Особливості використання статистичних методів у дослідженнях макробезхребетних піщаних пляжів Північно-Західної частини Чорного моря (огляд)
Вантажиться...
Дата
2019
Науковий керівник
Укладач
Редактор
Назва журналу
Номер ISSN
Номер E-ISSN
Назва тому
Видавець
Одеський національний університет імені І. І. Мечникова
Анотація
У статті розглянуто особливості дизайну відбору проб макробезхребетних
піщаних пляжів Північно-Західної частини Чорного моря і статистичного
опрацювання отриманих даних. Особливу увагу приділено актуальній і
маловивченій темі оцінці просторового розподілу цих організмів. Розглянуто
актуальні методи оцінки чисельності видів і впливу факторів середовища.
Relevance. The feature of the sandy beaches as the marginal biotope complex is that most of the environmental gradients act across the coastline. The low amplitude of the tides and the non-periodic wind tides determine the dynamics and specificity of these zones in the north-western Black Sea (NWBS). Because of this, the macroinvertebrate sampling design in the NWBS and the statistical processing of the results need special attention. Aim. The purpose of this paper is to review the methods of statistical data processing that will allow us to clarify and extend our understanding of the structure of macro-invertebrate sand beach communities of NWBS and the features of their use. Main results. The shape of the sampling device slightly affects the catch of macrofauna. It is acceptable to use both transects and randomized stratified sampling. A reliable estimation of quantity is provided by kriging methods and recalculation of density per length of the coastline. The coefficient b of the Taylor equation and the regression of mean crowding characterize species-specific aggregation. For estimation of distribution it is recommended to use spatial autocorrelation coefficients (Moran's I, Geary’s C), Ripley’s K function, JCA (for binary data) and SADIE. The relationship between the spatial distributions of species from one sampling station can be detected by the Mantel test. The biological diversity of macroinvertebrates of NWBS sandy beaches is relatively low, so unspecific fauna can significantly affect the value of indices. Influence of environmental factors on macroinvertebrates, their groups and communities can be measured using regression analysis or more specific methods. Suggestions. It is recommended to use a randomized stratified sampling design. Species-specific aggregation parameters should be evaluated by density-independent indices. Geostatistical methods are recommended for estimating the spatial distribution. It is suggested to estimate the number by the results, interpolated by kriging methods. For calculation of biodiversity, it is recommended to exclude randomly introduced species from the sample. It is recommended to use logistic, multiple linear regression and Mantel test to evaluate the impact of environmental factors. The analysis should evaluate the effect of factors at different time scales and use beach indices to characterize beaches.
Relevance. The feature of the sandy beaches as the marginal biotope complex is that most of the environmental gradients act across the coastline. The low amplitude of the tides and the non-periodic wind tides determine the dynamics and specificity of these zones in the north-western Black Sea (NWBS). Because of this, the macroinvertebrate sampling design in the NWBS and the statistical processing of the results need special attention. Aim. The purpose of this paper is to review the methods of statistical data processing that will allow us to clarify and extend our understanding of the structure of macro-invertebrate sand beach communities of NWBS and the features of their use. Main results. The shape of the sampling device slightly affects the catch of macrofauna. It is acceptable to use both transects and randomized stratified sampling. A reliable estimation of quantity is provided by kriging methods and recalculation of density per length of the coastline. The coefficient b of the Taylor equation and the regression of mean crowding characterize species-specific aggregation. For estimation of distribution it is recommended to use spatial autocorrelation coefficients (Moran's I, Geary’s C), Ripley’s K function, JCA (for binary data) and SADIE. The relationship between the spatial distributions of species from one sampling station can be detected by the Mantel test. The biological diversity of macroinvertebrates of NWBS sandy beaches is relatively low, so unspecific fauna can significantly affect the value of indices. Influence of environmental factors on macroinvertebrates, their groups and communities can be measured using regression analysis or more specific methods. Suggestions. It is recommended to use a randomized stratified sampling design. Species-specific aggregation parameters should be evaluated by density-independent indices. Geostatistical methods are recommended for estimating the spatial distribution. It is suggested to estimate the number by the results, interpolated by kriging methods. For calculation of biodiversity, it is recommended to exclude randomly introduced species from the sample. It is recommended to use logistic, multiple linear regression and Mantel test to evaluate the impact of environmental factors. The analysis should evaluate the effect of factors at different time scales and use beach indices to characterize beaches.
Опис
Ключові слова
просторовий розподіл, дизайн відбору проб, кількіснабіологія, spatial distribution, sampling design, quantitative biology
Бібліографічний опис
Вісник Одеського національного університету = Odesa National University Herald