Use this URL to cite or link to this record in EThOS:
Title: Predicting the impact of hydrological change on wetland vegetation
Author: Kennedy, Michael Patrick
ISNI:       0000 0001 3597 2510
Awarding Body: University of Glasgow
Current Institution: University of Glasgow
Date of Award: 2001
Availability of Full Text:
Access from EThOS:
Access from Institution:
During a three year field study (1997-2000) vegetation assemblages, collective vegetation variables, traits of dominant populations and hydrological and hydrochemical variables were repeat-sampled within seven wetland sites across Scotland and northern England. These ranged from the Irish Marshes, Inverness-shire in the north, to Tarn Moss, Cumbria at the southern extreme. Sampling was conducted at a total of fifty-six permanent sample stations located along a total of eleven transects. Vegetation groupings were defined using multivariate analyses, and were classified as various fen, mire, and swamp NVC community types. The various groups were characterised by the values for the range of variables measured, and significant differences were seen between a number of these variables for different groupings. In addition, certain separate groupings with the same community classification were also seen to have significant variations between them in terms of trophic status, and canopy height and biomass values. Collective vegetation variables and dominant population trait values were successfully predicted from physical and chemical variables measured within the groundwater and substrate during 1999. A number of specific models incorporating relatively large numbers of predictor variables were proposed alongside more general models incorporating fewer predictor variables. The greatest predictive power with R2 = 0.67 (p<0.001) for a model predicting stem density (m-2). Conversely, vegetation variables proved useful for predicting characteristics of the groundwater environment, for which specific and general models were against proposed. In this instance, the greatest predictive power was R2 = 0.79 (p<0.001) for a model predicting minimum water table level (i.e. maximum level of drawdown). The models were tested using data collected during 2000 from repeat sites and independent sites. Whilst some of the variables were predicted within noisy limits, predicted values generally corresponded well to observed values.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: QK Botany