Revegetation of peat excavations in a derelict raised bog.

Abstract

Thorne Waste, S. Yorkshire, is a large raised bog, subject to extensive commercial peat extraction, which retains some re-flooded, revegetated peat cuttings. The composition of the vegetation of some of these cuttings (96 sampled quadrats, each of 25 m2 ) has been related to a number of environmental variables. Revegetation sequences in the abandoned cuttings (all less than 60 years old and with less than 20 cm peat infill) were established by stratigraphical analyses. In some cuttings a basal layer of ericaceous peat apparently marks temporarily drier conditions preceding deeper inundation. Some former mire species (e.g. Sphagnum imbricatum have not survived the peat-cutting period even on uncut surfaces (baulks) but these have been much drained and burnt. The revegetated cuttings contained species typical of poor-ten (e.g. Carex curta) as well as of ombrotrophic mires (e.g. Andromeda polifolia); chemical analyses indicated weakly minerotrophic peat water. This may favour such species as Sphagnum recurvum. There was no indication of chemical enrichment from inflow from surrounding farmland, from underlying mineral ground or from the clay-lined canals excavated across the bog. Release of cations following drainage and re-flooding may provide some explanation. Only soluble reactive phosphorus (SRP) concentration in peat water samples showed consistent and significant seasonal variation, with the lowest values in summer. There were also only a few consistent differences in water chemistry amongst most of the vegetation units identified by TWINSPAN. Nodum 4 (Eriophorum-Sphagnum) stands (richest in ombrotrophic species) occupied the least minerotrophic conditions. One culling, dominated by Juncus, effuses, had no ombrotrophic species (nodum 6). Water concentrations of Ca2+ , Na+ , K+ NH4 + , SO4 2 and SRP were significantly larger here than in other cuttings and peat fertility (estimated phytometrically) and K, N and P concentrations in peat extracts were significantly greater than in a Sphagnum-dominated cutting. The J. effuses samples had strong negative loadings on axis 2 of a Principal Components Analysis (PCA) of the Adriatic data; concentrations of NH4 + , Na+ and SO4 2 in the water showed significant negative correlation with axis 2 ordination scores. Seasonal fluctuation of water level was similar in most cuttings (the J. effuses cutting was exceptional) but variation in depth relative to the peat surface (caused by excavation to different depths) was related to floristic variation. PCA axis 1 may reflect a water-table gradient with TWINSPAN noda 2 (Calluna Sphagnum recurvum) and particularly 1 (Pteridium Campylopus) occupying drier conditions than the others. The wetter cuttings (except the J. effusus site) supported most ombrotrophic and poor-fen species. They were not readily segregated into discrete vegetation types and the two TWINSPAN classes (noda 3 and 4) that accommodated them showed much overlap on the PCA ordination. Nodum 4 samples tended to have higher loadings on PCA axis 2 and to be more species-rich than nodum 3 samples. Species-richness was strongly positively correlated with PCA axis 2. This may reflect, in part, stand maturity, vegetation with most species having been disturbed least recently.