Various studies and surveys have examined the changes in botanical composition of set-aside fields in the past few years e.g. Poulton and Swash (1992), Fisher et al. (1992), Wilson (1992), Rew et al. (1992), Clarke and Cooper (1992), Shield and Godwin (1992), Brodie et al. (1992), Burch (1996), Turley et al. (1994), Lawson et al. (1994) and Welch (1991, 1994).
Poulton and Swash (1992) monitored the botanical composition for the first year of set-aside in England. It was found that core areas of set-aside were less diverse than the field boundaries. Furthermore, natural regeneration produced higher species richness and greater populations of notifiable weed species, such as Cirsium arvense, C. vulgare, Rumex crispis, R. obtusifolius and Senecio jacobaea. The most frequent species to occur were Agrostis spp.,Elymus repens, Cirsium arvense, Lolium perenne and Sonchus asper.
Fisher et al. (1992) looked at sown covers in Scotland and concluded that sown fields produced a lower frequency and ground cover of broad-leaved species. Very few annual weeds or volunteer crops were present in the sown covers, and it was predicted that perennial grasses would eventually predominate set-aside fields. Lechner et al.(1992) found that cover crops suppressed weeds on set-aside in Germany. 90% of total weeds consisted of just 8 species. Wilson (1992) studied naturally regenerated set-aside in Southern England and found that under natural regeneration annual species declined as the vegetation developed, and perennial species increased. Wilson also investigated the effect of adjacent of semi-natural habitats on the composition of set-aside and concluded that semi-natural habitats can have an important influence on the characteristics of set-aside. Soil characteristics had little effect on vegetation development.
Rew et al. (1992) examined the spatial distribution of vegetation changes in set-aside over a three year period in the south and south-east of England. Boundary vegetation consisted of a greater number of perennial species, and towards the core areas of fields, perennials decreased and annual and biennials increased. However, the annual Galium aparine was a notable exception and decreased from the boundary. It was concluded that there were three main sources of plant origin: Vegetation propagation (e.g. Agrostis stolonifera); increase in wind disseminated species including Sonchus spp. and Crepis spp.; and a decline in the diversity and cover of annuals originating from the seed bank over the three year period.
Clarke and Cooper (1992) examined weed levels in set-aside and the subsequent crops. On set-aside, frequent cutting altered the habits of Alopercurus myosuroides, Bromus spp. and volunteers. On naturally regenerated set-aside Bromus spp., Cirsium arvense, C. vulgare, Elymus repens, Poa spp., Senecio jacobaea, Sonchus spp. andStellaria media dominated the fields. Rotational set-aside had caused no problems for subsequent cropping, but on permanent set-aside perennial species became problematic to farmers.
Shield and Godwin (1992) studied set-aside on a traditional heavy soil cereal site previously in winter wheat on the Bedfordshire/Hertfordshire border. The developed vegetation sward was dominated by Elymus repens which achieved dominance through competition and the sensitivity of annual species to cutting regimes on the set-aside. The most sensitive species to cutting were volunteer wheat followed by Alopercurus myosuroides. Bromus sterilis was most difficult to control under the new conditions. The first year of set-aside was dominated by Bromus sterilis, Alopercurus myosuroides and volunteer wheat. By the second year these decreased and were replaced by Elymus repens. When Elymus repens occurs in high numbers, future crops can often be threatened by Gaeamannomyces graminis (the take-all fungus) and Pseudoncercosporella herpotrichoides (eyespot) which can remain in plant residues after cutting and cause significant reductions in yield.
Talling and Godwin (1994) found that on calcareous clay soils, where set-aside had been mown, decaying Elymus repens occurred in large amounts, and caused a significant crop yield reduction after set-aside. Decaying rhizomes can immobilise nitrogen. They suggest that a herbicide treatment should be carried out prior to set-aside, then a cover sown to prevent the spread of Elymus repens.
Brodie et al. (1992) studied set-aside in Cambridgeshire. The developed vegetation consisted of grasses characteristic of the weed flora of arable fields and disturbed ground. Although the number of species increased over the duration of the set-aside, the diversity index did not. The increase in the number of species was balanced by greater inequality in the relative abundance, i.e. a few species became extremely abundant, with infrequent areas producing higher diversity.
Turley et al. (1994) compared the development of flora during three years of set-aside under three cover managements: perennial rye grass; perennial rye grass and white clover, and natural regeneration. The sown covers suppressed the development of other grass and broad-leaved species, and by the second year sown covers had eliminated cereal volunteers. On the naturally regenerated fields establishment was slower, but the number of non-volunteer species and grass and broad-leaved plants increased over the three years. Within three years, perennial grasses dominated all sites, except for the those on the sandy soils. Weed levels in the first wheat following set-aside were similar, or lower than, those on land which had remained in arable rotation. However, at one of the sites, failure to control Elymus repens with glyphospate at the end of the set-aside period resulted in couch infesting the following wheat crop. The perennial rye grass and clover covers were very effective at excluding other species, after three years few other species were present in the sward. The naturally regenerated swards produced much greater species diversity (7 species present at Boxworth on heavy soil, 27 species at Gleadthorpe on light soil). Also, on the sown covers, thistles were unable to colonise.
Burch (1996) compared five establishment methods on the development of desirable species on a large scale field experiment at Wye College. Burch suggested that one of two establishment methods should be considered to produce the desired effect of balancing weed control with species enhancement. Where weed control is a principal concern, a grass cover should be established in the first couple of years, and then the spread of more desirable species can occur as the sward matures. In less weedy sites, a more “open matrix” approach will offer partial weed control and more favourable establishment conditions for desirable species. It was found that, perennial rye grass produces a dense sward within a couple of years, with associated species. Natural regeneration promotes growth of both desirable grassland species and pernicious weeds. If neither of these approaches produces the required sward, then sown covers of Festuca spp. and Cynosurus spp. can be used. However, Festuca rubra produces a dense cover which prevents other species colonising making it less suitable for achieving a spatial balance.
Lawson et al. (1994) examined set-aside management strategies on the soil seed bank and weed flora at Woburn. On fields at Woburn, naturally regenerated swards were dominated by grass species throughout the three years. Volunteer oats quickly colonised the set-aside, and acted in a similar way to a sown perennial rye grass cover in the first two years, to be replaced by Poa annua, Agrostis gigantea and Bromus sterilis in the third year. Cirsium vulgare, Sonchus spp. and S. Vulgaris appeared on the natural regeneration plots in the second and third years. It was not until the third year that volunteer oats declined and natural species could colonise the land. There was no evidence of the spread of sown species into the natural regeneration plots. Natural regeneration produced more seeds, more species and greater diversity in the seed bank than sown set-aside, indicating that volunteer oats had not been as effective in preventing seed bank expansion on their own as in association with a sown cover.
Welch (1991 and 1994) studied trends in the botanical composition of fields in N.E. Scotland. After one year of set-aside grasses predominated, with Holcus mollis and Poa annua being most abundant. Agrostis gigantea andElymus repens also reached high levels. Cirsium spp., Senecio jacobaea and Avena fatua occurred in most fields. Broom and gorse soon colonised the Scottish fields. Species richness seemed to be unaffected by type of crop last grown, although fields formerly in rape had the greatest average number of species recorded. Generally field margins were no more diverse than the core areas of fields. Colonisation by grasses was much more rapid than on studies in lowland England. After five years of set-aside grasses continued to dominate. Agrostis gigantea and Poa annua were replaced by grasses of permanent grassland, such as Agrostis capillaris, Dactylis glomerata and Holcus lanatus.White clover was the most frequent of the non-weedy dicotyledenous plants. Noxious weed frequencies remained low. Species richness declined during the five years of set-aside, due to the decline in dicotyledenous species. Many species only occurred in edge quadrats and often only a single species would be present. The process of colonisation is thought to be delayed by early cutting which prevents fruits ripening sufficiently for seed dispersal.
Aquilina and Clarke (1994) examined the effect of cutting date and frequency on perennial broad-leaved weeds on set-aside. Rumex spp. were found to increase in fields where the cover was cut at the full flower stage. Cirsium vulgarewas most affected by cutting and spraying treatments. Cirsium arvense was reduced in the second and third years by cutting treatments. On areas of poor plant cover, Cirsium arvense was less likely to be controlled. Sonchus arvensiswas controlled by most treatments, although cutting alone was not as effective as cutting and spraying.
With all treatments, a good plant cover aided the reduction of perennial broad-leaved weeds without which only a cutting and spraying treatment could reduce populations sufficiently.