Meiobenthos and Macrobenthos Are Discrete Entities and Not Artefacts of Sampling a Size Continuum: Comment on Bett (2013)
Total Page:16
File Type:pdf, Size:1020Kb
Vol. 505: 295–298, 2014 MARINE ECOLOGY PROGRESS SERIES Published May 28 doi: 10.3354/meps10830 Mar Ecol Prog Ser FREEREE ACCESSCCESS COMMENT Meiobenthos and macrobenthos are discrete entities and not artefacts of sampling a size continuum: Comment on Bett (2013) R. M. Warwick1,2,* 1Plymouth Marine Laboratory, Prospect Place, West Hoe, Plymouth PL1 3DH, UK 2Centre for Fish and Fisheries Research, School of Veterinary and Life Sciences, Murdoch University, 90 South Street, Murdoch, Western Australia 6150, Australia ABSTRACT: Species size distributions for metazoan benthic invertebrates conform to the highly conservative bimodal pattern, regardless of the sieve mesh sizes or numbers of sieves used in their extraction. This pattern is not an artefact of sampling a size continuum as suggested by computer simulations using just 2 fixed mesh sizes in Bett (2013; Mar Ecol Prog Ser 487:1−6). Meiobenthos and macrobenthos are coherent entities, each with a distinct suite of functional attributes, and should not be regarded as a single unit for ecological modelling purposes. KEY WORDS: Marine benthic invertebrates · Species size distributions · Meiobenthos · Macrobenthos · Functional traits Resale or republication not permitted without written consent of the publisher Introduction another potential artefact. He stated that this calls into question accepted ideas that meiobenthos and Using computer simulations, Bett (2013) demon- macrobenthos have coherent identities with distinct strated what is perhaps intuitively obvious: that ecological attributes and went on to conclude that a when size continua of perfect spheres (or a mixture of single mathematical formulation of standing stock and spheres and identically shaped cylinders in equal body size might be appropriate for bio geochemical proportions) are sampled using sieves with 2 mesh modelling purposes. sizes corresponding to those traditionally used by meiobenthic and macrobenthic researchers (i.e. 45 or 63 µm and 500 µm respectively), biomass size spectra Real size spectra are produced comprising 2 maxima with a trough between them. He therefore suggested that the pre- Nearly 30 years ago, before the ready availability viously proposed bimodal biomass spectrum across of computer simulation, Warwick (1984) addressed the meiobenthos-macrobenthos size range may be a the problem of potential sampling artefacts in the sampling artefact. He went on to extrapolate this above context. In studies of species richness size finding to account for species richness size spectra on spectra from 8 locations (examples in Fig. 1), more the basis that the number of species is likely to than 2 mesh sizes were used for the extraction of the increase with the number of specimens examined, fauna in all but one case (Northumberland offshore suggesting that bimodal species size distributions are mud; details in Warwick 1984). At 5 locations, mesh *Corresponding author: [email protected] © Inter-Research 2014 · www.int-res.com 296 Mar Ecol Prog Ser 505: 295–298, 2014 24 Northumberland mud Carmarthen Bay sand Algoa Bay sand 20 16 12 8 4 Number of species 0 2 6 10 14 18 22 26 30 2 6 10 14 18 22 26 30 2 6 10 14 18 22 26 30 X2 geometric weight classes Fig. 1. Examples of species richness size spectrum histograms for sublittoral samples of metazoan benthos, showing conservative pattern of 2 normal curves separated by a trough (redrawn from Warwick 1984) sizes of 63, 125 and 500 µm were used, the interme- tinct ecological attributes, since this contradicts what diate size employed specifically to enumerate species is known about size-related biological traits of ben- of intermediate size between meiobenthos and macro - thic metazoans. Despite differences in sampling pro- benthos. Interestingly, the size spectrum from sub- tocols, sediment types and the sets of sieve sizes tidal sand in Algoa Bay, South Africa (Fig. 1) was employed, the species size spectra shown in Fig. 1 based on mesh sizes of 63, 180, 1000 and 1300 µm are remarkably conservative, with 2 separate lognor- with additional hand collection by divers, which is mal distributions having modes at dry body weights quite similar to the 4-sieve scheme simulated in of 0.64 µg and 3.2 mg and a trough between them at Bett’s (2013) Fig. 4 (63, 180, 500 and 1400 µm). Bett’s 45 µg. There are no statistically significant differ- simulated data matched the size continuum with no ences at the 5% level in the values of these parame- trough between meiobenthos and macrobenthos, ters among locations (Clarke 1984). The form of these while the real data (Warwick 1984) conformed to lognormal curves is different from Bett’s (2013) flat- the highly conservative bimodal pattern found at all topped spectra for each of the 2 sieve sizes, with the locations. Furthermore, Warwick (1984) also consid- flat tops reflecting the underlying simulated size dis- ered the possible relationship between numbers of tribution. Because the pattern is so conservative, individuals and numbers of species: For each sample Warwick (1984) concluded that it resulted from evo- series and each mesh size, a sufficient number of lutionary mechanisms that determine the size struc- samples was analysed for the species numbers to ture of the regional species pool. This explanation reach an acceptable asymptote, so that a relatively was predicated by the early observation of Hutchin- large number of additional samples would have son & MacArthur (1959 p. 121) that needed to be examined to add an appreciable num- The ordinary symmetrical logarithmic curve we ber of species. Bett’s (2013) proposition that this con- take to imply that a given type of organization im - servative bimodal pattern is the result of a sampling plies a certain class of particularly appropriate niches and that as more extreme niches are occupied the artefact is therefore unacceptable. Bett’s simulations amount of evolutionary displacement needed is about also fail to explain the fact that this pattern is modi- as easy to achieve, with a given type of organization, fied by water depth and disturbance: The proportion by change in size of a given factor in either direction. of macrofauna species increases from disturbed inter- [emphasis added] tidal situations to more stable deeper water (Warwick Warwick (1984) thus postulated that 2 separate log- 1984), and mechanisms of resource partitioning and normal distributions would result if there were 2 diversity maintenance in the meiobenthos and macro- more or less separate types of organisation with no benthos may be affected differentially by sediment intermediates. Contrary to Bett’s (2013) assertion, disturbance (Warwick et al. 1990). many, if not most, functional traits of marine benthic animals switch at about 45 µg dry weight (Table 1; see Warwick 1984 for discussion of further implica- Meiobenthos and macrobenthos as discrete entities tions of this phenomenon). Credence to this explana- tion based on 2 levels of functional organisation is Most controversial is Bett’s (2013) questioning of the provided by the fact that, in freshwater sediments, well-established view that meiobenthos and macro - the meiofauna-macrofauna dichotomy does not seem benthos are discrete coherent entities each with dis- to exist, since the species size spectrum is unimodal Warwick: Meiobenthos and macrobenthos — Comment on Bett (2013) 297 Table 1. Size related biological traits that switch abruptly at about turbation in terms of species body size are 45 µg dry weight (final adult body size) in temperate shallow water also distinctly different. Using a series of marine benthos (from Warwick 1984) samples sieved using 5 mesh sizes (63, 125, 250, 500 and 1000 µm), Warwick et al. Less than 45 µg More than 45 µg (2006) showed that diversity profiles were Development Direct benthic Planktonic very consistent within the size classes of Dispersal As adults Planktonic larvae meiobenthos and macrobenthos of an inter- Generation time Less than 1 yr More than 1 yr tidal sand-flat, regardless of the mesh size Reproduction Semelparous Iteroparous (usually) used to extract the fauna, but were very dif- Feeding Discriminate use of Indiscriminate use of particles particles ferent between these 2 categories (Fig. 2). Resource Particle selection Spatial segregation and Bett’s theory would predict a gradual partitioning (size, shape, quality) particle size selection change in species composition when sam- Growth Reach asymptotic Continue growth pling a continuum of body sizes using a adult size throughout life series of sieves with increasing mesh size, Mobility Motile Sedentary or motile since different life stages of each species would be captured by more than 1 sieve (Strayer 1986, Stead et al. 2005). However, the repro- size. However, Warwick et al. (2006) demonstrated ductive characteristics of the species in volved are a dramatic change in species composition at the mei - very different: While most shallow-water temperate ofauna-macrofauna size boundary (250 to 500 µm), marine macrobenthos have planktonic larvae, many with only small gradual changes within the meio- species in the freshwater benthos are the larval benthic and macrobenthic size ranges (Fig. 3). Fur- stages of flying insects (Lopez 1988). Unimodal rela- thermore, high levels of organic enrichment result tionships of body size and diversity, with most spe- in dramatic increases in abundance of the largest cies being of intermediate size, occur globally, species of meiobenthos, mainly oncholaimid nema- regionally and locally in many environments and can todes