Methods for the Study of Marine Benthos

Anastasios Eleftheriou, Anastasios Eleftheriou

• Verlag: Wiley-Blackwell
• Jahr: 2013
• ISBN:
  9781118542378
• Auflage: 4
• Thema: Technik & Maschinenbau
• Thema: Fischerei & Aquakultur

The continuing global decline of the health of the sea, and the increasing depletion of marine resources and biodiversity, caused by human activity and climate change, have led to ever-increasing international concern. These changes in the marine environment highlight the importance of effective monitoring of the ecology of the benthos which has been shown to be a sensitive index of such alterations.

Completely revised and updated to include many new methods and technologies, this Fourth Edition of Methods for the Study of Marine Benthos provides comprehensive coverage on the tools and techniques available to those working in the area. Commencing with an overview of the design and analysis of benthic surveys, the book continues with chapters covering the sedimentary environment, imaging and diving techniques, macro- and meiofauna techniques, deep sea sampling, energy flow and production. An additional new chapter provided in this edition covers phytobenthos techniques.

Written by many of the world's leading authorities in marine sampling techniques and use, and edited by Professor Anastasios Eleftheriou, this comprehensive Fourth Edition is an essential tool for all marine and environmental scientists, ecologists, fisheries workers and oceanographers. Libraries in all research establishments and universities where these subjects are studied and taught will find this book to be a hugely valuable addition to their collections.

Chapter 1

Design and Analysis in Benthic Surveys in Environmental Sampling

Antony J. Underwood and Maura G. Chapman

Centre for Research on Ecological Impacts of Coastal Cities, University of Sydney, Sydney, Australia

Abstract:
Measuring environmental impacts affecting benthic habitats requires detection of specific patterns of statistical interactions in data sampled before and after a potential impact, in the potentially impacted place and in control or reference locations. This is complex because ecological assemblages and populations vary at many spatial and temporal scales. Here, we introduce methods to ensure appropriate, independent replication of sampling at hierarchical scales in space and time. For statistical analysis, the logic of sampling design is critical. Determining precision of estimates and maximising power to detect impacts require care in the design, analysis and interpretation of the relevant data.

Keywords:
benthic variance, environmental impact, environmental sampling, independence, precautionary principle, precision, replication, scale, sampling design, statistical interaction

1.1 Introduction

Quantified sampling, particularly to test applied and logically structured hypotheses about patterns and processes in marine habitats, is of increasing importance in underpinning the understanding of natural processes and to predict changes in response to environmental influences. There have been numerous advances in soul-searching (Peters, 1991), methods of analysis (Clarke, 1993; Anderson, 2001), commentaries on logic (Underwood, 1990; Resetarits & Bernardo, 1998; Lawton, 1999) and the need for better understanding of environmental impacts (Schmitt & Osenberg, 1996; Sparks, 2000). As a result, it is not possible in a general summary to review comprehensively even the new material, let alone everything relevant to the topic of improved benthic sampling. Suffice it to say that it is also essential to help with management and conservation of diversity, natural resources, systems and functions. Taking care in the acquisition of quantitative information should, therefore, be of paramount importance to all marine biologists and ecologists.

This chapter, therefore, presents a general overview of the issues concerned. It is not, nor could it be, a ‘cookbook’ of procedures that might work. Rather, it is an attempt to consider fundamental issues of replication, in space and time, the nature of variables examined, issues about designing comparative sampling programmes and so forth.

These topics are considered against a general background of the logical structure underpinning sampling methodology. The issue is a simple one – unless the aims and objectives of any study are clearly identified at the outset, the least damaging outcome will be wastage of time, money and resources. The worst outcome would be a complete lack of valid information on which to build understanding, predictive capacity and managerial/conservatory decision-making. Where aims are vague, designs of sampling are usually (if not always) inadequate, data do not match necessary assumptions, analyses are invalid and conclusions suspect.

In contrast, where aims and purposes are logical, coherent and explicit, it is usually possible to design a robust, effective, efficient and satisfactory sampling programme, which will allow aims to be achieved with minimal uncertainty. This seems such common sense that it does not need to be stated – but common sense indicates that the world is flat and that the sun rotates round the earth. Common sense is not enough.

For example, Hurlbert (1984) published a devastating critique of the failure of many published studies to demonstrate a valid basis for reaching conclusions, because the samples analysed were inappropriately or not at all replicated. His study was confined to the published studies, in refereed journals, subject to independent scrutiny. The overall situation, taking into account rejected papers, less intensively scrutinised journals and the flood of unreviewed grey literature, was clearly much worse. It is clear, from practical inspection of more recent literature and through reviewing manuscripts and applications for grants, that the situation (though now better) has not improved substantially (Hurlbert, 2004).

The starting point for studies needing quantitative sampling is that objectives should be clear, the variables to be measured should be defined and the sorts of patterns anticipated in the data should be clearly identified (as testable hypotheses). Wherever possible, as much information as is available should have been collected (and understood) about the operational processes operating, their spatial and temporal scales and about the biological interactions in assemblages and responses to environmental variables. Ideally, the constraints of money, time and equipment should also have all been taken into account. In other words, the professional components of scientific work should all be in place. Under these circumstances, it should be possible to design sampling to achieve minimal probabilities of error in analyses.

As a result, the focus in this chapter is on general issues and procedures to provide help and guidance with setting objectives, formulating hypotheses and designing sampling, particularly for measuring ecological impacts. This will serve as an aide-memoire for contemplating issues of logic when dealing with spatial and temporal variability in biological systems. It will also provide an introduction into the broader literature where many advances have been made in methodologies dealing with the problems of biological complexity in the real world.

1.2 Variability in benthic populations

Surveys must always be designed to take into account the fact that benthic animals and plants are extremely patchy in distribution and abundance. Patchiness is caused by processes external to the assemblage, particularly disturbances and recruitment, in addition to processes operating within the existing assemblage. Although anthropogenic disturbances are often very severe (e.g. large-scale trawling or dredging can cause extreme changes in benthic assemblages; Hall & Harding, 1997; Lindegarth et al., 2000), natural disturbances are common and can be important contributors to spatial variability of populations. These vary from small-scale disturbances, e.g. being overturned by waves affects the assemblage on a boulder (Sousa, 1979) and potentially the assemblage in the sediment below it, to large-scale processes, such as the erosion of nearshore sediments (Shanks & Wright, 1986) and the destruction of assemblages in response to large storms (Underwood, 1998).

However, the most important contribution to patchiness is probably due to unpredictable and variable patterns of recruitment (Underwood & Denley, 1984). Both settlement itself and post-settlement mortality typically vary at a hierarchy of spatial scales (Caffey, 1985; Gaines & Bertness, 1992). Patterns at larger scales tend to be more predictable because most species are confined to particular habitats within a biogeographic range (Brown & Gibson, 1983). However, at small scales within patches of habitat, there is considerable variability in recruitment (Keough, 1998), caused by local environmental variation (e.g. topographic features of habitat or localised water currents) or by the existing assemblage itself (e.g. gregarious settlement in response to conspecific adults or consumption of larvae by large numbers of sessile species). The numbers of larvae competent to settle that arrive in any site are themselves influenced by a multitude of external processes, many of which act in the water column well away from the site of settlement. The localised processes, both physical and biological, that influence recruitment are interactive, so recruitment is extremely variable in space and time.

In addition, numerous interactions within the assemblages themselves continually alter patterns of abundances and these effects, too, occur at a range of spatial scales. For example, though predation may decrease abundances at the scale of a shore or habitat, within that habitat predation may eliminate species from certain patches, but leave other patches alone. Feeding by eagle rays or shore birds can create extreme small-scale patchiness in abundances of their prey, although these effects are complicated by environmental factors, such as currents and movement of sediment (reviewed by Thrush, 1999). Even in areas with heavy predation, prey may settle in particular microhabitats, where they can grow large enough to escape predation (Dayton, 1971).

Competition, either for space among sessile animals or plants or for food among mobile animals, also contributes greatly to patchiness of assemblages. Therefore, over-growth or dislodgment of one species by another (Keough, 1984) causes very patchy assemblages of sponges, ascidians and other colonial animals in subtidal habitats and of barnacles and various types of algae in intertidal habitats. Although species that are superior competitors for space or food may eliminate inferior species, the relative strength of interspecific competition may be balanced with that of intraspecific competition. This ensures that neither species is eliminated, but that both persist in very variable and patchy numbers.

Processes such as these are better understood for benthic assemblages living on hard surfaces because, first, the patterns are often readily visible and, second, the processes are relatively easily investigated experimentally. They are, however, also important for assemblages in soft sediments, where recruitment may be equally variable (Skilleter, 1992; Whitlatch et al., 1998) and local disturbances, competition and predation alter local abundances, causing very patchy distributions at a hierarchy of spatial and temporal scales (Morrisey et al., 1992a, 1992b; Ysebaert & Herman, 2002; Fig. 1.1). To estimate abundances of benthic animals and plants accurately, measures must be made at the range of spatial scales relevant for the species, assemblage or process under consideration.

FIGURE 1.1 Mean (Standard Error; SE) abundance of two species of amphipods (a) and (b) between sites (S1, S2; tens of metres apart) in each of two locations (L1, L2; 100 m apart) in each of three mangrove forests (F1, F2, F3; kilometres apart). Note that each species shows significant variation at each spatial scale, but these differ between the two species. Patterns of variation at the scale of sites and locations also vary from one mangrove forest to another.

vsp -6pt?

1.3 Appropriate replication

Appropriate spatial replication

Whatever the hypothesis being tested and the ultimate use of the data from sampling, spatial replication is a mandatory component of any benthic study. The large variability in numbers...