WEATHERING THE STORM: THE IMPLICATIONS OF WAVE EXPOSURE ON THE DISTRIBUTION, PHENOTYPE AND GROWTH OF A TEMPERATE REEF FISH

<p>Disturbance is a fundamental process that affects the structure and dynamics of populations. Wave action is an important agent of disturbance in coastal marine systems, and the frequency and severity of wave-associated disturbances is forecasted to increase with climate change. Understanding the effects of waves on coastal marine ecosystems, and the ability of organisms to adapt to wave action, is of growing importance. This is particularly true for intertidal/shallow subtidal species that are subjected to varying, sometimes intense, wave action. Most studies to-date have focused on species with limited mobility (e.g., algae and invertebrates), and have used estimates of wave dynamics that are not always relevant to the spatial scales of these organisms and their home ranges. My thesis focuses on the common triplefin, Forsterygion lapillum, an abundant benthic marine fish inhabiting shallow subtidal and intertidal rocky reefs throughout New Zealand. I develop and implement a protocol to characterise wave climates on an ecologically relevant scale. I evaluate the effects of waves on abundance, phenotype, performance, and behaviour of a reef fish. In Chapter 2, I develop and implement a protocol to characterise wave climate at an appropriate scale. The Wellington south coast is exposed to storm waves that develop in the Southern Ocean and propagate up the east coast of New Zealand. I deployed low-cost HOBO acceleration loggers at two depths within each of six locations along the Wellington south coast to record a time series of wave action at twelve sites. Data from my loggers showed substantial spatial and temporal variation in water acceleration due to interactions between waves and local topography. I used a clustering analysis to characterise my 12 sites as either ‘exposed’ or ‘sheltered’. Assignments to these exposure categories did not match with a priori predictions of exposure, suggesting that wave forces experienced by organisms in the shallow subtidal environment may be difficult to assess from surface-based observations of waves. Data were generally well-correlated with an offshore buoy at all sites, and these correlations were stronger for more exposed sites. In Chapter 3, I explored variation in fish density and phenotype through time and as a function of wave exposure. Densities peaked in summer (corresponding to seasonal recruitment) and declined over winter (consistent with increased losses during high-wave periods), and were generally greater at sheltered locations. While body condition was generally highest for fish sampled from exposed sites (consistent with a density-dependent effect on condition and/or enhancement of foraging with increasing water acceleration), other morphological characteristics did not consistently vary with wave exposure. In Chapter 4, I used otoliths to reconstruct of growth histories of individuals to further elucidate the influence of wave exposure on triplefin phenotypes. Recent growth was not influenced by wave exposure, but this was confounded by strong seasonal variation in growth rates. Lifetime growth rate also did not differ with wave exposure, and was strongly influenced by hatch date. I used mixed effects models to appropriately account for the potentially confounding effects of other features on growth, and found that daily growth rates were slightly positively correlated with site-specific daily measures of wave action. This result can potentially account for the elevated body condition of fish at exposed sites (Chapter 3), and it has important implications for fish inhabiting wave exposed coasts. In Chapter 5, I conducted a lab experiment to evaluate feeding ability in relation to simulated wave action. I used fish of a range of sizes, sampled from either a wave-sheltered or a wave-exposed site, and measured their consumption of prey in calm (low flow) conditions, disturbance (high flow) conditions, and immediately following a period of disturbance. Fish consumed fewer prey during disturbance, and more prey during calm conditions (and a similar consumption rate was observed for fish that were assayed after a period of intense wave action). While this pattern held for fish sampled from both populations, fish from wave-exposed sites consumed more prey than fish from sheltered sites, suggesting phenotypic traits (e.g., behavioural or morphological) that shape their feeding efficiency. Collectively my results suggest that organisms that inhabit wave-exposed coastlines may be intimately linked to wave climate. Waves may have direct effects on numbers (reducing densities via induced mortality) and/or indirect effects on the traits, foraging opportunities, and/or body condition of survivors. Species such as the common triplefin may exhibit plasticity in phenotypic traits that enable them to adapt to dynamic and unpredictable environments. Overall, this thesis provides insight into the ability of an intertidal/shallow subtidal species to cope with variable wave action. Such species may exhibit resilience with increasing wave action due to climate change.</p>

WEATHERING THE STORM: THE IMPLICATIONS OF WAVE EXPOSURE ON THE DISTRIBUTION, PHENOTYPE AND GROWTH OF A TEMPERATE REEF FISH

<p>Disturbance is a fundamental process that affects the structure and dynamics of populations. Wave action is an important agent of disturbance in coastal marine systems, and the frequency and severity of wave-associated disturbances is forecasted to increase with climate change. Understanding the effects of waves on coastal marine ecosystems, and the ability of organisms to adapt to wave action, is of growing importance. This is particularly true for intertidal/shallow subtidal species that are subjected to varying, sometimes intense, wave action. Most studies to-date have focused on species with limited mobility (e.g., algae and invertebrates), and have used estimates of wave dynamics that are not always relevant to the spatial scales of these organisms and their home ranges. My thesis focuses on the common triplefin, Forsterygion lapillum, an abundant benthic marine fish inhabiting shallow subtidal and intertidal rocky reefs throughout New Zealand. I develop and implement a protocol to characterise wave climates on an ecologically relevant scale. I evaluate the effects of waves on abundance, phenotype, performance, and behaviour of a reef fish. In Chapter 2, I develop and implement a protocol to characterise wave climate at an appropriate scale. The Wellington south coast is exposed to storm waves that develop in the Southern Ocean and propagate up the east coast of New Zealand. I deployed low-cost HOBO acceleration loggers at two depths within each of six locations along the Wellington south coast to record a time series of wave action at twelve sites. Data from my loggers showed substantial spatial and temporal variation in water acceleration due to interactions between waves and local topography. I used a clustering analysis to characterise my 12 sites as either ‘exposed’ or ‘sheltered’. Assignments to these exposure categories did not match with a priori predictions of exposure, suggesting that wave forces experienced by organisms in the shallow subtidal environment may be difficult to assess from surface-based observations of waves. Data were generally well-correlated with an offshore buoy at all sites, and these correlations were stronger for more exposed sites. In Chapter 3, I explored variation in fish density and phenotype through time and as a function of wave exposure. Densities peaked in summer (corresponding to seasonal recruitment) and declined over winter (consistent with increased losses during high-wave periods), and were generally greater at sheltered locations. While body condition was generally highest for fish sampled from exposed sites (consistent with a density-dependent effect on condition and/or enhancement of foraging with increasing water acceleration), other morphological characteristics did not consistently vary with wave exposure. In Chapter 4, I used otoliths to reconstruct of growth histories of individuals to further elucidate the influence of wave exposure on triplefin phenotypes. Recent growth was not influenced by wave exposure, but this was confounded by strong seasonal variation in growth rates. Lifetime growth rate also did not differ with wave exposure, and was strongly influenced by hatch date. I used mixed effects models to appropriately account for the potentially confounding effects of other features on growth, and found that daily growth rates were slightly positively correlated with site-specific daily measures of wave action. This result can potentially account for the elevated body condition of fish at exposed sites (Chapter 3), and it has important implications for fish inhabiting wave exposed coasts. In Chapter 5, I conducted a lab experiment to evaluate feeding ability in relation to simulated wave action. I used fish of a range of sizes, sampled from either a wave-sheltered or a wave-exposed site, and measured their consumption of prey in calm (low flow) conditions, disturbance (high flow) conditions, and immediately following a period of disturbance. Fish consumed fewer prey during disturbance, and more prey during calm conditions (and a similar consumption rate was observed for fish that were assayed after a period of intense wave action). While this pattern held for fish sampled from both populations, fish from wave-exposed sites consumed more prey than fish from sheltered sites, suggesting phenotypic traits (e.g., behavioural or morphological) that shape their feeding efficiency. Collectively my results suggest that organisms that inhabit wave-exposed coastlines may be intimately linked to wave climate. Waves may have direct effects on numbers (reducing densities via induced mortality) and/or indirect effects on the traits, foraging opportunities, and/or body condition of survivors. Species such as the common triplefin may exhibit plasticity in phenotypic traits that enable them to adapt to dynamic and unpredictable environments. Overall, this thesis provides insight into the ability of an intertidal/shallow subtidal species to cope with variable wave action. Such species may exhibit resilience with increasing wave action due to climate change.</p>

Designing accurate and effective means for marine ecosystem monitoring incorporating species distribution assessments

<p>Monitoring marine ecosystems is essential for the conservation and management of marine biodiversity as it is central to the development of sustainable management practices and for assessing the effectiveness of the increasing number of marine reserves (MR) globally. Monitoring data are often collected in MRs to assess the state of natural marine systems in the absence of anthropogenic disturbance or to assess recovery of previously impacted species. In recent years, MR designation has attempted to move away from ad hoc approaches to MR establishment and towards using existing species distribution and abundance data to define protected areas. Given the logistics and cost of collecting biological data in the marine environment, effective methods are required to successfully demonstrate changes associated with MRs and to identify the spatial distribution of organisms and habitats for the planning of further MRs. The aim of this thesis was to identify effective protocols for the monitoring of fish and invertebrate species inside MRs in New Zealand, and to develop and apply methodologies to identify spatial distribution patterns relevant to marine spatial planning. Using baseline data of fish and invertebrate species abundances for the Taputeranga MR I performed prospective power analyses to identify the most cost-effective monitoring approach for subsequent monitoring. Based on before-after-control-impact (BACI) tests the power to conclude statistically that abundances were higher at MR sites was low for even large simulated changes in abundance (two-fold or four-fold increases) for most species. Due to differences in baseline abundance and spatio-temporal variance terms, power varied considerably among species, highlighting the difficulty of monitoring all species to the same degree, whilst also remaining cost-effective. Furthermore, the results highlight the need for temporally replicated survey designs as “one-off” surveys had much lower power than those that were temporally replicated. Longer term monitoring effectiveness was analysed using three long-term datasets from MRs in the South Island of New Zealand. I analysed the power of alternate underwater visual census (UVC) monitoring configurations to conclude statistically that there were increasing/decreasing trends in abundance, as well as the precision and accuracy of trend estimates. Overall even the highest replication designs considered had low power (< 80%) to conclude there was a non-zero trend even when simulated data represented trends equivalent to the population doubling or halving over ten years. The most cost-effective monitoring design varied among species and MRs, further highlighting that monitoring choices need to be location- and species-specific. A general finding, however, was that increasing the number of sites was almost always more beneficial than increasing the number of transects per site. Based on these results, I recommend that monitoring design planning focuses more specifically on assessments of precision and accuracy of estimated parameters, with less focus on power, as this places greater emphasis on interpreting monitoring data in terms of potential biological significance rather than testing for statistical significance. Monitoring can never achieve complete coverage of large areas therefore methods for extrapolating or predicting species or habitats to un-surveyed locations are necessary for evaluating large-scale spatial distributions. To address this I used modelling techniques to identify the spatial variation in species and habitats along the Wellington south coast, with a particular focus on elucidating the potential and realised effects of wave exposure. A wave simulation model (SWAN) was used to identify the spatial variation in wave exposure relevant to intertidal and subtidal communities. In particular the spatial variation in wave forces was compared to the distribution of two subtidal macroalgal species, Macrocystis pyrifera and Ecklonia radiata, taking into consideration the biomechanical thresholds of damage for these plants. Despite considerable wave forces during winter storms, healthy E. radiata is unlikely to be damaged, whilst larger (>15 m stipe length) M. pyrifera plants are likely to be damaged in certain locations dependent on local sheltering effects. Furthermore, the distribution of M. pyrifera from aerial imagery coincided with areas that were predicted to have lower wave forces, suggesting that the distribution of M. pyrifera may be related to wave exposure. I subsequently constructed species distribution models revealing the relationship between intertidal species distributions and environmental factors, as a predictive baseline of the current distributions of species. The abundances of Chamaesipho barnacle species were found to be best described by wave exposure, with increased cover correlated with increasing wave exposure, while contrasting patterns were observed for C. brunnea and C. columna with respect to distance from the harbour entrance, suggesting differential larval supply or differential responses to changing water column characteristics. Macroalgal assemblage composition was explained predominantly by wave exposure, with a rich macroalgal assemblage at the less exposed locations, and more exposed locations exhibiting a community consisting of coralline algal species and the large brown alga Durvillaea antarctica. The predictive models were then used to predict species distributions for a section of coastline demonstrating how this form of modelling can be used to maximise the potential of monitoring data. Finally, a literature keyword search along with methodological developments and results from previous chapters are used in the final chapter to develop a framework for the collection of data from the planning phase all the way through to long-term monitoring of MRs.</p>

Designing accurate and effective means for marine ecosystem monitoring incorporating species distribution assessments

<p>Monitoring marine ecosystems is essential for the conservation and management of marine biodiversity as it is central to the development of sustainable management practices and for assessing the effectiveness of the increasing number of marine reserves (MR) globally. Monitoring data are often collected in MRs to assess the state of natural marine systems in the absence of anthropogenic disturbance or to assess recovery of previously impacted species. In recent years, MR designation has attempted to move away from ad hoc approaches to MR establishment and towards using existing species distribution and abundance data to define protected areas. Given the logistics and cost of collecting biological data in the marine environment, effective methods are required to successfully demonstrate changes associated with MRs and to identify the spatial distribution of organisms and habitats for the planning of further MRs. The aim of this thesis was to identify effective protocols for the monitoring of fish and invertebrate species inside MRs in New Zealand, and to develop and apply methodologies to identify spatial distribution patterns relevant to marine spatial planning. Using baseline data of fish and invertebrate species abundances for the Taputeranga MR I performed prospective power analyses to identify the most cost-effective monitoring approach for subsequent monitoring. Based on before-after-control-impact (BACI) tests the power to conclude statistically that abundances were higher at MR sites was low for even large simulated changes in abundance (two-fold or four-fold increases) for most species. Due to differences in baseline abundance and spatio-temporal variance terms, power varied considerably among species, highlighting the difficulty of monitoring all species to the same degree, whilst also remaining cost-effective. Furthermore, the results highlight the need for temporally replicated survey designs as “one-off” surveys had much lower power than those that were temporally replicated. Longer term monitoring effectiveness was analysed using three long-term datasets from MRs in the South Island of New Zealand. I analysed the power of alternate underwater visual census (UVC) monitoring configurations to conclude statistically that there were increasing/decreasing trends in abundance, as well as the precision and accuracy of trend estimates. Overall even the highest replication designs considered had low power (< 80%) to conclude there was a non-zero trend even when simulated data represented trends equivalent to the population doubling or halving over ten years. The most cost-effective monitoring design varied among species and MRs, further highlighting that monitoring choices need to be location- and species-specific. A general finding, however, was that increasing the number of sites was almost always more beneficial than increasing the number of transects per site. Based on these results, I recommend that monitoring design planning focuses more specifically on assessments of precision and accuracy of estimated parameters, with less focus on power, as this places greater emphasis on interpreting monitoring data in terms of potential biological significance rather than testing for statistical significance. Monitoring can never achieve complete coverage of large areas therefore methods for extrapolating or predicting species or habitats to un-surveyed locations are necessary for evaluating large-scale spatial distributions. To address this I used modelling techniques to identify the spatial variation in species and habitats along the Wellington south coast, with a particular focus on elucidating the potential and realised effects of wave exposure. A wave simulation model (SWAN) was used to identify the spatial variation in wave exposure relevant to intertidal and subtidal communities. In particular the spatial variation in wave forces was compared to the distribution of two subtidal macroalgal species, Macrocystis pyrifera and Ecklonia radiata, taking into consideration the biomechanical thresholds of damage for these plants. Despite considerable wave forces during winter storms, healthy E. radiata is unlikely to be damaged, whilst larger (>15 m stipe length) M. pyrifera plants are likely to be damaged in certain locations dependent on local sheltering effects. Furthermore, the distribution of M. pyrifera from aerial imagery coincided with areas that were predicted to have lower wave forces, suggesting that the distribution of M. pyrifera may be related to wave exposure. I subsequently constructed species distribution models revealing the relationship between intertidal species distributions and environmental factors, as a predictive baseline of the current distributions of species. The abundances of Chamaesipho barnacle species were found to be best described by wave exposure, with increased cover correlated with increasing wave exposure, while contrasting patterns were observed for C. brunnea and C. columna with respect to distance from the harbour entrance, suggesting differential larval supply or differential responses to changing water column characteristics. Macroalgal assemblage composition was explained predominantly by wave exposure, with a rich macroalgal assemblage at the less exposed locations, and more exposed locations exhibiting a community consisting of coralline algal species and the large brown alga Durvillaea antarctica. The predictive models were then used to predict species distributions for a section of coastline demonstrating how this form of modelling can be used to maximise the potential of monitoring data. Finally, a literature keyword search along with methodological developments and results from previous chapters are used in the final chapter to develop a framework for the collection of data from the planning phase all the way through to long-term monitoring of MRs.</p>

Hybridisation in the Brown Alga Carpophyllum: Investigating Morphology, Distribution and Wave Exposure

<p>Hybridisation can result in new hybrid lineages, parental species extinctions, the transfer of adaptations, or the merging of parental lineages. Subsequently hybridisation has important implications for the species involved. Hybridisation has recently been confirmed between the Fucalean brown algae Carpophyllum angustifolium and Carpophyllum maschalocarpum using the ITS2 marker. This study conducted a detailed morphometric analysis combined with molecular data to investigate morphology distribution and exposure at two sites on the East Cape. Hybridisation was also morphologically investigated at Leigh, where the previous work had been unable to resolve hybrids using the ITS2 marker. Carpophyllum angustifolium, C. maschalocarpum and their hybrids had distinct and intermediate morphologies, and could be identified by stipe width alone. Individuals with hybrid genotypes with distinctive C. angustifolium morphotypes were also found, which suggests asymmetrical introgression is occurring. Some aspects of C. angustifolium and C. maschalocarpum morphology were found to be correlated with wave exposure. In the more exposed zones C. angustifolium individuals were longer, while C. maschalocarpum individuals were shorter, had thinner stipes and less frequent vesicle presence. There were also nonsignificant trends of C. maschalocarpum individuals having thinner lamina, and lower branch presence in higher wave exposures. The distributions of C. angustifolium, C. maschalocarpum and their hybrids were found to be correlated with exposure. Carpophyllum angustifolium was distributed only in the relatively exposed zones, while C. maschalocarpum was distributed mainly in the more sheltered zones. Hybrids were distributed in intermediate exposure zones where both parental species were present. The hybrid distributions could be a reflection of environmental selection or of the parental contact zone. Morphological evidence was found for hybridisation at Leigh, although there were differences between the morphologies of East Cape and Leigh clusters of C. angustifolium and hybrids. These differences could be due to environmental differences, genetic differentiation or different levels of introgression between the two locations. The general findings in this study support the existing literature on hybridisation, which mainly comes from terrestrial plant and animal species complexes.</p>

Hybridisation in the Brown Alga Carpophyllum: Investigating Morphology, Distribution and Wave Exposure

<p>Hybridisation can result in new hybrid lineages, parental species extinctions, the transfer of adaptations, or the merging of parental lineages. Subsequently hybridisation has important implications for the species involved. Hybridisation has recently been confirmed between the Fucalean brown algae Carpophyllum angustifolium and Carpophyllum maschalocarpum using the ITS2 marker. This study conducted a detailed morphometric analysis combined with molecular data to investigate morphology distribution and exposure at two sites on the East Cape. Hybridisation was also morphologically investigated at Leigh, where the previous work had been unable to resolve hybrids using the ITS2 marker. Carpophyllum angustifolium, C. maschalocarpum and their hybrids had distinct and intermediate morphologies, and could be identified by stipe width alone. Individuals with hybrid genotypes with distinctive C. angustifolium morphotypes were also found, which suggests asymmetrical introgression is occurring. Some aspects of C. angustifolium and C. maschalocarpum morphology were found to be correlated with wave exposure. In the more exposed zones C. angustifolium individuals were longer, while C. maschalocarpum individuals were shorter, had thinner stipes and less frequent vesicle presence. There were also nonsignificant trends of C. maschalocarpum individuals having thinner lamina, and lower branch presence in higher wave exposures. The distributions of C. angustifolium, C. maschalocarpum and their hybrids were found to be correlated with exposure. Carpophyllum angustifolium was distributed only in the relatively exposed zones, while C. maschalocarpum was distributed mainly in the more sheltered zones. Hybrids were distributed in intermediate exposure zones where both parental species were present. The hybrid distributions could be a reflection of environmental selection or of the parental contact zone. Morphological evidence was found for hybridisation at Leigh, although there were differences between the morphologies of East Cape and Leigh clusters of C. angustifolium and hybrids. These differences could be due to environmental differences, genetic differentiation or different levels of introgression between the two locations. The general findings in this study support the existing literature on hybridisation, which mainly comes from terrestrial plant and animal species complexes.</p>