Thumbnail Image

Novel approaches for modelling changes in phytoplankton diversity and lake ecosystem function

Ecosystem function represents the collective outcome of many different processes. Function may be interrupted by events that originate from outside a system, influencing biological diversity dynamics. Difficulties in expressing how a system is functioning originate firstly from being able to define a normative status for a dynamic system and secondly from the accuracy of common metrics of biodiversity changes. In this thesis, I used a numerical model and high-frequency ecological observations to express functioning of a system. Chapter 2 used biogeochemical parameter perturbations in a lake ecological model to identify seasonal parameter sensitivity variabilities. A set of internal process parameters of calibrated shallow eutrophic Lake Waahi DYRESM-CAEDYM ecological model was used to apply Monte-Carlo perturbation. Analysis was conducted by examining the collective results variability, a “spread” of the ensemble results from the iteration. The results showed that the spreads were small when lake inflows had high discharge, suggesting that lake internal dynamics had lesser effect on water quality and inflows dominated the system dynamics. Due to the simplicity of the methods, regular use of perturbation methods is suggested to assess model uncertainty and to better understand the model. Chapter 3 used interdisciplinary methods to identify changes in dissolved oxygen (DO) observations caused by biological processes. DO in lakes is a key indicator of ecosystem function. Methods used in this chapter included expert panel decision making, Symbolic Aggregate approXimation (SAX) analysis, and text classification. The use of an expert panel was motivated by the common practice of DO data visual assessment. Variability in experts’ boundaries for data quality were observed by data survey, reinforcing the necessity of robust and reproducible methods for unbiased analysis. Surface DO sensor data from 18 global lakes were used to create day-long data segments. The modelling framework successfully simulated the expert panel decisions on these segments, automatically labelling data to indicate when the signal is likely dominated by biological activities. In Chapter 4, species-neutral biological assemblage metrics were developed to account for phytoplankton changes associated with changes in species abundance. Every species’ population changes were converted into binary metrics (i.e., increases or decreases) to identify the “constituents” of species richness, to allow robust assessments of population dynamics. Four lakes (Lakes Annie, Feeagh, Esthwaite and Mendota) from different regions were analysed. The results showed several previously undocumented features. Species recruitment was proportional to the number of species that were increasing. The number of species that were decreasing did not immediately increase the number of species that went extinct. The rate of increase was logarithmically distributed from the fastest to the slowest growing species, with the distribution shape being strongly influenced by number of species that were increasing. Such species-neutral community metrics, along with abundance distribution and diversity, are helpful to assess mechanistic community ecology models. This thesis provides toolsets useful for future studies to understand relationships between forcing and functioning of ecosystems and changes in biodiversity, by providing means to assess ecosystem function and demonstrating examples of species-neutral community structural changes.
Type of thesis
Muraoka, K. (2019). Novel approaches for modelling changes in phytoplankton diversity and lake ecosystem function (Thesis, Doctor of Philosophy (PhD)). The University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/13155
The University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.