Benthic habitat dynamics and models on Australias North West Shelf

As management of marine living resource moves beyond simple single species resource utilisation concerns to ecosystem-based management, consideration of habitat dynamics is becoming an integral part of marine resource management. Previous studies have found that habitat can play a critical role in both single species and community level dynamics of species of commercial concern (Sainsbury, 1987; Sainsbury, 1988; Auster & Malatesta, 1995; Freese et al. 1999; Lindholm et al. 1999; Jackson et al. 2000; Sainsbury et al. 2000). Moreover, benthic habitat is becoming a conservation concern in its own right (Environment Protection and Biodiversity Conservation Act 1999). Useful first steps in understanding local benthic habitat dynamics is to collect observation (preferably through time) of the benthos and then to attempt to create dynamic models that capture the broadscale dynamics of the habitat of interest. Just such an exercise was undertaken for the major benthic habitat types in the North West Shelf of Australia (specifically epibenthic, mainly sponge, habitats, seagrass, macroalgae and mangroves). Between 1983 and 1997 photographic data on benthic habitats were collected on the North West Shelf of Australia by CSIRO Marine Research. These data were used to calculate proportional coverage of small (25 cm) epibenthos on the seabed between depths of 20 and 200 m. These observations and the fisheries effort data for the Taiwanese (1973 to 1981) and domestic fleets (1987 to 1997) were pooled onto a spatial grid of 10 by 10 nautical minutes with a temporal scale of a year. A multivariate analysis of the main factors associated with the distribution of the benthic habitats was undertaken (as a guide for factors to include in the final habitat dynamics model). The observations suggested that there was a strong depth-dependent gradient in the biomass and coverage of benthic habitat, which did not appear to be related to bottom stress, but may have been associated with sediment substrate properties. Given the importance of bottom stress in shaping benthic habitats in many other locations around Australia (Pitcher et al. 2002; Pitcher et al. 2004a; Pitcher et al. 2004b and Phillip England, CSIRO Marine and Atmospheric Research, pers. comm.) it is surprising that the analyses showed it to be a non-significant physical factor in determining proportional coverage on the North West Shelf (NWS). During the model development phase of the study a dynamic age-structured metapopulation model was created. This habitat model includes depth and substrate dependent recruitment, growth natural mortality and removal rates by fishing and cyclones. The parameters used in this model were either taken from literature or estimated by minimising the sum of squares between the observed and estimated proportional coverage. The model results easily reproduced the observed patterns of strongly depth related recruitment. It also showed that trawl fishing effort (both by Taiwanese and domestic fleets) was probably a significant factor in shaping the current distribution of benthic habitats on the NWS. There were issues with the models ability to predict recovery rates that match the empirical data. This is almost undoubtedly the result of poorly spatially resolved historical catch time series and a too coarse model resolution. Recasting future analyses and modelling efforts on finer (or more irregular) grids should go a long way to rectifying these issues. Nevertheless, even as is, the model still performs acceptably, particularly within an MSE framework. The bulk of the data (and subsequent modelling efforts) dealt with epibenthic (mainly sponge) habitats. The same model was also applied (in a more limited extent) to seagrass, macroalgae and mangroves. There was substantially less data available for these groups and the models were parameterised from the literature and expert knowledge.随着海洋生物资源管理从单纯的单一物种资源利用考量转向基于生态系统的管理，生境动态（habitat dynamics）的考量正成为海洋资源管理不可或缺的组成部分。此前已有研究表明，生境在具有经济价值的物种的单物种种群动态与群落水平动态中均发挥关键作用（Sainsbury, 1987; Sainsbury, 1988; Auster & Malatesta, 1995; Freese et al. 1999; Lindholm et al. 1999; Jackson et al. 2000; Sainsbury et al. 2000）。此外，底栖生境（benthic habitat）本身也已成为独立的保护议题（《1999年环境保护与生物多样性保护法》（Environment Protection and Biodiversity Conservation Act 1999））。理解局地底栖生境动态的有效首要步骤是收集底栖生物的观测数据（最好具备时间序列），并尝试构建能够捕捉目标生境大范围动态的动态模型。针对澳大利亚西北陆架的主要底栖生境类型——具体为表栖（主要为海绵）生境、海草、大型藻类与红树林，研究团队就此开展了相关研究工作。1983年至1997年间，CSIRO海洋研究（CSIRO Marine Research）在澳大利亚西北陆架收集了底栖生境的摄影数据。这些数据被用于计算20米至200米水深范围内，海床上小型（25厘米）表栖生物的占比覆盖率。前述观测数据，以及中国台湾船队（1973年至1981年）与澳大利亚本土船队（1987年至1997年）的捕捞努力量数据，被整合至10×10航海分的空间网格中，时间分辨率为1年。研究团队开展了与底栖生境分布相关的主要影响因素的多元分析，以此作为最终生境动态模型中纳入影响因素的参考依据。观测结果显示，底栖生境的生物量与覆盖率存在显著的深度依赖梯度，该梯度似乎与底应力（bottom stress）无关，但可能与沉积物底质属性相关。考虑到底应力在澳大利亚其他众多区域塑造底栖生境过程中的重要性（Pitcher et al. 2002; Pitcher et al. 2004a; Pitcher et al. 2004b以及Phillip England, CSIRO海洋与大气研究中心，私人通信），分析结果显示其在澳大利亚西北陆架（North West Shelf, NWS）的覆盖率占比预测中并非显著物理因素，这一点令人意外。在本研究的模型开发阶段，团队构建了动态年龄结构集合种群模型（metapopulation model）。该生境模型纳入了依赖于深度与底质的种群补充、生长、自然死亡率，以及捕捞与气旋引发的移除率。模型中使用的参数要么取自公开文献，要么通过最小化观测覆盖率与估算覆盖率之间的残差平方和进行估算。模型结果成功复现了观测到的深度依赖型种群补充模式。研究还表明，拖网捕捞努力量（包括中国台湾船队与本土船队）可能是塑造澳大利亚西北陆架当前底栖生境分布的重要因素。模型在预测恢复速率方面存在与经验数据不符的问题，这几乎无疑是由于历史捕捞时间序列的空间分辨率不足，以及模型分辨率过于粗糙所致。未来采用更精细（或更不规则）的网格开展分析与建模工作，将能在很大程度上解决这些问题。尽管如此，即便以当前形式，该模型仍表现良好，尤其是在管理策略评估（Management Strategy Evaluation, MSE）框架中。本研究的大部分数据及后续建模工作聚焦于表栖（主要为海绵）生境。同一模型也被在更有限的范围内应用于海草、大型藻类与红树林生境。针对这些类别的可用数据相对较少，模型参数通过文献与专家知识进行设定。