Open semi-natural habitats, like those inhabited by the two relict butterfly species studied here, must be managed to prevent forest recolonization. The real question of interest to people in charge of the design and implementation of biodiversity conservation plans is how to determine the management protocols that would yield the best conservation results. Mowing and grazing are currently used in many biodiversity conservation and management actions, including Ardenne humid grasslands. However, we showed here that both can have very detrimental effects on target species, if used without caution, and they must therefore not be blindly applied in any case.
Some studies showed that mown grasslands present a reduced butterfly richness and abundance compared with abandoned ones (Erhardt 1985; Erhardt and Thomas 1991; Munguira and Thomas 1992; Feber and Smith 1995, Feber et al. 1996), while other studies showed that hay meadows can harbor a rich and original fauna, including rare species, compared with other biotopes (Swengel 1996; Wettstein and Schmid 1999; Stefanescu et al. 2005). This study and other results of the same experiment (Goffart et al. 2001) suggest that the impact of mowing on populations may be highly negative, dramatically diminishing the butterfly densities (> 50% loss). P. eunomia appears especially sensitive to mowing, with a real avoidance of mown areas by adults and a disastrous effect on early instars. The likely explanation is that this species is linked to habitats where structural elements like grass tussocks can be found (Goffart et al. 2001; Turlure et al. 2009). Its populations can be affected by a widespread application of cutting management in the presently abandoned humid grasslands. The persistence of populations in the Ardenne valleys at a time when mowing practices were widespread can probably be explained by the far bigger extent of its habitat in the past, allowing for regional persistence of populations despite the decrease in population density and/or extirpation of local populations due to mowing. Indeed, movements of individuals from occupied sites allowed for metapopulation dynamics through rescue effect (immigrants increase population density) and recolonization (Hanski and Gilpin 1997). The non-mechanical way of cutting and its temporal spread might also have had a lower impact than current mowing techniques. Thus, the traditional way of management of these biotopes cannot be retained as an ideal solution for the long-term conservation of P. eunomia populations.
Light grazing, being a more "natural" way of management (although not a traditional way) and generating heterogeneity in the vegetation cover (Van Wieren 1991, 1995; WallisDeVries et al. 1998), therefore appears as a prime alternative. Indeed, several studies showed that grazing management can be very efficient to conserve butterfly populations or communities and their habitats, when used with light or moderate stocking level (Dolek and Geyer 1997; Ellingsen et al. 1997; Weiss 1999; WallisDeVries and Raemakers 2001; Ellis 2003; Pöyri et al. 2004; Öckinger et al. 2006). The present study indicates that grazing can also be a management option for P. eunomia and L. helle, the two threatened species linked to a food plant highly sensible to grazing and trampling, if the following conditions are applied: (1) very light stocking rates are used: < 0.2 LU ha-1, as recommended already by others in other contexts (WallisDeVries and Raemakers 2001; Ellis 2003), (2) preferably used in a rotational basis, and (3) avoiding crucial parts of the butterfly life cycle (April to July for P. eunomia and L. helle).
However, despite its effectiveness in conserving habitats of these target butterflies with relatively limited impacts on their populations, grazing should not be considered as "the" only appropriate and adequate management option in Ardenne humid grasslands. Indeed, light grazing is difficult to apply on small areas, where the risk of overgrazing is higher (Schtickzelle et al. 2007), and this method is better applied in large areas (Bakker and Londo 1998; Ellingsen et al. 1997). Actually, the impact of grazing on butterfly abundance or richness is highly variable and clearly dependent on its intensity and the modalities of its application (Kruess and Tscharntke 2002; Pöyri et al. 2006).Very detrimental effects have been reported at high stocking levels (Wettstein and Schmid 1999; Kruess and Tscharntke 2002; Ellis 2003; Schtickzelle et al. 2007). Mowing may then replace grazing on small areas (< 1 ha), if used with caution, on a rotational basis, with a pluriannual regime.
More generally, the choice of a single management method (i.e., mowing or grazing) is not advisable in a conservation perspective at the landscape scale. The use of different management practices and intensities, including permanent or temporary abandonment, can be fruitful for individual species (Schwarzwälder et al. 1997) as well as entire communities (Swengel 1998; Dolek and Geyer 1997; Wettstein and Schmid 1999; Balmer and Erhardt 2000; Pöyri et al. 2005; Vögel et al. 2007). The particular context of each site, and especially the area of the habitat in the landscape, is clearly an important parameter regarding the choice of the management method aimed at the conservation of viable butterfly populations and of a rich fauna in this kind of biotope.
Was this article helpful?