Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.695590
Title: Model forest system : even-aged Douglas fir plantation with invasive Rhododendron
Author: Haffenden, Austin
ISNI:       0000 0004 5990 0265
Awarding Body: Keele University
Current Institution: Keele University
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis investigated the ecology and dynamics of Douglas fir (Pseudotsuga menziesii var. menziesii) dominated plantations at Coed-y-Brenin, Wales, whose transformation to continuous cover forestry has been delayed due to the presence of non-native invasive rhododendron (Rhododendron ponticum L). The literature review highlighted the complementary evolution of silviculture and modelling concepts to meet societal expectations and advance knowledge and understanding. It highlighted the development of hybrid, individual-based models, facilitated by advances in technology and complex systems theory to address contemporary pressures on forests. Three permanent sample plots of different ages, collectively covering 1.6ha of forest, were resurveyed in 2011, producing a 5-year growth series of mensuration data to parameterise an individual-based spatially-explicit forest growth model, SORTIE-ND. Vegetation, soil and light were surveyed to construct a statistical model of rhododendron seedling establishment and determine Ecological Site Classification. The limiting factors identified by the ecological site classification were a slightly dry Soil Moisture Regime and a poor Soil Nutrient Regime with moderate to high phosphorus and low nitrogen availability. SORTIE-ND was successfully parameterised using a maximum likelihood technique and simulated annealing. Parameterised relationships achieved an excellent level of fit to the data (R2 of 0.785 to 0.971), and the stand simulations produced DBH density plots comparable to observed size distributions. A Bayesian statistical model of rhododendron seedling establishment was produced. The cross-validated model predicted 81.3% of 24 survey stations with rhododendron seedlings, and 75% of 16 stations without seedlings, with an overall accuracy of 77.5%. Seedling establishment probability increased with soil O layer depth and decreased with increasing summed tree height in a 12x12 m neighbourhood. SORTIE-ND was then parameterised, using published literature, for the rhododendron lifecycle, to highlight knowledge gaps and as the novel first step towards full parameterisation from field data and the creation of decision support systems enabling foresters to interrogate the impact of different harvest regime scenarios on rhododendron invasion vulnerability.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.695590  DOI: Not available
Keywords: SD Forestry
Share: