The impact of patchiness on plant-flow interactions and consequences for seagrass habitat suitability: a multiscale study

Abstract

Seagrasses are among the most productive marine ecosystems, providing essential ecological services. As ecosystem engineers, these flowering plants modify their environment through physical, chemical, and biological processes. A key physical mechanism involves stabilizing sediments by reducing near-bed flows, which result from the resistance created by plant stems. Flow attenuation increases sedimentation and reduces erosion of seabed, making seagrasses valuable for coastal protection in the face of climate change. Additional ecological services provided by seagrasses include improving water quality by cycling nutrients, providing habitat and feeding grounds for marine species, and sequestering carbon. However, the ability of seagrasses in providing these ecological services is strong influenced by the extent and spatial distribution of vegetation coverage. Patchiness in seagrass beds alters interactions between seagrass and water flow, with areas of varied flow resistance creating heterogeneous flow and turbulence conditions. This thesis assesses the impacts of patchiness, and in particular, the effect of gaps (bare sediment) in otherwise, continuous vegetation coverage, on plant-flow interactions through field measurements, numerical hydrodynamic modelling, and machine-learning classification methods. A multiscale approach is employed: First, a small-scale, field-based study explores the impacts of a bare patch on flow and turbulence, and the links with the spatial distribution of the surrounding seagrass. Second, a suite of numerical modelling simulations was undertaken to investigate the hydraulic interactions of multiple bare patches in a meadow-scale domain. Finally, estuary-scale species distribution models were developed based on topographic and sediment property variables to evaluate seagrass habitat suitability in a large tidal estuary, while also correlating habitat suitability with seagrass patchiness at a larger spatial scale. The first chapter examines the effects of seagrass patchiness on plant-flow interactions. Data on vegetation cover, bed elevation, and hydrodynamics were collected within and around a bare gap in the natural intertidal seagrass meadows of Tauranga Harbour, New Zealand. The small-scale, high-frequency measurements of flow and turbulence revealed faster flow speeds within bare gaps, and slower flow speeds above the surrounding seagrass, during the incoming flood phase of the tidal cycle. Although flow speeds were slower above the seagrass adjacent to the gap, turbulence, as characterised by the dissipation rate of turbulent kinetic energy, remained elevated above the vegetation These turbulence levels were dominated by down-deceleration events, which directed water and nutrient fluxes downward into the canopy, potentially explaining the presence of denser seagrass along the sides of the gap. We observed a spatial alignment between tidal flows and spatial distribution of seagrass: NDVI-derived maps of seagrass coverage demonstrated that the axis of seagrass continuity coincided with the dominant flow direction during flood tide at both the gap scale (O(100)m) and meadow scale (O(102)m), thus suggesting that small-scale plant-flow interactions are also strongly modulated by large-scale processes, such as tidal asymmetry. These findings suggest a feedback mechanism in which seagrass patchiness and spatial structure are shaped by flow transience, which then also influence ecological conditions associated with the survival and growth of the plants. The work examining the impact of a single bare patch on plant-flow interactions was extended to investigate the collective impacts of multiple bare gaps at the meadow scale. This section used numerical modelling, focusing on unidirectional flows within emergent vegetation. Simulations with single gaps showed that horizontal flows were accelerated within the modelling domain, with the flow acceleration increasing up to a critical gap size. Beyond this critical size, further gap enlargement did not increase flow speeds. Paired-gaps simulations revealed that flow interactions occurred when the gaps were separated by distances less than or equal to two gap diameters, especially when aligned in the streamwise direction, where wakes of accelerated flow interconnected, and flow changes were maximised. Minimal flow changes were observed for laterally aligned gaps. Hydrodynamic simulations involving multiple gaps assessed the effects of fragmentation, defined by the number and size of gaps, on flow at fixed levels of bare bed coverage. The results showed that the flow dynamics within aquatic canopies if also driven by spatial configuration of gaps, rather than bare bed coverage alone. Models with fewer, larger gaps caused greater flow acceleration within the idealised meadow than those with more numerous smaller gaps at the same coverage ratio. These effects were more pronounced at higher bare coverage levels, demonstrating that the arrangement of gaps is one of the driving factors in modifying flow and influencing ecosystem stressors linked to flow modification and habitat degradation. Lastly, as feedback mechanisms are species- and system-specific, understanding the relationship between seagrass distribution and environmental factors across different systems is essential. To address this need, a regional-scale study was conducted to explore how topography and sediment composition determine the distribution and the habitat suitability of the temperate, shortleaf species Zostera muelleri in Tauranga Harbour, New Zealand. Using Random Forest species distribution models, the influences of twenty topographic and substrate variables on seagrass habitat suitability were assessed across a large (~240 km²), predominantly intertidal (~60%) estuary. The models showed strong predictive performance as assessed by a mean area under the curve (AUC) of 0.84. Elevation emerged as the most significant factor (mean importance score = 0.29), followed by sediment composition (mean importance score = 0.17 to 0.22), while other topographic variables had a lesser impact (mean importance score < 0.17). Suitable habitats were predominantly located in sandy areas near mean sea level, within a narrow optimal elevation range. The study also revealed that broad-scale patchiness (minimum patch size = 0.1 ha) was significantly influenced by these environmental factors, with seagrass forming smaller, clustered patches in areas of lower habitat suitability, where patches also exhibited a higher presence of bare gaps. The combined findings of the three chapters highlight the controls of the spatial heterogeneity of aquatic vegetation, and in particular, the presence of gaps on the seagrass dynamics across scales. Understanding the links between spatial distributions and environmental drivers can be used to inform frameworks for targeted conservation and restoration of Zostera muelleri.