An evaluation of the evidence for linkages between mangroves and fisheries: A synthesis of the literature and identification of research directions

There is a widely held paradigm that mangroves are critical for sustaining production in coastal fisheries through their role as important nursery areas for fisheries species. This paradigm frequently forms the basis for important management decisions on habitat conservation and restoration of mangroves and other coastal wetlands. This paper reviews the current status of the paradigm and synthesises the information on the processes underlying these potential links. In the past, the paradigm has been supported by studies identifying correlations between the areal and linear extent of mangroves and fisheries catch. This paper goes beyond the correlative approach to develop a new framework on which future evaluations can be based. First, the review identifies what type of marine animals are using mangroves and at what life stages. These species can be categorised as estuarine residents, marine-estuarine species and marine stragglers. The marine-estuarine category includes many commercial species that use mangrove habitats as nurseries. The second stage is to determine why these species are using mangroves as nurseries. The three main proposals are that mangroves provide a refuge from predators, high levels of nutrients and shelter from physical disturbances. The recognition of the important attributes of mangrove nurseries then allows an evaluation of how changes in mangroves will affect the associated fauna. Surprisingly few studies have addressed this question. Consequently, it is difficult to predict how changes in any of these mangrove attributes would affect the faunal communities within them and, ultimately, influence the fisheries associated with them. From the information available, it seems likely that reductions in mangrove habitat complexity would reduce the biodiversity and abundance of the associated fauna, and these changes have the potential to cause cascading effects at higher trophic levels with possible consequences for fisheries. Finally, there is a discussion of the data that are currently available on mangrove distribution and fisheries catch, the limitations of these data and how best to use the data to understand mangrove-fisheries links and, ultimately, to optimise habitat and fisheries management. Examples are drawn from two relatively data-rich regions, Moreton Bay (Australia) and Western Peninsular Malaysia, to illustrate the data needs and research requirements for investigating the mangrove-fisheries paradigm. Having reliable and accurate data at appropriate spatial and temporal scales is crucial for mangrove-fisheries investigations. Recommendations are made for improvements to data collection methods that would meet these important criteria. This review provides a framework on which to base future investigations of mangrove-fisheries links, based on an understanding of the underlying processes and the need for rigorous data collection. Without this information, the understanding of the relationship between mangroves and fisheries will remain limited. Future investigations of mangrove-fisheries links must take this into account in order to have a good ecological basis and to provide better information and understanding to both fisheries and conservation managers.

The application of adaptive partitioned random search in feature selection problem

Feature selection is one of important and frequently used techniques in data preprocessing. It can improve the efficiency and the effectiveness of data mining by reducing the dimensions of feature space and removing the irrelevant and redundant information. Feature selection can be viewed as a global optimization problem of finding a minimum set of M relevant features that describes the dataset as well as the original N attributes. In this paper, we apply the adaptive partitioned random search strategy into our feature selection algorithm. Under this search strategy, the partition structure and evaluation function is proposed for feature selection problem. This algorithm ensures the global optimal solution in theory and avoids complete randomness in search direction. The good property of our algorithm is shown through the theoretical analysis.

Environmental conditions and variation in levels of sun exposure among children in child care

Sun exposure in childhood is I of the risk factors for developing skin cancer, yet little is known about levels of exposure at this age. This is particularly important in countries with high levels of ultraviolet radiation. (UVR) such as Australia. Among 49 children 3 to 5 years of age attending child care centers, UVR exposure was studied under 4 conditions in a repeated measures design; sunny days, cloudy days, teacher's instruction to stay in the shade, and a health professionals instruction to apply sunscreen. Three different data collection methods were employed: (a) completion of questionnaire or diary by parents and researcher, (b) polysulphone dosimeter readings, and (c) observational audits (video recording). Results of this study indicated that more than half the children had been sunburnt (pink or red) and more than a third had experienced painful sunburn (sore or tender) in the last summer. Most wore short sleeve shirts, short skirts or shorts and cap, that do not provide optimal levels of skin protection. However, sunscreen was applied to all exposed parts before the children went out to the playground. Over the period of I hr (9-10 a.m.) the average amount of time children spent in full sun was 22 min. On sunny days there was more variation across children in the amount of sun exposure received. While the potential amount of UVR exposure for young children during the hour they were outside on a sunny day was 1.45 MED (Minimum Erythemal Dose), they received on average 0.35 MED, which is an insufficient amount to result in an erythemal response on fair skin even without the use of sunscreen.