Leon Metzeling

Classification and prediction of macroinvertebrate assemblages from running waters in Victoria, Australia

We constructed predictive models using 2 macroinvertebrate data sets (for both species and family) from bankside habitats at 49 undisturbed reference sites from 6 Victorian river basins; data were accumulated over 4 to 6 sampling occasions. Classification (by unweighted pair-group arithmetic averaging with the Bray-Curtis association measure) showed 3 site groups were evident at the species level and 4 at the family level. A subset of 5 of 22 environmental variables provided maximum discrimination (using stepwise discriminant analysis) between the 3 species site groups; these variables were: conductivity, altitude, substrate heterogeneity, distance of a site from source, and longitude. Four variables discriminated between the 4 family site groups: conductivity, catchment area upstream of site, mean annual discharge, and latitude. From the discriminant analysis, it was possible to predict the group into which an unknown site (specified only by measurements on the 4 or 5 variables just noted) would be placed and thus the probabilities of occurrence of taxa at this site. To test predictive ability, 4 sites were removed at random from the 2 data sets and the classification and discriminant models were recalculated. This process was repeated 5 times. The identity and number of taxa observed at each of these sites were compared with those predicted with a probability of occurrence >50% and the results expressed as a ratio of numbers observed to numbers expected (O/E). This ratio varied from 0.75 to 1.05 at the species level and from 0.83 to 1.12 at the family level, indicating that the fauna conformed with expectation (O/E near 1.0). To test such predictive models on independent data, O/E ratios were also calculated for family data collected in spring at 18 sites from a basin not used in the original models. Two new discriminant models based on single sets of samples from the reference sites taken in spring were constructed for this purpose. O/E ratios varied from 0.09 to 1.01 for the 18 sites and were inversely correlated (r = -0.4 to -0.8) with a range of water quality variables, the values of which increased as water quality deteriorated. The O/E ratio could thus be considered a sensitive measure of disturbance.

A trait database of stream invertebrates for the ecological risk assessment of single and combined effects of salinity and pesticides in South-East Australia

We compiled a database on a priori selected traits for South-East Australian freshwater macroinvertebrate families and used this data for the development of a biotic indicator for the detection of the effects of salinisation on freshwater communities (SPEAR(salinity)) and for the adaptation of the existing SPEAR(pesticides) index for South-East Australian taxa. The SPEAR(salinity) indicator showed a reasonably high relationship (0.38≤r(2)≤0.5) with salinity in terms of logarithmic electrical conductivity (log EC) using field biomonitoring data from 835 pools and riffle sites in Victoria and South Australia. Several other biotic indexes that were calculated for comparison purpose exhibited a lower relationship with log EC. In addition, SPEAR(salinity) was the only indicator that did not respond to other water quality variables and was therefore most selective. We used log EC data and modelled pesticide exposure for sites in Victoria in concert with SPEAR(salinity) and the existing SPEAR(pesticides) index to assess whether pesticides interacts with effects of salinity on invertebrate communities and vice versa. No interaction with pesticides was found for the effect of log EC on SPEAR(salinity), whereas EC interacted with the estimated pesticide exposure on the invertebrate communities. To foster the development of further trait-based ecological indicators, we suggest a conceptual model that predicts response traits based on the disturbance regime and disturbance mode of action of the stressor. Biotic indicators based on a priori selected traits represent a promising biomonitoring tool even for regions where ecological information is scarce.

The definition of species richness used by species sensitivity distributions approximates observed effects of salinity on stream macroinvertebrates

The risk of chemicals for ecological communities is often forecast with species sensitivity distributions (SSDs) which are used to predict the concentration which will protect p% of species (PCp value). However, at the PCp value, species richness in nature would not necessary be p% less than at uncontaminated sites. The definition of species richness inherent to SSDs (contaminant category richness) contrasts with species richness typically measured in most field studies (point richness). We determine, for salinity in eastern Australia, whether these definitions of stream macroinvertebrate species richness are commensurable. There were strong relationships (r2≥0.87) between mean point species, family and Ephemeroptera, Trichoptera and Plecoptera species richness and their respective contamination category richness. Despite differences in the definition of richness used by SSDs and field biomonitoring, their results in terms of relative species loss from salinity in south-east Australia are similar. We conclude that in our system both definitions are commensurable.

Temporal persistence of benthic invertebrate communities in south-eastern Australian streams: taxonomic resolution and implications for the use of predictive models

Benthic macroinvertebrates are commonly used to monitor the condition of rivers and streams. Predictive models and biological objectives used in environmental policies, two tools for assessing stream condition, rely on the assumption that the communities from which the invertebrates are derived are sufficiently stable or persistent over time for valid comparisons to be made with test sites sampled years afterwards. There has only been limited testing of this in Australia and there has been no formal programme established to assess long-term changes in aquatic ecosystems. In this paper, data sets collected from the Latrobe and Yarra river systems in south-eastern Australia, sampled over periods of up to 20 years between initial and final samplings, were examined. Using multivariate analyses, it was found that stream communities were persistent at the taxonomic level of family, but significant temporal changes were apparent using species-level data. Tests of rank abundance (Friedmans and Spearman rank correlation) gave mixed results but generally lead to the same conclusions. The implications of these results are discussed with regard to changes in faunal composition, stability of rank abundances, impacts on the use of predictive models and biological objectives, and possible links to major environmental features.

Macroinvertebrate regionalisation for use in the management of aquatic ecosystems in Victoria, Australia

The development of a broader, more holistic approach to aquaticecosystem management has been called for in recent years. Physical and chemical objectives alone are no longer consideredsufficient for the protection of aquatic ecosystems and shouldbe supplemented with biological objectives. The ubiquitousand sedentary nature of macroinvertebrates, combined with theirmeasurable response to environmental conditions, favour their use as important indicators in environmental policies. To establish biological objectives, there is a need for a regionalframework to limit the variability between ecosystems. Past studies have demonstrated that an a posteriori regionalisationapproach may be more useful than an a priori approach in explaining single component (e.g. macroinvertebrates) patternsacross ecosystems. This is particularly important as aquaticresource management agencies often focus on one or twocomponents of the ecosystem to assess environmental health. This study uses an a posteriori method to delineate and describebiological regions based on edge and riffle macroinvertebrate data. The regionalisation will provide a framework for settingbiological objectives, based on the range of reference conditionsmeasured within each separate region. The objectives will includeregional checklists for taxa and biotic indices. Predictive modelling in the style of RIVPACS or AUSRIVAS will also be usedwithin each region to develop objectives, incorporating local, regional and systematic features as predictor variables.

River conservation in a changing world: invertebrate diversity and spatial prioritisation in south-eastern coastal Australia

Concentration of human populations with likely impacts of climate change present major challenges for river conservation in the south-eastern coastal region of Australia. Quantitative methods for spatial prioritisation of conservation actions can play a major role in meeting these challenges. We examined how these methods may be applied to help plan for potential impacts of climate change in the region, using macroinvertebrate assemblages as surrogates of river biodiversity. Environmental gradients explaining broad-scale patterns in the composition of macroinvertebrate assemblages are well represented in protected areas; however, their effectiveness for conserving river biodiversity with climate change depends on linking management inside and outside protected areas. Projected increases in temperature and sea level may be used to prioritise conservation to counter likely major impacts in high-altitude zones and the coastal fringes, whereas elsewhere, considerable uncertainty remains in the absence of better downscaled projections of rainfall. Applying such spatial prioritisations using biodiversity surrogates could help river-focussed conservation around the world.

A comparison of diatom and macroinvertebrate classification of sites in the Kiewa River system, Australia

Diatoms and macroinvertebrates are both commonly used for biological assessment of stream condition. As the use of biological assessment techniques increases, resource managers will need to make decisions on which biological tool to use for a particular study. In a study of the Kiewa River, Victoria, Australia we assessed these two components of the biota—macroinvertebrates and diatoms—using indices and pattern analysis, and comparing them with an a priori landscape classification. We also assessed the relationship exhibited between the biological results and environmental variables which are usually significant in stream ecosystems. To make the data comparable we used categorical abundances for both data sets. The pattern analyses showed complementary results, with diatoms more closely related to water quality variables, whereas macroinvertebrates were primarily related to catchment and habitat features. An analysis of a combined data set (diatoms plus macroinvertebrates) showed no extra information was gained. Using categorisation to create consistency between data sets was shown to reduce the information and affect results from the diatom analyses. The results suggested that the locally derived bioassessment models and indices provided a more accurate assessment of the sites than the overseas-derived diatom index. The outcomes are complicated by issues of data weighting, whereby a presence/absence diatom index may have performed better than abundance-weighted indices due to strong dominance of one or two species at a site. Future comparisons will benefit from an increase in the knowledge of regional diatom taxonomy and autecology.

Regional and local species diversity patterns for lotic invertebrates across multiple drainage basins in Victoria

Regional (RSR) and local species richness (LSR) was recorded for stream invertebrate communities at reference sites in 25 drainage basins in Victoria. Regional species richness was defined as the total number of species recorded at all reference sites within a basin, and LSR as the total numbers of species recorded at a single reference site. Records were obtained from bank and channel habitats and analysed separately. Regressions between LSR and RSR indicated a proportional or linear relationship in both habitats. This applied to the whole data set and to subgroups representing Ephemeroptera, Plecoptera and Trichoptera (EPT as a group), Hemiptera, Diptera, and Coleoptera. All data sets thus represented communities in which no upper limit to LSR was observed. Multiple regressions between LSR and RSR, number of samples per site (N) and seven physical variables showed that RSR and N were nearly always significantly related to LSR. Few of the physical variables were significant except conductivity (for EPT and Coleoptera). Multidimensional scaling ordinations revealed an east-west gradient in compositional similarity of invertebrates, upon which variations in RSR had a major influence. Investigation of factors that regulate RSR will thus be necessary for a broad scale view of what regulates LSR.

Evaluation of the sample size used for the rapid bioassessment of rivers using macroinvertebrates

Stream biological monitoring programs using benthic invertebrates have been implemented in all States and Territories of Australia in recent years. Although some variations exist in sorting strategies, all have adopted a rapid bioassessment approach with the collection of a single large sample from specified habitats at a site. However, the adequacy of the size of the sample collected has never been assessed. In this study, we examined data collected from rivers in three different States (Victoria, Queensland and Western Australia – each from a distinct bioregion of Australia), from four habitats (riffles, edges, macrophytes and pool rocks) and collected over four sample sizes (smaller and larger than that the national standard of 10 m or 10–20 pool rocks). We also used a subset of the data to examine the interaction of taxonomic resolution (family vs species) with sample size. All samples were collected using a live sorting approach which aimed at maximizing taxa richness while collecting about 200 animals.We found that the current recommended sample size adequately described the invertebrate community at a site in comparison to samples of other sizes. There were some differences between the States and these varied with habitat. In some instances, smaller sized samples would be adequate for monitoring purposes. Taxonomic level had little effect with only the riffle species samples showing a significant difference between sizes in contrast to the family level data which showed no difference.

Risk assessment of salinity and turbidity in Victoria (Australia) to stream insects' community structure does not always protect functional traits

Ecological risk assessments mostly consider measures of community composition (structure) across large spatial scales. These assessments, using species sensitivity distributions (SSDs) or the relative species retention (RSR), may not be protective of ecosystem functions and services at smaller spatial scales. Here we examine how changes in biological traits, as proxy for ecosystem functions/services, at a fine spatial scale relate to larger scale assessment of structure. We use functional traits of stream insect species in south-east Australia in two habitats (riffle and edge/pool). We find that the protection of community structure in terms of 95% of species over multiple sites against adverse effects of salinity (as electrical conductivity) and turbidity will mostly, but not always, protect traits at smaller scales. Considering different combinations of trait modalities, contaminants and habitat, a mean of 17.5% (range 0%-36.8) of cases would result in under-protection of trait modalities despite protecting species composition (in terms of Jaccards Index). This under-protection of trait modalities is only because of the different spatial scales that community structure and the traits were considered. We recommend that where the protection of biological traits, ecosystem functions or ecosystem services from stressors is a management goal, protective targets should not be solely set using measures of community structure such as SSDs or RSR. To protect both structural and functional attributes separate risk assessments should be done.