School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 6 of 6
-
Item
-
Item
-
Item
-
ItemStochastic structured models for Banksia goodii and Anas rhynchotis rhynchotis populations( 1994)Models were constructed for Banksia goodii and Anas rhynchotis rhynchotis in the hope that they may be useful for the management of the species. Banksia goodii is an endangered plant species distributed in Western Australia. Anas rhynchotis rhynchotis is a native duck species under harvesting management in Victoria, South Australia, and Western Australia. All simulations in both models were carried out using the RAMAS/stage program, developed by Ferson (1992; see Appendix 3). The Anas rhynchotis rhynchotis model is stochastic, structured, two-sex, and discrete. It assumes that the populations in eastern Australia and Tasmania are a single population. The stages include male juveniles, female juveniles, male adults, and female adults. Parameters were estimated based on personal communications and published data for the species or were inferred from other related species. The hatching rate, female juvenile survival, female adult survival, sex-ratio of new offspring, and proportion of number of breeding pairs which breed during breeding season are important and should be estimated as precisely as possible. A hunting rate of below 0.03 may result in a downward population growth trend. A daily bag size of 2 may be too high for the population. The Banksia goodii model is stochastic, structured, and discrete. The stages consist of seedlings, 19 immature stages, and adults. Parameters were determined based on information from B. Lamont and published data. The growth rate of Banksia goodii populations is very slow and is determined very much by survival of adults. The effect of fire on population growth is inhibitive. The species is likely to persist for 50 years but it may not be able to recover from loss of adults. Artificial regeneration via seeding may be helpful for the restoration of adult loss. The adult survival rate and adult survival rate after a fire are important and should be estimated as precisely as possible. A\ model for ramets seems more appropriate than a model for genets. The species needs rehabilitation and protection. Population dynamics modelling should be encouraged in all management of species.
-
ItemA comparison of the efficiency of three models to estimate water yield changes after forested catchment conversion( 1994)This study used Cropper Creek (located southwest of Myrtleford, Victoria, Australia) data to compare the efficiency of a statistically based "black box" model derived from an antecedent precipitation index (API) model, a pseudophysical model (SFB), and a paired catchment regression model in predicting the effects of catchment land use change at times when data from the neighbouring catchment are not available, and to analyse whether the pseudophysical model can be used to estimate changes in the internal hydrology of the treated catchment due to the land use change. Each of the models was viewed as a representative of a different philosophic approach. The three models were calibrated using the pre-treatment record. A recession factor of 0.882 and threshold value of 0.1 mm of the antecedent precipitation index were found to satisfactorily estimate hydrograph response at the Clem Creek catchment during the calibration period. The parameters of SFB model (S = 200 mm, F = 80 mm/day, and B = 0.01 ) during the calibration period were obtained by the exhaustive gridding method in parameter space. The input of the paired catchment regression model included only streamflow of Ella Creek. The calibration paired catchment regression model was found to reliably estimate Clem Creek water yield during the calibration period. The paired catchment regression model gave a satisfactory representation of the hydrograph at Clem Creek during pre-treatment and post-treatment periods. The API and SFB models failed to estimate water yield during the calibration period and to predict water yield changes after forested catchment conversion. This failure is probably due to limitations of the conceptualisation of the process. Variations of seasonal and monthly streamflow, and low flow (December - May) were analysed during the study period. The water yield of Clem Creek increased after the treatment. Relative increases in streamflow of Clem Creek were greater in winter and spring than summer and autumn. Annual average increased streamflow was 51.3 megalitres for the paired catchment regression model after the treatment. Low flow changed after the treatment, but it was insignificant for the three models.
-
ItemEvaluation of the concepts and methods of response farming using crop growth simulation models( 1989)This thesis considers aspects of the application of the CERES-Maize simulation model of the growth and yield of the maize crop to the analysis of "response farming" in the semi-arid region of East Kenya. Response farming comprises a set of management tactics by which sowing density and nitrogen application can be targeted to the yield potential of the season based upon the timing and nature of the opening rains. Sowing density and nitrogen application are set by the timing of the opening rains and then further adjustments are made four to five weeks after sowing. The in-season adjustments involve additional nitrogen fertilizer in seasons that promise high yield potential and crop thinning when yield promises to be low. These combinations have the purpose of maximizing yield in seasons of high potential and minimizing fertilizer and seed input in seasons in which the rainfall is too low for it to have any advantage. The scheme is possible, it is proposed, because the timing and nature of the opening rains are closely correlated with total seasonal rainfall and hence yield potential. The thesis discusses the potential that simulation models hold in the analysis of alternative cropping strategies (Chapter 2). It presents a validation study of the model with experimental data collected in the region (Chapter 3) and discusses the relationships between maize agronomy in the region and the components of response farming (Chapter 4). The tuned model is shown to provide a good explanation of the response of the common cultivar of maize to a range of combinations of water and nitrogen supply and a reasonable capacity to describe the response of the crop to thinning (Chapter 5). The model is then used (Chapter 6) to compare response farming with conventional practices at various levels of inputs. It shows that the common practice of the region by which-crops are sown at low density without fertilizer has the least risk of failure but low expected yields. Fertilizer is required to increase yields, but its use incurs a greater chance of crop failure. Compared to alternative strategies using fixed or variable sowing dates at comparable levels of crop density and nitrogen fertilizer, it is shown that response farming does provide higher mean yield with less chance of failure. It is proposed that continuing experimentation with the current and improved versions of the model have an important role to play in the identification of better management strategies for East Kenyan farmers.