dale et al., Reference Dale, Steifetten, Osiejuk, Losak and Cygan2006). In relatively isolated populations there is little opportunity for recruitment of females from elsewhere (Steifetten & Dale, Reference Steifetten and Dale2006). Declines and local population fluctuations are also driven by males undertaking relatively long-distance breeding dispersal in search of females (Dale et al., Reference Dale, Lunde and Steifetten2005; Dale & Christiansen, Reference Dale and Christiansen2010; Dale & Steifetten, Reference Dale and Steifetten2011). Loss of females from a population can only be mitigated by increasing the availability and suitability of habitat patches within breeding areas, which requires detailed knowledge of the species’ habitat and foraging requirements (Steifetten & Dale, Reference Steifetten and Dale2006).
Illegal captures during migration
The ability to constitute fat reserves rapidly before autumn migration seems to be an idiosyncrasy of the ortolan bunting, a characteristic known for centuries in gastronomic circles around Europe (Bastien, Reference Bastien1798; Kumerloeve, Reference Kumerloeve1954; Claverie, Reference Claverie1955). The fact that the ortolan bunting is the only species of bunting in Western Europe that undertakes long-distance migration may indicate specific adaptations for storing fat reserves. Historically, large numbers of ortolan buntings were trapped during the autumn and, to a lesser extent, spring migration, mostly in southern Europe (Claessens, Reference Claessens1992). Small traps known as matoles are used that are baited exclusively with nearly ripe oat stalks (Claverie, Reference Claverie1955). It is likely that the species has become a delicacy because of its propensity to lay down fat relatively quickly when fed grain (oat and millet) ad libitum in captivity (Claverie, Reference Claverie1955; Claessens, Reference Claessens1992; Dale, Reference Dale1997; Steifetten & Dale, Reference Steifetten and Dale2006). After fire a wild form of oat occurs en masse in some Mediterranean habitats (R. Arlettaz, unpubl. data) and may have constituted an important food source prior to the expansion of agriculture.
In some areas of south-west France, despite the species now being protected, trapping and fattening continues unabated, with an estimated 50,000 birds illegally captured per year until at least the early 1990s (Claessens, Reference Claessens1992). No studies have been conducted on the consequences of this regionally intensive poaching on the demography of temperate European populations. Although it is unlikely that birds from all declining European populations of the ortolan bunting cross these areas during migration, in a wide-scale metapopulation system these losses may affect overall population dynamics and thus also have regional consequences for distant populations.
Environmental changes in wintering areas
Although the migration phenology is well documented for Europe (Stolt, Reference Stolt1977; Cramp & Perrins, Reference Cramp and Perrins1994; Stolt & Fransson, Reference Stolt and Fransson1995; Yosef & Tryjanowski, Reference Yosef and Tryjanowski2002; Bairlein et al., 2009), the wintering areas of the ortolan bunting are poorly known. Habitat changes in wintering areas because of climate variation and/or anthropogenic impacts on land (e.g. pesticides, Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005; Zwarts et al., Reference Zwarts, Bijlsma, van der Kamp and Wymenga2009) may also contribute to the observed population declines (Conrads, Reference Conrads1977; Kutzenberger, Reference Kutzenberger, Tucker and Heath1994; Busche, Reference Busche2005; Vepsäläinen et al., Reference Vepsäläinen, Pakkala, Piha and Tiainen2005; Lang, Reference Lang2007). However, the fact that some populations in the Mediterranean have been recently expanding (Brotons et al., Reference Brotons, Herrando and Pons2008) seems to indicate that the problem may lie primarily with the quality of the breeding grounds in Europe, possibly compounded by environmental changes in African wintering areas. Identification of the wintering areas of this species is imperative for understanding factors that may be affecting the species outside the breeding season and whether birds from temperate and Mediterranean populations winter in different areas.
Discussion
Knowledge gaps and recommendations for future research
Further information on the diet of the ortolan bunting across its range, particularly quantification of nestling diet in relation to prey availability in the main foraging habitats, is required for a full understanding of the species’ ecological requirements. In particular, quantification of the abundance and availability (the latter being abundance modified by accessibility) of major invertebrate groups in relation to the stages of vegetation succession following events such as fire may provide information on why the species colonizes these disturbed habitats during specific time windows (Pons & Clavero, Reference Pons and Clavero2010). Understanding the relationships between timing of breeding and prey phenology/availability would also elucidate the potential effects of weather and climate variation on reproductive output. Shifts in insect phenology could potentially lead to a mismatch between breeding season and prey availability, a phenomenon that may particularly affect long-distance migrants (Both et al., Reference Both, Van Turnhout, Bijlsma, Siepel, Van Strien and Foppen2010).
More data are required on survival and movement patterns in areas where populations are stable (Pons, Reference Pons, Estrada, Pedrocchi, Brotons and Herrando2004). As most detailed studies on population structure and dynamics have been conducted in northern Europe (particularly Norway: Dale, Reference Dale2001; Steifetten & Dale, Reference Steifetten and Dale2006; Dale & Steifetten, Reference Dale and Steifetten2011), a comparison between eastern European, Mediterranean and temperate populations would facilitate an understanding of the demographic factors limiting populations, such as the propensity for females to disperse away from certain areas. Investigation into the extent of continued poaching would elucidate the potential effects this may have on the demography of temperate European populations as a whole.
Climate variation has already affected the ecology and distribution of some bird species (Arlettaz et al., Reference Arlettaz, Schaad, Reichlin and Schaub2010; Both et al., Reference Both, Van Turnhout, Bijlsma, Siepel, Van Strien and Foppen2010). It is uncertain what effect predicted climate change scenarios would have on the ortolan bunting, especially given the paucity of data on future precipitation regimes (Easterling et al., Reference Easterling, Meehl, Parmesan, Changnon, Karl and Mearns2000) and the impact of weather on the species’ reproductive success. Studies at the edges of the species' range could provide insights into possible colonization of higher latitudes and altitudes. In Mediterranean Europe temperature increases may lead to abandonment of the warmest areas. Such studies are needed to disentangle the future effects of climate modification and ecosystem changes. Identification of the species’ African wintering grounds by use of new light-weight tracking techniques such as geolocators (Bächler et al., Reference Bächler, Hahn, Schaub, Arlettaz, Jenni and Fox2010) would facilitate the assessment of any potential environmental issues that may be contributing to the decline of the species outside its breeding areas and elucidate the connectivity of European breeding populations.
Conservation recommendations
Until we know more about the specific factors limiting ortolan bunting populations, we recommend application of evidence-based conservation measures (Pullin & Knight, 2001) to counteract vegetation encroachment and increase the proportion of patches of bare ground within vegetated patches close to ortolan bunting breeding areas. This could be achieved through extensive grazing, controlled fire (Wilson et al., Reference Wilson, Whittingham and Bradbury2005; Schaub et al., Reference Schaub, Martinez, Tagmann-Ioset, Weisshaupt, Maurer and Reichlin2010) and forestry interventions such as short-rotation coppicing (Berg, Reference Berg2002). However, attention should be paid to protecting sufficient dense grass sward to support prey populations. Patches of bare ground in cultivated fields close to breeding areas may also be produced by spring sowing, decreasing seed sowing density or increasing the distance between rows in sown fields. Although prescribed fire may be a cost-effective management option that is already used in several countries to counteract vegetation encroachment (Montané et al., Reference Montané, Casals, Taull, Lambert and Dale2009), further research is required to determine the potential detrimental effects of prescribed burning on other aspects of biodiversity. In southern Switzerland extensive grazing, forestry measures (coppicing), sowing of oat fields and controlled fire have recently been applied simultaneously to halt the decline of a rare butterfly species and the ortolan bunting (E. Revaz & R. Arlettaz, unpubl. data). Thus in certain habitats conservation measures targeting the ortolan bunting may have broader benefits for biodiversity.
Acknowledgements
We acknowledge O. Roth and C. Marti for assisting with sourcing literature. V. Braunisch, T. Reichlin and M. Berman assisted with translation of German and French literature. MHMM was supported by an Australian Postgraduate Award, and grants from the Holsworth Wildlife Research Endowment and the School of Plant Biology at the University of Western Australia. We also thank R.D. Phillips and two anonymous referees for their constructive comments.
Biographical sketches
Myles Menz has a broad interest in landscape ecology and conservation biology of birds, insects and plants, with particular interest in pollination ecology and the restoration of plant-pollinator mutualisms. Raphaël Arlettaz has wide interests in biodiversity conservation. His research is aimed at providing the necessary rigorous evidence-based guidance to maintain and restore ecosystems and their emblematic species, especially vertebrates and invertebrates in agro-ecosystems and Alpine ecosystems. He is also committed to bridging the great divide that exists between research and action in conservation biology, developing integrated research-implementation programmes, mostly within Switzerland.
