As it is normal to characterise grassland habitats on the basis of their plants, it is perhaps unsurprising that goals for re-creation have focused almost exclusively on re-establishing the desired botanical composition (Bakker and Berendse 1999; Walker et al. 2004). The assumption has been that successful re-creation of the plant communities will be followed by other components of these habitats, such as the invertebrates (Woodcock et al. 2008; Woodcock and McDonald In press). In the case of the invertebrates, this assumption reflects their dependence on the plant communities for the provision of both host plants and as structural refuges (Mortimer et al. 1998; Morris 2000; Mortimer et al. 2006). While there is certainly a relationship between the establishment of plants and invertebrates during habitat re-creation, this relationship is complex (Woodcock et al. 2008). By understanding how more than one trophic level responds to re-creation management, the potential for successfully replicating target communities - in this case a species-rich chalk grasslands - should be increased (Forup and Memmott 2005; Woodcock et al. 2008).
The practice of using seed mixes from commercially available sources has been shown to be effective in both promoting the rapid establishment of sward cover and reducing the occurrence of pernicious weeds (Stevenson et al. 1995). Potentially, such seed mixes may contain both grass and forb components (Stevenson et al. 1995; Pakeman et al. 2002; Walker et al. 2004). In this study we have considered a simple grass-only seed mix, as this is common practice in UK agri-environment schemes. This grass-only seed mix has proved popular to farmers as, in addition to being effective in controlling pernicious weeds, it is relatively inexpensive (DEFRA 2005). However, the impact of this seed mix was to reduce overall plant species richness, independent of whether this approach was used in combination with hay spreading or brush harvesting techniques used to introduce a broader spectrum of local provenance seeds. This suggests that while the grass-only seed mix may be effective in reducing undesirable weed establishment (Stevenson et al. 1995), it was also limiting the establishment of desirable chalk grassland species (Crofts and Jefferson 1999; Willems 2001). The impact of the grass-only seed mix was also seen for beetle species richness, which tended to be lower where the grass-only seed mix had been sown in combination with either of the methods used for introducing local provenance seeds. While the effect of the grass seed mix on the beetles was small, its impact reflects the strong dependence of the beetles on the establishment of key forb host plants, a factor reduced where the grass-only seed mix was sown (Woodcock et al. 2006; Woodcock et al. 2008).
The impact of the grass-only seed mix on the plants was limited not only to the species richness of the sward, but also affected its botanical composition. Unsurprisingly, this was characterized by strong associations with the sown grass species F. ovina, C. cristatus, A. capilaris and Dactylis glomerata L. where the grass-only seed mix had been used, either in combination with the local provenance seeds or on its own. However, where the grass-only seed mix was sown on its own, plant community structure was largely dominated by these grasses, and so distinct from when sown in combination with local provenance seeds.
However, the use of the grass-only seed mix acted as a competitive filter limiting forb establishment when sown in combination with either hay spreading or brush harvesting methods. For example, Centaurea nigra L. (Asteraceae) was successful in becoming established only when introduced in combination with the grass-only seed mix (Pakeman et al. 2002).
The response of beetle assemblage structure to the grass-only seed mix shows parallels to that of the plants, although in contrast there were far fewer significant responses to the treatment interactions with local provenance seed treatments. The parallel was most apparent in terms of a divergence in beetle assemblage structure where management had either included or not included the grass-only seed mix. Typically, the beetle assemblages associated with the grass-only seed mix contained species that were found ubiquitously across all treatments, with few beetles being strongly associated with this management practice. Given that the majority of UK phytophagous beetles found within grasslands are dependent on forbs as host plants, rather than grasses (Bullock 1992), the poor establishment of forbs where the grass-only seed mix was sown would explain this trend (Mortimer et al. 2006; Woodcock et al. 2008).
The use of local provenance seeds during grassland re-creation had several advantages over the use of commercially available seed mixes, such as those described above (Stevenson et al. 1995; Pakeman et al. 2002; Walker et al. 2004). Firstly, seeds will be sourced from local biotypes of plants and so adapted to local environmental conditions. It has been suggested that this will improve survival of these plants once established as well as also contributing to the preservation of local genetic diversity (Jones et al. 1999; Gustafson et al. 2004; Walker et al. 2004). Additionally, methods used in the collection of local provenance seeds will harvest species present within the swards of the donor sites which are not commercially available (Morgan and Collicutt 1994; Jones et al. 1999).
Relative to the grass-only seed mix, the benefits of using hay spreading and brush harvesting techniques was in general seen in terms of the establishment of forb species harvested from the donor site chalk grassland. However, both the method of sourcing local provenance seeds and the application rates affected the success of re-creation. In particular the use of low rates of application for the local provenance seeds tended to result in benefits to the plant community that did not persist for the duration of the study. For example, plant species richness was always higher than the control treatment after the initial application of local provenance seeds, although after three years this initial difference only remained where high application rates has been used. High rates of application of local provenance seeds are therefore likely to be a limiting factor to enhancing plant species richness (Edwards et al. 2007).
Beyond simple enhancement of species richness, differences in community structure for the plants where local provenance seeds has been sown was determined primarily by whether or not they had been applied in combination with the grass-only seed mix. Such an overriding impact on the structure of the plant communities in response to the addition of the grass-only seed mix reflects the competitive dominance of some of these sown grass species (Willems 2001; Bakker et al. 2003). However, the relatively small differences in community structure between the hay spreading and brush harvested seed treatments, either with or without the grass-only seed mix, seem to be characterized by underlying differences in the composition of seeds collected by these two methods (Morgan and Collicutt 1994; Jones et al. 1999; Edwards et al. 2007). These relatively small differences should be weighed against the relative scarcity of the machinery required for brush harvesting (Morgan and Collicutt 1994). Whether the costs associated with such specialized machinery are warranted is arguable, although the overall benefits associated with this technique relative to the more conventional hay spreading methods were apparent, at least in terms of the enhancement of plant species richness.
The success of re-creation was measured relative to an existing example of a species-rich chalk grassland (Edwards et al. 2007; Woodcock et al. 2008). For both the plants and the beetles, the methods used for harvesting local provenance seeds played an important role in influencing the long term success of chalk grassland re-creation. For the plants, similarity to the target grassland increased for the control, hay spreading and brush harvesting methods over the three years period after re-creation management was initiated. The high rates of application of both hay spreading and brush harvesting resulted in the highest similarity to the target grassland, while the control had the lowest. This predictable pattern reflected expected benefits associated with high sowing rates (Stevenson et al. 1995; Jones et al. 1999; Edwards et al. 2007). Although high application rates of local provenance seeds were clearly superior for the plants, botanical similarity to the target grassland under the low application rates remained superior to the control treatment. This was in contrast to the beetles which were reliant on high application rates of hay spreading or brush harvested seeds if similarity to the target grassland was to remain higher than the control over the same three years. Even where high rates of application had been used for the local provenance seeds, the brush harvesting method was only slightly superior to the control in replicating the beetles of the target grassland. Why high rates of brush harvesting were less effective in establishing chalk grassland beetle assemblages than hay spreading is not immediately clear. The probable explanation is that brush harvesting may have failed to introduce key forb species that supported large number of beetles typical of chalk grasslands (Edwards et al. 2007).
Was this article helpful?