Elsevier

Ecological Engineering

Volume 31, Issue 2, 1 October 2007, Pages 79-91
Ecological Engineering

The effects of turf translocation and other environmental variables on the vegetation of a large species-rich mesotrophic grassland

https://doi.org/10.1016/j.ecoleng.2007.05.003Get rights and content

Abstract

This paper reports on the translocation of 5.6 ha of a 9.1 ha dry mesotrophic grassland in 1998. For 4 years before and 5 years after translocation, the vegetation was monitored annually by randomised quadrat sampling and compared with an area of the site which remained untranslocated. Associated soil and climatic variables were also monitored. The hypothesis to be tested was that changes in the vegetation ascribable to translocation were of a similar amplitude to those ascribable to other causes. Soil monitoring showed that it had been possible to reconstruct both vegetation and substrate with only minor changes in soil compaction and soil fertility. The frequencies of 28 species in the source meadow in the 4 years before translocation varied significantly. Three areas of the translocated vegetation monitored separately showed significant changes in frequency in a similar number of species, as did the untranslocated area. Species density gradually declined on the translocated turf for the first 4 years after translocation but recovered in the fifth year, although at least one species has so far failed to recover. Multivariate analysis showed that the translocated vegetation diverged from the untranslocated portion for the first 4 years but that this divergence ceased and partially reversed in the final year. The divergence involved an increase in the cover-abundances of several grasses and a parallel fall in a range of forbs. Weather records suggest that this pattern correlates with 4 years without water stress followed by a much drier growing season in the fifth year. It is concluded that there was a definite response to the 1998 translocation which appears to be comparable in severity with the pre-translocation changes in terms of the effects on species frequency. The untranslocated area also changed suggesting that the changes were not all the result of translocation, and the results for the last year of monitoring imply an element of recovery. The results support the use of turf translocation as a restoration technique where the alternative is the complete loss of the site and where the process of reconstructing the functioning of the ecosystem meets certain standards. Monitoring is continuing for 5 more years.

Introduction

Community (often ‘habitat’) translocation, the movement of assemblages of species, usually together with their associated substrates, from their original site to a new location, is typically attempted in mitigation of development. It is expensive, the results are unpredictable and in the UK it is regarded by the statutory nature conservation agencies as never being an acceptable alternative to in situ conservation (McLean, 2001). On the other hand, community translocation offers opportunities to study the factors affecting the functioning of specific ecosystems on largely intact species assemblages and therefore to contribute to our understanding of the control of biodiversity in ecosystem restoration.

Assessing the success of translocation depends to some extent on defining success. There will always be a loss in naturalness and in the original context. Maintaining the translocated habitat “unaltered” (Byrne, 1991) or “intact” (Bullock et al., 1997, Bullock, 1998) is probably unattainable; vegetation may change whether translocated or in situ. Bullock (1998) rejected this ‘preservation’ aim and suggested an alternative ‘mitigation’ aim of preserving the more important (however defined) plant species and some of the other taxa such that the translocated community resembles the pre-translocated state. Wilson (1998) and Box (2004) suggested that objectives of maintaining (or enhancing) the nature conservation value of the habitat as measured by application of Ratcliffe's criteria (Ratcliffe, 1977), the SSSI guidelines (Nature Conservancy Council, 1989) and comparisons with the National Vegetation Classification (Rodwell, 1992). Box (2004) also suggested that evaluation should include detailed species monitoring for an appropriate period of time, such as 10 years.

Only 14 grassland turf projects overall had sufficient monitoring data to be included in three reviews of community translocation in the UK by Bullock (1998), Anderson (2003) and Box (2004). At least three of the receptor sites received little preparation and at least seven of the translocated grasslands received little or inappropriate after-management, making it difficult to assess reasons for change. Bullock (1998) reviewed 10, mostly grassland projects across England and Wales. Monitoring had taken place typically for 3–4 years after translocation. He found that all grassland communities had shown both gains and losses in species. He concluded that the receptor sites were often poorly investigated and prepared and that the initial surveys, at best, started immediately before translocation, limiting comparison. He considered that there is an inherent time-lag in response, that turf depth may be important for some species and that dry grassland translocation seemed most successful. Box (2004) followed Hodgson (1989) in considering the critical factors controlling the outcome to be the similarity in the environmental context of the donor and receptor site, the translocation technique and the habitat management of the translocated habitat.

Anderson (2003) reviewed a similar set of turf translocations in the UK and considered that in many instances there were irreconcilable differences between the donor and receptor sites which affected the communities and that responses to variation in climate or management could have been confused with responses to translocation. In particular the disturbance resulting from translocation has an impact on soil nutrient cycling. Although there are soil quality thresholds to ecosystem recovery on degraded soils (Garten and Ashwood, 2004), in old meadows on mature soils, botanical diversity is most likely to be adversely affected by the levels of available nitrogen and phosphorus being too high (Marrs, 1993, McCrea et al., 2004), a situation which might arise from the disturbance resulting from translocation and often has to be ameliorated in successful ecosystem restoration (Gilbert et al., 2003).

There are a few examples of turf translocation which offer long-term detailed monitoring data. Most consist of relatively small areas of vegetation translocated short distances in very deep turves. Circa 0.3 ha of 400–500 mm deep turves of magnesian limestone grassland translocated at Thrislington in Durham, UK were assessed for 8 years after translocation (Cullen and Wheatear, 1993). Initial changes in both flora and invertebrates appeared to be followed by at least partial recovery although considerable fluctuation in the balance between species was recorded. 0.4 ha of montane grassland in the Harz mountains in Germany was translocated in turves 500 mm deep and monitored for 5 years, with minimal appearance of disturbance indicators, successful preservation of the flora including almost all regionally threatened species and with maintenance of four different vegetation types but with loss of the original spatial pattern (Bruelheide and Flintrop, 2000).

The most thorough and long-term monitoring of a translocation of mesotrophic grassland seems to be at Brocks Farm, Devon, UK, where 0.5 ha of species-rich mesotrophic grassland were translocated onto a nearby site as 200 mm deep turves after soil stripping and replacement with subsoil from the donor site (Jefferson et al., 1999). The vegetation was monitored annually for 9 years alongside a similar neighbouring ‘control’ area of grassland maintained in situ and another area translocated as turf fragments. The results suggested that over the 9 years of monitoring, stress-tolerant species declined on the turf transplant and that of 31 species for which comparisons could be made, 24 behaved differently on the translocated and untranslocated sites. It was concluded that floristic changes had occurred which had taken many years to become discernable. This interpretation was supported by a multivariate analysis of the Brocks Farm data (Gibson, 1997), which gave some indication that the translocated and untranslocated areas of vegetation were becoming more dissimilar with time. However, many of the species declining on the turf did not decline on the turf fragment area suggesting that at least some of the deleterious changes on the turves were not an inevitable consequence of translocation.

Both at Brocks Farm and at Thristlington, marked fluctuations in the frequencies of individual species were noted from year to year. Other studies have shown large fluctuations in apparently stable grassland, as in the monitoring of control plots in the long-term experiment at Park Grass at Rothamsted between 1929 and 1979 reported by Dodd et al. (1995), who concluded that the data suggested that the concept of a stable plant community needed re-evaluation. In some cases these fluctuations can be ascribed to annual fluctuations in the weather. Dunnett et al. (1998) showed that the control plots of the under-managed dry mesotrophic grassland road verges near Bibury, Gloucestershire (Willis, 1972, Willis, 1988) showed many significant correlations in variation in the species with weather variables, particularly ones recorded for spring and summer seasons.

The examples suggest that turf transplant monitoring projects should monitor associated soil and weather variables both at donor and receptor sites for possible explanations of variation in the vegetation associated with the transplant period. They also suggest that success in turf translocation might be defined in the following terms:

  • (a)

    Changes in the vegetation should be of a comparable magnitude to those recorded in the vegetation before translocation.

  • (b)

    Changes in the vegetation should be of a comparable magnitude to those recorded in areas of the vegetation which have not been translocated.

  • (c)

    Translocated and untranslocated areas of the vegetation should not become more dissimilar with time.

The present study concerns the translocation of a large area of species-rich grassland turf at Durnford Quarry, Long Ashton, near Bristol in SW England (national grid reference ST 538715) in winter 1998–1999. Following four seasons of vegetation monitoring before translocation, the vegetation and selected environmental variables have been monitored for 5 years at the donor and receptor sites after translocation and this exercise is still continuing. The aims of the present paper are to describe quantitatively and to attempt to evaluate year-on-year changes in the vegetation in terms of the transplant event and the background variation in the environment in order to evaluate the success of the translocation. The hypotheses to be tested is that through the monitoring period changes in the vegetation and soils ascribable to translocation are of a similar amplitude to those ascribable to other causes.

Section snippets

The translocation scheme

Top Park Field (National Grid reference ST544717) at Durnford Quarry included 9.1 ha of dry, species-rich grassland, conforming to the UK National Vegetation Classification (NVC) mesotrophic grassland community Cynosurus cristatus–Centaurea nigra (MG5), Lathyrus pratensis subcommunity (Rodwell, 1992), forming part of the Ashton Court Estate Site of Interest for Nature Conservation. In winter 1998–1999, 5.6 ha was translocated to Ashton Hill Field, also at Durnford Quarry, to allow the extension

Environmental comparison between the donor site, Top Park Field and the receptor site, Ashton Hill Field

Although varying in magnitude, measurable differences in temperature were recorded between the two sites each year (1999–2003), especially in spring. Ashton Hill Field is not a topographic frost hollow, but the woods and hedges can trap katabatic air, while Top Park Field allows much freer movement of cold air. Often ground frosts continue into June in Ashton Hill Field, while ceasing in April in Top Park Field. These conditions may slow plant growth, resulting in plant damage if severe frosts

Discussion

The climatological comparison of the source site Top Park Field and the receptor site Ashton Hill Field suggested that the latter is more sheltered, with small but measurable tendencies to higher humidity, greater frost liability and higher summer temperatures. The soil penetrometer analysis showed that following translocation there was early compaction of the surface soil on the translocated turves of Ashton Hill Field which gradually ameliorated over the next 3 years and may even have

Conclusions

The results of the monitoring programme, although not yet complete, give some support to the use of turf translocation as a technique of value in the mitigation for loss and in nature conservation where the alternative is the complete loss of the site. The results obtained so far support the hypothesis that changes in the vegetation ascribable to translocation are largely of a similar amplitude to other changes taking place on the site and that the vegetation has been translocated successfully.

Acknowledgements

The authors would like to thank E.V.J. Cohn and N. Musgrove for help in producing the paper, M.J. Carter Associates and Alaska Environmental Contracting for access to the translocation plans, Pioneer Aggregates and Tarmac Southern Ltd. for funding the monitoring project, G. Wilson, W. Bond, E.V.J. Cohn and J. Box for comments on early drafts and three anonymous referees for comments on a late draft.

References (35)

  • Cox, J.H.S., Leach, S.J., Byrne, S.A., Blake, C.P., Buckingham, H.G., 1992. Progress reports on monitoring of grassland...
  • R. Cullen et al.

    The flora and invertebrate fauna of a relocated grassland at Thrislington Plantation, County Durham, England

    Restoration Ecol.

    (1993)
  • M. Dodd et al.

    Community stability: a 60-year record of trends and outbreaks in the occurrence of species in the Park Grass experiments

    J. Ecol.

    (1995)
  • N.P. Dunnett et al.

    A 38-year study of relations between weather and vegetation dynamics in road verges near Bibury, Gloucestershire

    J. Ecol.

    (1998)
  • C.W.D. Gibson

    Grassland Translocation Contract No VT5.1M—Brocks Farm

    (1997)
  • J.P. Grime et al.

    Comparative Plant Ecology

    (1988)
  • J.G. Hodgson

    Selecting and managing plant materials used in habitat construction

  • Cited by (29)

    • A comparison of different soil transfer strategies for restoring a Mediterranean steppe after a pipeline leak (La Crau plain, South-Eastern France)

      2014, Ecological Engineering
      Citation Excerpt :

      This has already been demonstrated for certain grasslands in Great Britain where positive results were achieved (Bullock, 1998; Box, 2003; Trueman et al., 2007). For example, in 18 experiments, the rate of transfer for plant species ranged from 54% to 91% after 3 to 7 years with a high success rate in particular for the transfer of turf (Bullock, 1998) which may be as high as 100% after 5 years for a dry grassland restored using this method in Great Britain (Trueman et al., 2007). On the other hand, this technique is certainly not sustainable, since soils are not considered a renewable resource.

    • Can ecological engineering restore Mediterranean rangeland after intensive cultivation? A large-scale experiment in southern France

      2014, Ecological Engineering
      Citation Excerpt :

      In contrast, soil transfer partly confirmed our hypothesis. Soil nutrient content was not affected by soil transfer, contrary to what sometimes occurs in habitat or turf translocation (Trueman et al., 2007). Only pH and calcium were higher than at the soil donor site.

    • Experimental wet heath translocation in Dorset, England

      2011, Ecological Engineering
      Citation Excerpt :

      This was not possible for the Gadle Knapp translocation because the entire mire system was destroyed at the donor site and there was not any other appropriate mire system in the immediate vicinity. It is known that fluctuation in weather conditions (precipitation in particular) from one year to the next correlate with changes in the species structure of grassland communities (Trueman et al., 2007). The same may be true for wet heath and mire communities, but there appears to be no direct published evidence.

    View all citing articles on Scopus
    View full text