Mapping the ecological and biophysical character of seabed habitats of the Paraninihi Marine Reserve, Taranaki, New Zealand.

Abstract

Habitat mapping is important for determining the spatial distribution of biological and physical components of the seabed. Conventional surveying methods, such as diver or drop camera surveys are time consuming and constrained by factors such as depth, water clarity, currents, and weather conditions, which means that is not practical to survey large tracks of sea floor using these methods. Consequently, a substantial proportion of the world’s seafloor remains undescribed. In recent years, multibeam sonar (MBES) has revolutionised the way we image, map and understand the marine environment. However, the quantitative characterisation of MBES backscatter imagery for seafloor and habitat mapping remains a developing field. This thesis examines the utility of MBES backscatter imagery as a tool for the characterisation and mapping of biogenic habitats. Pariokariwa Reef, located within Paraninihi Marine Reserve, Northern Taranaki, was chosen as the location for this study because it supports a range of distinct habitats (including sponge gardens of unusually high biomass and diversity) against which to assess our ability to use MBES backscatter imagery to recognise biogenic seabed habitats.

This thesis describes the collection of spatially coincident MBES data (bathymetric and backscatter) within Paraninihi Marine Reserve and outlines techniques used to process and transform this data. Acoustic data was used to generate a predictive habitat map that was linked to the habitat classes derived from observations made on Pariokariwa Reef, over fine spatial scales. Results from the survey, showed MBES successfully produces high resolution bathymetric imagery that revealed the reefs unique morphology. The resolution of the backscatter imagery was fine enough to identify four dominant seabed classes on the reef, but not fine enough to accurately map heterogeneous habitat over small spatial scales. Results from the study suggest that image-based backscatter classification shows promise for the interpretation of MBES backscatter data, for the production of habitat maps. However, this study revealed a new challenge associated with habitat mapping, which is acoustic surveying over complex reef topography. Hence for complex or heterogeneous topographies, MBES data must be generated at a finer resolution in order to acquire the same level of detail that is available in predictive habitat models created from acoustic surveys conducted over flat, homogenous terrain.

I also examined the distribution of biological assemblages over a smaller spatial scale, to that examined using MBES. The purpose of this exercise was to test whether the reefs complex terrain influences biological community composition and distribution. Visual imagery obtained from drop camera and scuba diver surveys, revealed heterogeneous habitat over small spatial scales, across the morphology of the reef. Community composition and distribution significantly changed with reef aspect, with percentage sponge and biogenic reef appearing to be significantly higher over the vertical face of the reef, and within reef overhangs. Percentage silt was highest below the reef, and appears to be a dominant environmental factor influencing the composition and distribution of sponge communities on the Pariokariwa Reef. The findings from this study suggest multibeam sonar can be used as a tool to map biogenic seabed habitat. However, there are challenges associated with acoustic seabed classification across complex terrain, and therefore requires in situ surveys, conducted over smaller spatial scales.