Ecological scales, levels and wide-range scaling There is a mushrooming ecological literature on the problem of spatial and temporal scales and its relationship to ecological 'level'; the collections (Powell and Steele, 1995; Peterson and Parker, 1998) contain many recent examples. At first sight, the distinction seems clear enough: scale refers to the spatial extent of an ecological field (such as biomass density, population number density) or to the duration of an ecological process. In contrast, the notion of 'ecological level' refers to a (possibly) corresponding level of 'organization' ('trophic level', 'landscape level', 'canopy level', etc.), and hence to a specific

No abstract

The Optical Plankton Counter (OPC) was used to count individual animals in situ, and to produce a one-dimensional spatial-series from which gap relationships could be quantified at the rnillimeter scale and above, using a Distance to Next Encounter (DNE) technique. Both DNE and onedimensional neighbor analyses indicated that zooplankton distributions in all transects were significantly (p < 0.0001) aggregated into patches. Within patches, zooplankton were effectively (r2 = 0.94) randomly distributed, resulting in important implications for some of the newer foraging models concerning zooplankton. The DNE frequency distributions all exhibited a distinct pattern that would not be expected from single Poisson distribution, indicating patchiness at the meter scale. This allowed calculation of various statistics used to describe in situ patchiness such as: relative percentage of a transect occupied by patches (79 to 89%) and gaps, estimates of patch size (-2 m diameter), and patch densities (7000 to 14 000 organisms m-3).

Owing to a lack of basic information on the biology of zebrafish (Danio rerio), lab managers must often base decisions regarding the care and use of this species on anecdotal information. In an effort to provide researchers with context-specific behavioral information, the authors evaluated shoaling and spawning behaviors in small groups of zebrafish. In each shoaling assay, a fish was given a choice to shoal with either a single fish or a group of three fish. Females preferred to shoal with a group of three individuals rather than with a single individual, regardless of the sex of the other fish. Males preferred groups of three males over single males but preferred single females to groups of three females. In spawning assays, zebrafish were placed in breeding tanks in one of three sex ratios (1 male:1 female; 3 males:1 female; 1 male:3 females). Reproductive efficiency did not differ among groups, but aggression (evaluated according to presence of shed scales) was more frequently observed in the male-dominated treatment group.

[1] Spectral analysis was performed on a series of oceanographic transects collected using an optical plankton counter and conductivity-temperature-depth probe. The ''physical'' time series (i.e., temperature) power spectra showed a single passive scaling relationship across the entire range of sampling scales (1-8192 s) that was expected from turbulence. However, the ''biological'' time series possessed more than one scaling region. The Chl a fluorescence had three scaling regions, a flattened (whitened) intermediate range bound by passive regions at scales approximately <30 and >300 s. Cross-spectral analysis indicated that the chlorophyll-temperature spectra were similar at these scales. The zooplankton biomass had a single break in the power spectrum and was passive only at scales >300 s, the zooplankton-temperature spectra being similar only at these scales. The zooplankton-chlorophyll cross-spectra were often negatively coupled at the intermediate (300-30 s) scale giving a strong indication that zooplankton grazing was affecting the phytoplankton distributions.