Community Level Patterns

Body Size-Abundance Distributions

Body size-abundance distributions describe the variation of some measurements of individual abundance with individual body mass. The measurements of abundance used are number and biomass of individuals of each population within a guild or a community, number or biomass of individuals in successful populations within a species range, at the regional, continental, and global scale, number or biomass of individuals within a community and number or biomass of individuals in base 2 logarithmic body-size classes. Whatever criteria for grouping individuals are selected, within populations, communities or size classes, at the guild, community, landscape, continental or global scale, as number or biomass, a negative relationship between individual density and body size is generally observed. However, the shape and coefficient of these relationships, the mechanisms involved and the ecological significance vary according to the criteria selected, and each single body-size pattern provides different information, contributing to a better understanding of the role of individual body size in structuring and organizing ecological communities.

The selection of either species populations or body-size classes as a grouping criterion creates two main categories of size-abundance distributions: 'taxonomi-cally based and nontaxonomically based'. The latter are commonly referred to as 'size spectra'. Studies of terrestrial ecosystems have preferentially used body size-abundance distributions based on the taxonomic grouping ofindividuals into populations and communities, whereas studies of aquatic ecosystems have preferentially used body size-abundance distributions as 'taxon-free' patterns, grouping individuals into logarithmic body-size classes independently oftheir taxonomy.

Taxonomically based size-abundance distributions

On average, population densities (PDs) scale with individual body size (BS) according to the allometric equation

PD = aiBSbl where b1 is typically lower than 0 and a1 is the specific density. a1 expresses the combined action of factors such as average energy transfer efficiency, average energy availability, and temperature-driven shifts in the metabolic rates ofthe populations in question.

Broadly speaking, taxonomically based size-abundance distribution derives from the notion that since the energy requirement ofindividuals (Met) increases with individual body size according to a well-known allometric equation

Met = a2BSi2

where b2 has been consistently found to be close to 0.75, the number of individuals of each population supported by the available resources must decrease with average individual body size. Assuming that resource availability is homogeneous across species and body sizes, the slope of the body size-abundance distribution (b1) is expected to be -0.75.

The processes underlying body size-abundance distributions, and hence their information content and ecological meaning, depend on whether they account for density values and average body sizes of species on a regional, continental, or global scale (hereafter, 'global-scale size-abundance distributions'), for density and average individual body size of co-occurring populations within guilds or communities (hereafter, 'local-scale size-abundance distributions'), or for average population densities and individual body sizes of entire guilds or communities along ecological, climatic, or biogeographic gradients (hereafter, 'cross-community size-abundance distributions').

Global-scale size-abundance distributions are among the most extensively studied. They cover regional, continental, and global scales, and the broadest range of taxonomic variation, with a bias toward birds and mammals, for which more extensive databases of population densities and body sizes are available at every spatial scale. Data used to compile global-scale size-abundance distributions typically describe densities of successful populations within the species' geographic range, which may be close to the maximum carrying capacity. Most commonly, populations included in the global-scale size-abundance distributions do not coexist, and affect each other through vertical or horizontal interactions. For large compilations of population densities, population density generally scales very closely with body size, with a slope near the value of-0.75. The close agreement between the slope observed for global-scale size-abundance distributions and that expected on the grounds of simple energetic arguments confirms that at the continental and global scales, availability of resources or energy is not correlated with species body size. The homogeneity of resource or energy availability across species body sizes is an interesting, but far from straightforward, aspect of global size-abundance distributions. It implies that the advantage for large species arising from their wider niches (and thus greater availability of resources) with respect to small species, is counter-balanced by the presence of other body-size-related factors which compensate.

These include the resource density perceived by individuals and the individuals' exploitation efficiency, both of which decrease with increasing body size. Intercepts of global size-abundance distributions express the average energy-use efficiency of the group of populations considered. Compilations of global size-abundance distributions for ectothermic and endothermic species show different intercept (a1) values, the former having less negative intercepts than the latter due to the cost ofbeing homoeo-thermic. Similarly compilations of size-abundance distributions of herbivores have higher a1 values than those obtained for carnivores, reflecting the overall efficiency of energy transfer in food webs.

When size-abundance distributions are compiled at the local level, where the body size and abundance of each species (N) is measured at the same location, body size generally explains only a small part of the variation in population abundance, and the regression slope is much higher than the expected -0.75. The observed deviations from the expected slope in local size-abundance distributions are suggestive of size biases in resource acquisition that could be driven by size asymmetry in competition. An alternative hypothesis to explain the deviation of local size-abundance distributions from global ones is that the former typically examines a smaller range of sizes than the latter. Observing a smaller portion of the overall relationship accentuates the noise in the local sample. This could explain why local size-abundance distributions in aquatic environments, covering a larger spectrum of body sizes than terrestrial ones, are also generally stronger. In fact, at the local scale, triangular-shaped size-abundance distributions are much more commonly observed than simple allometric relationships. Triangular distributions have three major attributes: an 'upper bound', a 'lower bound', and a dispersion of points in the size-abundance space (Figure 1a). The 'upper bound' of the triangular-shaped size-abundance distributions is determined by the body-size scaling of the dominant species' population densities. The 'upper bound' has been used as a proxy of the complete local size-abundance distribution, under the assumption that the ecological role of rare and occasional species, being weak, is unclear. The procedure may be useful for applied purposes but since most species are rare the assumption is not generally acceptable. The body-size dependency of the minimum viable population may explain an expected 'lower bound', which is difficult to measure because of the problems with correctly quantifying the rarity of populations. The density of points between these two bounds is determined mainly by regional processes and horizontal and vertical partitioning rules. The ecological information carried by the intercepts of the size-abundance distributions is of lower value at local than at global scale because whenever slopes are different, as often













c c












ci e a.







ie ci






















iz si




o B



100 150 Species rank

Figure 1 Body-size patterns of macroinvertebrate guilds of transitional water ecosystems in the Mediterranean and Black Sea Eco-regions. Both local size-abundance distributions (a) and body size-species distributions (b) are triangular shaped. The 'upper bounds' of the triangular distributions are reported. The graph (c) emphasizes that species of transitional water macroinvertebrates are clumped around the mode of the body size-species distribution, with 74% of the species being grouped in 2 out of the 5 order of magnitudes occurring between the size of the smallest and the largest species.

100 150 Species rank

Figure 1 Body-size patterns of macroinvertebrate guilds of transitional water ecosystems in the Mediterranean and Black Sea Eco-regions. Both local size-abundance distributions (a) and body size-species distributions (b) are triangular shaped. The 'upper bounds' of the triangular distributions are reported. The graph (c) emphasizes that species of transitional water macroinvertebrates are clumped around the mode of the body size-species distribution, with 74% of the species being grouped in 2 out of the 5 order of magnitudes occurring between the size of the smallest and the largest species.

occurs when comparing local size-abundance distributions, comparisons between intercepts are not possible.

Classifying all the individuals in a population into guilds or communities and averaging their mass, we may then describe every guild and community with two simple parameters: mean organism size (BS) and total community abundance (Ntot). The scaling of total community abundance with mean organism size leads to cross-community size-abundance distributions. Cross-community size-abundance distributions were first studied in self-thinning plant and sessile communities, where, as organisms grow, there is space for fewer and fewer individuals, determining a negative relationship between BS and Ntot. In general cross-community size-abundance distributions tend to be well described by allometric equations, whose slopes tend to be similar to the inverse of the scaling exponent of metabolic rates with individual body size. A similarity between observed and expected slopes has been also detected in guilds and communities which are not regulated by self-thinning rules, such as bird and phytoplankton guilds. However, since much fewer data are available for cross-community size-abundance distributions than for global and local size-abundance distributions, the underlying mechanisms remain to be determined.

Nontaxonomically based size-abundance distributions

In large aquatic ecosystems, early studies of body size-abundance distributions focused on energy transfer (i.e., how information on productivity and energy transfer may be gained from body-size data, which can be collected relatively easily). In accordance with this objective, they dealt with particles rather than with species, dividing particles suspended in the water column into logarithmic-base 2 size classes, irrespective of species and including nonliving organic particles. Thus the nj particles in the jth body-size class of average mass BS; may represent more than one species, and every species can occur in more than one class. Nontaxonomic size-abundance distributions (hereafter referred to as size spectra) have been quantified for many different guilds and communities, including plankton, benthos and fish guilds, woodland and forest plant guilds, as well as marine, freshwater, and terrestrial ecosystems; however, a large proportion of the ecological literature addressing size spectra deal with the pelagic marine environment.

According to the classification reported for taxonomi-cally based body size-abundance distributions, almost all size spectra are local, being determined at the guild or community scale. Size spectra can be compiled with two different types of data, that is, biomass and number of individuals. Both biomass-size spectra and number-size spectra can cover different body-size ranges, describing either entire communities or single guilds.

Regarding biomass-size spectra, the amount of biomass has been shown both empirically and theoretically to be constant when plankton individuals are organized into logarithmic size classes. As a result of this equal partitioning of biomass, the slope of a straight line fitted to plankton biomass-size spectra is expected to be 0; this relationship is known as the 'linear biomass hypothesis', which has strong experimental support in aquatic pelagic environments, particularly when a large spectrum of sizes and trophic levels are considered. Often the data is subjected to a normalization procedure, which consists of dividing the biomass in each size class by the width of

the size class. In normalized biomass-size spectra, biomass in each size class decreases isometrically with the average class size, the slope being close to —1. The linear biomass hypothesis implies that in pelagic systems, the number of individuals within logarithmically increasing size classes declines linearly with average body size. The slope of the allometric equation tends to be close to -1; when number-size spectra are normalized, the expected slope is equal to -2. Nevertheless, within pelagic-size spectra, a series of dome-like distributions are typically detected, corresponding mainly to different functional guilds within which there is a poor fit with linear statistical regressions.

'Dome-like' distributions and gaps in number- and biomass-size spectra occur not only between but also within functional groups, such as phytoplankton and zooplankton, even when they are not attributable to incomplete censuses of species or to systematic underestimation of intraspecific size variation. Dome-like patterns of biomass distribution have been observed both in freshwater and marine ecosystems, as well as in macro-zoobenthos and fish. Therefore, by restricting the range of body size considered and addressing specific functional groups, size spectra tend to have a shape similar to the triangular shape of local size-abundance distribution. Most commonly, the maximum number and biomass of individuals, either partitioned into species or irrespective of species, occur at some small but intermediate body size, rather than at the smallest size.

Two kinds of scaling in the relationship between body size and abundance within size spectra may be recognized. A unique and primary slope reflecting the size dependency of metabolism ('metabolic scaling'), and a collection ofsecondary slopes which represent the scaling of numerical or biomass abundance with body size within groups of organisms having similar production efficiencies ('ecological scaling'). Size-dependent coexistence relationships are likely to be representative of the secondary slopes, leading to a dominance of large cells/species, and slopes that are less negative than predicted by the 'linear biomass hypothesis'. Ecological scaling can also produce dome-like patterns in size spectra within the size range of each functional group.

Body Size-Energy Use Distributions

The body-size dependence of both metabolic rates and population densities makes it possible to evaluate populations' rates of energy use and how they scale with individual body size. Indeed, the rate at which energy flows through a population (E) can be evaluated as the product of individual metabolism (Met) and population density (PD), as follows:

E — Met x PD = a2BSh x a1BSbl — (a2 x a1)BS{b'2+h)

Since b2 has been found to be consistently close to 0.75, the scaling of energy-use rates with individual body size depends on b1, which is generally expected to be negative, since, at every spatial scale of ecological organization, many small and few large individuals occur.

Assuming that resource availability is homogeneous across species and that species do not limit each other's resource availability and have optimized the efficiencies of resource exploitation and use, then population densities are expected to scale with individual body size with a slope (b1) of-0.75, and the amount of energy each species uses per unit of area is expected to be independent of body size:

The independence of energy use per unit area from body size is known as the energetic equivalence rule (EER). Whenever b1 is consistently lower, more negative, than -0.75, small species dominate energy use. Conversely, if b1 is consistently larger, less negative, than -0.75, large species make a disproportionately large use of the available energy per unit of area.

Global size-abundance distributions seem to agree with the EEF. At the global scale, the energy use of the most successful populations within the species range seems to be actually independent of the body size of individuals within populations. On the other hand, local size-abundance distributions, which commonly show scaling exponents higher, less negative, than -0.75 show that within local guilds and communities large species normally dominate energy use. Dominance of small species has also been detected at the local scale usually in relation to some degree of stress. Therefore, the shape and slope of local size-abundance distribution, and consequently the body-size scaling of energy use, can have practical applications in ecology.

Body-Size-Species Distributions

Understanding biodiversity is a major goal of ecology. Since many small and few large species occurs in the biosphere, at every scale, from the community to the continental and global level, describing and understanding the scaling of biodiversity patterns with individual body size is also a key topic.

Basically, whenever organisms perceive a two-dimensional (2D) habitat, they sample habitats on a grid proportional to the reciprocal of the square of their linear dimension (L). Therefore, the likelihood of niches being opened up to species specializing in particular resources and habitat patches is proportional to «L~2 or to BS-067. Consequently, the number of species (S) is expected to decrease with individual body size according to ~L" . Whenever organisms perceive a 3D habitat, the species number is expected to be proportional to «L-3 or «BS-1. Considering that the linear dimension of individuals represents the 'ruler' (L) they use to sample the habitat and that habitats are rarely completely homogeneous at every scale ofperception, individuals perceive the habitat to be fractal. The perceived 2D habitat scale is , where D is the fractal dimension of the habitat as well as of the resources. The fractal dimension is a habitat property that in many field studies has been found to be close to 1.5. This would mean that in 2D habitats the number of species (S) is expected to be between L-D and L~2D, that is, between L~1' and L-3 0, where L-2 is analogous to the 'upper bound' of size-abundance distributions. In 3D habitats the number of species (S) is expected to be proportional to L-3D, that is, to L-~5. Assuming D« 1.5, a tenfold decrease in individual size determines a threefold increase in the perceived length ofeach habitat edge, a tenfold increase in apparent habitat surface and a maximum tenfold increase in S.

Available data on both the full range of taxa and particular groups of species consistently show that the species-size distributions are humped, with the mode in some small but intermediate size class (Figure 1b). The underestimation of the number of existing small species may be an explanation of humped distributions covering the whole scale of size from the smallest to the largest species. Underestimation of small species is less likely to explain humped distributions observed within restricted taxonomic groups, such as invertebrates, birds, and mammals. Within restricted taxonomic groups, it seems likely that an optimal body size exists, where species and individuals perform optimally and tend to be clumped (Figure 1c). An optimal body size of between 100 g and 1 kg has been proposed for mammals and an optimal body size of 33 g has been proposed for birds. Two hypotheses have been proposed to explain the size dependency of species performance at every scale: the energy conversion hypothesis, addressing optimal size according to the size dependency of the efficiency of energy conversion into offspring; and the energy control hypothesis, addressing optimal body size according to the species' performance in monopolising resources.

Body-Size Ratios

Coexisting species of potential competitors commonly differ in body size. Using consumer body size as a proxy of resource size, this difference may explain competitive coexistence; in his famous paper 'Homage to Santa Rosalia: or why are there so many kind of animals' Hutchinson proposed that in order to coexist species must be spaced in size with a ratio between their linear dimensions of at least 1.28 (2.0-2.26 in biomass), which is commonly referred to as the 'Hutchinson ratio'. Patterns of body-size spacing between coexisting species pairs consistent with the 'Hutchinson ratio' have been observed for many groups of animals, including, birds, desert rodents, and lizards. The 'Hutchinson ratio' corresponds to a limiting similarity threshold; therefore, the average size ratio between species is expected to vary with resource limitation. Actually, the average size ratio between species pairs decreases with increasing richness and with decreasing guild trophic level.

That co-occurring species within guilds tend to have different body size, with an average size ratio close to the expected 2-2.26, is a very general observation in ecology. However, the ecological relevance of the 'Hutchinson ratio' has been questioned, mainly due to two key critical observations, apparently in contrast with the interpretation that size ratios between species correspond to a low-enough niche overlap between species pairs to allow interspecific coexistence: (1) many nonliving things in nature as well as many objects built by humans, from nails to musical instruments, are scaled in size according to the 'Hutchinson ratio'; and (2) size spacing between species pairs does not always seem to be related to niche spacing. The latter is the most critical issue. However, a functional link between the body-size ratios of coexisting species and competitive coexistence conditions may also be derived independently of any niche spacing. Body-size-mediated coexistence between species differing in size may result from simple energetic constraints on individual space use regardless of any a priori resource partitioning: that is, size ratios between species may critically affect species coexistence even if niche spacing is not detected.

Was this article helpful?

0 0
Project Earth Conservation

Project Earth Conservation

Get All The Support And Guidance You Need To Be A Success At Helping Save The Earth. This Book Is One Of The Most Valuable Resources In The World When It Comes To How To Recycle to Create a Better Future for Our Children.

Get My Free Ebook

Post a comment