An Investigation into the importance of propagule provenance in restoration ecology.

Abstract

There is increasing concern among restoration ecologists that using non-local propagules in revegetation schemes may influence the success of grassland restoration. This research programme investigated the importance of propagule provenance in restoration schemes. In particular the work focused on two areas. Firstly, the study investigated the significance of propagule provenance for plant establishment and persistence using Lotus corniculatus as a model species. Secondly, it investigated the practical implications of restricting seed procurement to local sources, using a field trial which assessed the success of limestone grassland re-creation under different treatments. Propagules were collected from two contrasting habitats within each of six regions in the British Isles and a common garden experiment was used to quantify the genetic component of provenance as estimated by plant morphology and fitness. There were differences in survival, growth habit, pubescence, leaf shape, plant size and fecundity between plants grown from seeds sourced from different regions. In addition these populations differed in the extent to which they were damaged by seed herbivores. Differences between plants grown from seed sourced from contrasting habitats were limited to morphology. These findings suggested that populations would be likely to perform differentially in a restoration environment. Both geographical location and the ecological conditions at the seed source should be considered when procuring seeds for a restoration scheme. A field trial was set up to establish whether local propagules exhibited higher fitness than those collected from non-local sources. Propagules were collected from two contrasting habitats in each of fifteen regions. Two restoration environments at a single site were investigated; one was treated with a dressing of topsoil, the other site was untreated bare clay substrate. Differences between populations were measured in terms of both geographical and ecological distance. Results for both sites demonstrated that although there was no home-site advantage in terms of geographical distance, plants from more distant populations were smaller and less fecund in the restoration environment. An investigation into the relationship between ecological distance and plant performance produced different results on the treated and untreated plots. On the treated plot there was no significant relationship but on the untreated plot, plants from more distant populations were larger and more fecund. The contrast in the results obtained for the effect of geographical and ecological distance on performance in the untreated restoration environment is interesting. The enhanced performance of geographically local populations agrees with findings from previous studies. The findings for ecological distance are unexpected based on other work. However, it is postulated that the initial success of non-local populations maybe misleading, as environmental conditions which are infrequent but typical of the area may lead to high mortalities in the longterm. The range of what can be considered local is rarely considered in studies that investigate plant provenance, but work from other areas suggests that there is sufficient variation over small distances to warrant seed collection within 100m. Fine scale phenotypic variation over 200m in populations of L. corniculatus was investigated, however there was no evidence to suggest that seed collection should be restricted to 100m, consequently seeds could be safely collected up to 200m from a restoration site. If seed procurement is to be restricted to local seed then it is possible that the seed application rate and species mix available for a restoration project may be limited. A field trial investigating the effect of different treatments showed that it is possible to establish an appropriate plant community using a low sowing rate. However, that community will be more vulnerable to changes in the environment during the establishment phase and more open to invasion by colonizing species which are likely to be weedy in the first few years. A comparison of two seed mixes showed that a diverse seed mix resulted in an increased diversity and evenness of vegetation. There was no benefit in adding a nurse grass to compensate for a low application rate of local species. In conclusion, although it is an advantage to use seeds of local provenance in restoration schemes, it will be necessary to balance this with the likelihood of successful re-vegetation given the seed available