AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS
Aquatic Conserv: Mar. Freshw. Ecosyst. 14: S43–S58 (2004) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/aqc.649
Monitoring with benthic macroinvertebrates: advantages and disadvantages of body size descriptors ALBERTO BASSET*, FRANCA SANGIORGIO and MAURIZIO PINNA Department of Biological and Environmental Sciences and Technologies, University of Lecce, 73100 Lecce, Italy
ABSTRACT 1. The search for simple and effective descriptors of biological ecosystem components is a major challenge of monitoring aquatic ecosystem health. 2. The relevance of body-size-related descriptors of benthic invertebrate guilds in monitoring the health of transitional aquatic ecosystems is discussed. The rationale is that macroinvertebrate body size relates body-size–abundance distributions to disturbance pressures through individual energetics, population dynamics, interspecific interactions and species coexistence responses. 3. Body size is generally easy to measure and amenable to intercalibration procedures, it is comparable across taxa, guilds and sites, and, as a community feature, it is expected to vary on disturbance gradients, according to energetic and ecological constraints. 4. The mechanistic relevance of individual body size as a community feature, through coexistence relationships, still requires field and laboratory tests; standard methods to analyse body-size– abundance distributions are not yet fully developed. 5. Field experiments on coastal lagoons and freshwater ecosystems of southern Italy, which were designed to test the relevance of body-size-related constraints on the organization of detritus-based benthic guilds, are reviewed. 6. Study cases emphasized a number of interesting features of body size and related descriptors, that support their relevance as benthic invertebrates descriptors of ecosystem health: (a) body-size– abundance distributions are consistently less variable than taxonomic composition; (b) the width of body-size–abundance distribution is mainly due to the interspecific component; (c) the descriptors of body-size–abundance distributions seem to respond on environmental gradients and generally covary with species density, richness and diversity, on which most of the monitoring programmes actually rely. Copyright # 2004 John Wiley & Sons, Ltd. KEY WORDS:
coastal lagoons; benthic macroinvertebrates; body-size–abundance distributions; ecosystem health
*Correspondence to: A. Basset, Department of Biological and Environmental Sciences and Technologies, University of Lecce, 73100 Lecce, Italy. E-mail:
[email protected]
Copyright # 2004 John Wiley & Sons, Ltd.
Accepted 12 February 2004
S44
A. BASSET ET AL.
INTRODUCTION Benthic macroinvertebrates play an important role in many ecological processes in transitional waters, rivers and lakes (Plante and Downing, 1989; McCall and Soster, 1990; Griffiths, 1991); they integrate environmental changes in physical, chemical and ecological characteristics of their habitat over time and space (Cook, 1976; Milbrink, 1983). That is why they have been attractive targets of biological monitoring of environmental quality in aquatic ecosystems (Hellawell, 1986; Metcalfe, 1989; Rosenberg and Resh, 1993). The use of benthic organisms to assess water quality was first proposed around 40 yr ago (Hynes, 1959; Hawkes, 1979) and its validity has been repeatedly confirmed (Rutt et al., 1989). There are many advantages of using macroinvertebrates in bio-monitoring: (a) they are ubiquitous (Lenat et al., 1980); (b) they have a basically sedentary behaviour, which allows a spatial analysis of pollutant (Slack et al., 1973; Hellawell, 1986; Abel, 1989); (c) they have long life cycles when compared with other groups, which allows an analysis of temporal changes caused by perturbations (Gaufin, 1973; Lenat et al., 1980); (d) they can be affected by environmental perturbations in many different aquatic ecosystems, which allows an analysis of a spectrum of responses to environmental stress (Rosenberg and Resh, 1993). In spite of these advantages, many researchers have found some difficulties in using benthic invertebrates in bio-monitoring. These are related mainly to the requirements of a great expertise in taxonomy (in order to minimize any bias in the assessment of species composition and diversity) and of a high number of samples (in order to account for the distribution of these populations which is commonly contagious). Bio-monitoring with benthic invertebrates mainly followed the so-called ‘indicator species’ approach, either qualitative, based on the presence–absence of taxa sensitive to perturbations, or quantitative, based on the numerical or taxonomic abundance and including also some weighted combinations of numerical and taxonomic abundance through biotic indices (e.g. Rosenberg and Resh, 1993). Several biotic indices have been proposed so far (e.g. Rosenberg and Resh, 1993); among these, the Trent biotic index (Woodiwiss, 1964) and its slightly modified version, the extended biotic index (Woodiwiss, 1978), are very popular in several countries (Ravera, 2001). They are based on different sensitivities of some taxa of macroinvertebrates (Norris and Georges, 1993) and are used extensively. On the other hand, they have a generally low sensitivity to weak disturbances, being unsuitable for detecting early signs of stress (e.g. Balloch et al., 1976; Murphy, 1978; Hellawell, 1986). A second major bio-monitoring approach with benthic macroinvertebrates is represented by the river invertebrate prediction and classification system (RIVPACS) (Logan and Furse, 2002). RIVPACS, which fundamentally relies on the Grinnell–Hutchinson niche concept (sensu Leibold, 1995), was originally expressed as the macroinvertebrate community structure of a large number of pristine, reference stream ecosystems in the UK (i.e. 614 reference sites; Wright et al., 1997) as a function of the niche in the environment at those sites, measured from a small number of physical dimensions: distance from source, slope, mean substrate particle size, altitude, discharge category, mean width, mean depth, latitude, longitude, mean air temperature, air temperature range. RIVPACS models, which are now developed for other regional areas (Wright et al., 2000), being based on community taxonomy, are sometimes restricted to the habitat type or the regional area from which pristine sites were studied and models developed, which is a major drawback for a classification system requiring extensive field studies for its setup. In the last three decades, theoretical and experimental studies have focused on body-size–abundance distributions, biomass-size spectra or dimensional structures as structural community features (Damuth, 1981; McMahon and Bonner, 1983; Peters, 1983a; Lawton, 1990; Schmid et al., 2000). Body-size– abundance distributions can provide tools for the evaluation of aquatic ecosystem health (Rasmussen, 1993), which can offer an alternative or complementary perspective to taxonomic analysis. The size-based approach was initially utilized in a marine environment (Sheldon et al., 1972) and lake plankton communities (Peters, 1983b; Sprules and Munawar, 1986; Ahrens and Peters, 1991); more recently, on the Copyright # 2004 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 14: S43–S58 (2004)
MONITORING MACROINVERTEBRATES AND BODY SIZE DESCRIPTOR
S45
basis of these studies, benthic ecologists have used size structure of invertebrate assemblages as an alternative to taxonomic description of these communities (Strayer, 1991), both in freshwater lake (Hanson, 1990; Rodriguez and Magnan, 1993; Morin et al., 1995) and stream ecosystems (Morin and Nadon, 1991; Bourassa and Morin, 1995; Morin, 1997; Solimini et al., 2001). Theoretical advances in community ecology supported the relevance of body size on coexistence relationships and community organization, i.e. the body size–energy constraints hypothesis (Basset, 1995a), also called body size–optimal foraging hypothesis, (Tokeshi, 1999). The body size-energy constraints hypothesis of coexistence allows a number of predictions on body-size–abundance distribution descriptors that could be very relevant for conservation issues and which will be considered in this paper: *
*
Body-size–abundance distributions are expected to be at a higher hierarchical level than taxonomic composition of communities. Therefore, they would be independent of the taxonomic composition and, consequently, less variable. Guilds are expected to be organized hierarchically on a body size gradient. Therefore, the width of the size–abundance distributions is expected to decrease with increasing direct, or cascade, disturbance pressure.
On the other hand, the predictions of the body size-energy constraints hypothesis on body-size–abundance distribution descriptors are based on the assumption that body-size–abundance parameters are determined by interspecific more than intraspecific components. This assumption, however, has to be proven for benthic macroinvertebrate taxa, whose individual body size varies greatly during their life cycle. The current study set out to review field experiments carried out to test these predictions in coastal lagoons and freshwater ecosystems of southern Italy. Body-size–abundance distributions in coastal lagoons and brackish lakes, which have received little attention in the literature, were focused on here and compared with freshwater ecosystem data. Preliminary investigations were also carried out into the relative contributions of intra- and inter-specific individual body size variability to body-size–abundance distributions in brackish water and freshwater ecosystems of the Mediterranean region.
ANALYSIS OF MACROINVERTEBRATE BODY-SIZE–ABUNDANCE DISTRIBUTIONS The body-size–abundance distributions of benthic macroinvertebrates colonizing leaf detritus in transitional and freshwater ecosystems of southern Italy have been studied. Body-size–abundance distributions were drawn on data of individual mass (micrograms of ash-free dry weight) after gravimetric determinations. Benthic macroinvertebrates were sampled using the leaf packs technique (Petersen and Cummins, 1974) from the River Tirso and River Mannu-Cixerri basins (Sardinia, Italy) and at the Lake Alimini complex (Apulia, Italy). The macroinvertebrates associated with each leaf pack were gently removed, sorted according to taxon and functional role (Cummins, 1974) and enumerated. Each individual was then ovendried at 608C for 72 h and weighed on a Sartorius MC5 microbalance at the nearest 1 mg. Ash content was determined on groups of co-specific individuals after muffle furnace combustion at 4508C for 24 h. Data on individual masses were logarithmically transformed and grouped into unitary size classes. Position and width of log-normal distributions and descriptive statistics computed on raw data were used to describe body size structure of macroinvertebrate guilds. Detailed biotic and abiotic features of the study sites are published elsewhere for the Lake Alimini complex (Sangiorgio and Basset, 2001; Vadrucci et al., 2001; http//:www.ecomuseoalimini.unile.it), the River Mannu-Cixerri (Basset, 1995b; Ponti et al., 1996), and the River Tirso (Pinna et al., 2000, 2003). Data on taxonomic and body size structure of macroinvertebrate guilds were compared using proportional Copyright # 2004 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 14: S43–S58 (2004)
S46
A. BASSET ET AL.
similarity indices (Renkonen, 1938; Feinsinger et al., 1981) after normalization of both body size structure and taxonomic composition into the same number of classes.
BODY-SIZE–ABUNDANCE DISTRIBUTION VARIABILITY Body-size–abundance distributions can be described at both population (Damuth, 1981; Peters, 1983a; Lawton, 1990; Schmid et al., 2000) and individual levels (Basset, 1995b; Bourassa and Morin, 1995; Solimini et al., 2001). At the population level, the number of populations, the population density and the average body size of individuals are related through commonly decreasing functions. Many species are rare and only a few are common; many are small and few are large; moreover, small species are expected to be dense and large species sparse. These very simple and widely accepted relationships have an intrinsic contradiction, which is apparent when body-size–abundance distributions are studied at guild or community level where most species are rare, most species are small and only a few small species are actually dense. Studies of size– abundance distributions at regional, continental or global scale, which gave the strongest support to the expected inverse relationship between body size and population density (Damuth, 1981; Peters, 1983a), generally take into consideration only relatively dense populations, with a bias towards large species, and neglect the fact that most small species are sparse. Sparse, small species could represent confounding variables in body-size–abundance distributions. In the Lake Alimini complex, benthic macroinvertebrate guilds showed decreasing species–abundance distributions; 63.4% of taxa had a relative abundance lower than 0.05% of the entire guild and can be considered rare or occasional (Figure 1). When taxa for each class of abundance were divided into groups of small and large taxa (as characterized by an individual body size smaller or larger than the average for that class of abundance), 78% of rare and occasional taxa can also be considered small (Figure 1). Sparse small taxa of benthic macroinvertebrates had a profound influence on body-size–abundance distributions in Lake Alimini. Globally, population abundance and body size of individuals in the macroinvertebrate guilds studied were not statistically related (Figure 2(A)); (slope of the best fit b=0.26),
Figure 1. Species–abundance distribution in the Lake Alimini complex. Species are divided into rare, common and very common. Dashed boxes represent the abundance of large species in the rare group. Dashed line approximates the expected log-normal distribution when small rare species are removed. Copyright # 2004 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 14: S43–S58 (2004)
MONITORING MACROINVERTEBRATES AND BODY SIZE DESCRIPTOR
S47
No individuals
A -0.257
100000
y = 23.32x r = 0.171; d.f.= 69; n.s.
10000 1000 100 10 1
0.01
1
100
10000
Body size (mg, AFDW)
Size-abundance residuals
B 100
-0.276
y = 19.83x r = 0.944; d.f.=69; P
