Antoine Morin

Empirical models predicting primary productivity from chlorophyll a and water temperature for stream periphyton and lake and ocean phytoplankton

Published data on stream periphyton, lake phytoplankton, and ocean phytoplankton were analyzed to 1) quantify regression models relating daily gross primary production (GPP) to chlorophyll a (chl a) standing stock and water temperature, 2) compare regressions across assemblages, and 3) compare the precision of regression estimates of daily primary production and of production integrated over time to those obtained by measurements using radioisotopes. Regression models predicting daily GPP explained between 29% and 86% of the variance in Log GPP with chl a accounting for 28 to 85% of the explained variance. Regression models differed significantly across assemblages. Chlorophyll-specific production, corrected for the effect of temperature, declined with increasing chl a standing stock presumably because of increased self shading, and was lower in stream periphyton than in lake or marine phytoplankton presumably because of reduced nutrient diffusion in algal mats. Gross primary production was more intensely related to water temperature in stream periphyton (Q10 = 2.5) than in either ocean phytoplankton (Q10 = 1.2) or lake phytoplankton (Q10 = 1.4). Precision, measured as the error factor (EF) by which means have to be multiplied or divided to obtain the limits of a 95% confidence interval, was lower for regression estimates of daily production (EF = 3.4-6.7) and for production integrated over time (EF = 3.4) than for measurements of daily production (EF = 1.2-2). Considering the reduced effort required to obtain estimates of primary production using these regression models, we argue that they could be useful when coarse production estimates are sufficient or when adequate resources are not available to make direct measurements.

Relationships between Size Structure of Invertebrate Assemblages and Trophy and Substrate Composition in Streams

Nine stream sites, spanning the range of water total phosphorus concentrations ([TP]) found in Eastern Ontario and Western Québec, were sampled in June 1992 to describe how the size distribution of invertebrates varied along a nutrient gradient and among substrate categories ranging from sand to small boulders. Despite clear differences in the taxonomic composition among sites, the shape of the size distribution was remarkably similar in all sites. Substrate composition affected the overall abundance of invertebrates more than their size distribution. Overall density peaked on fine and coarse gravel and was lower on sand and boulders. Total abundance of invertebrates was highest in eutrophic sites ([TP] > 40 μg/L), although only organisms larger than 1 mm (approximately 1 μg dry mass) showed higher abundances in richer sites. The increase in total invertebrate abundance with increasing total phosphorus concentration suggests that these streams are phosphorus limited. Animals smaller than 1 mm long account for less than 3% of the respiration of the benthic assemblage on all substrates in our streams. Size distribution of invertebrates, although affected by substrate composition and trophy, is remarkably constant and predictable in our streams. Most of the biomass, and consequently the consumption, respiration, and secondary production, is found in relatively large invertebrates. Meiofauna is relatively rare and consequently has little direct influence on community metabolism in the streams we studied.

A Simple Model to Estimate Growth Rate of Lotic Insect Larvae and Its Value for Estimating Population and Community Production

Data from the literature on growth of lotic insects were used to develop an empirical model predicting the instantaneous growth rate (g, in d-1) as a function of individual dry mass (W, in mg) and water temperature (T, °C). Growth rates were found to decrease with increasing size, and to increase with an increase in temperature with an average Q10 of 1.78. The equation Log10g = -2.09 - 0.27 Log10 W + 0.025 T explained 49% of the observed variance in logtransformed growth rates from field studies. The relation between growth rate and body mass and temperature was found to vary among insect orders. On average, Diptera grow 1.5× faster than what is predicted by the general model, whereas Trichoptera grow at only 0.7× the predicted rate. The coefficients for body mass and temperature also varied among orders. On average, body mass affects less the field growth rates of Diptera and Ephemeroptera than rates of Trichoptera. The growth models can be combined with information on the biomass size distribution of populations of single species or communities to estimate secondary production. Estimates so obtained agree closely (r = 0.93, n = 31) with those obtained by using standard methods for populations of single species. For production of communities, computer simulations show that errors in predicted growth rates decrease the precision of estimates but suggest that estimates of production of communities of tens of species calculated from monthly description of biomass size distribution for a year would on average be 70-140% of the true value. Thus, the model may be used to estimate production for species or groups of species with indistinguishable cohorts when estimates of growth rates are not available.

Size distribution of epilithic lotic invertebrates and implications for community metabolism

The size structure of epilithic animal communities was examined in 12 streams of the Ottawa-Hull region, Ontario-Québec border. In all streams, the size distribution of invertebrates in late summer was approximately lognormal, with most animals <2 mm long (about 10 μg dry mass), but with most of the biomass in size classes >2 mm. Size structure differed significantly among streams, but these differences were not correlated with total invertebrate biomass or periphyton abundance. Normalized biomass spectra in all streams were peaked, unlike those reported for marine and freshwater benthic and planktonic communities. It is likely, however, that addition of bacterial, algal, and meiofaunal components to obtain spectra for the entire epilithic community may linearize the spectraaa, yielding approximately equal biomasses in logarithmic size classes. When combined with allometric relationships describing metabolism, the observed size spectra imply that metazoans <1 mm long make negligible contributions to production, consumption, or respiration of the epilithon.

Empirical Models Predicting Population Abundance and Productivity in Lotic Systems

Stream ecologists have yet to produce general predictive models of abundance and productivity of lotic organisms. The large number of factors affecting species abundance, differences in fauna and flora among ecoregions, and the cost of calibrating models combine to put in question our hope of producing general mechanistic models that yield quantitative predictions. Similar seemingly insurmountable difficulties have been recognised in other systems and, at least in limnology, an alternative to mechanistic models has clearly emerged: the use of a set of relatively simple regression models describing empirical relationships that are combined to yield predictions of interest. Although smaller, a similar set of empirical models exists for running water systems, and published data permit the estimation of many more. Here I review a subset of published and unpublished models to illustrate how they can be used to: 1) serve as reference values for the interpretation of measurements made in particular conditions, 2) yield estimates of abundance and productivity of populations and assemblages that can be used directly for management decisions or to test ecological theories, and 3) serve as heuristic tools to generate hypotheses about factors that affect community structure. Two major weaknesses are made clear by this treatment. The 1st is the large uncertainty associated with most predictions. This uncertainty will be reduced only by models that include better predictors, or sacrifice generality for precision. The 2nd weakness is the existing gap in the body of empirical models that prevents some useful, and perhaps more interesting, predictions. This gap will only be filled by a concerted effort of stream ecologists to produce more empirical relationships, to report them in a useful format, and to create data and model repositories. In alleviating these 2 weaknesses lies one of the most challenging tasks of our discipline.

Sieve retention probabilities of stream benthic invertebrates

Abstract Replicate samples of cobbles and loose inorganic and organic matter collected from 3 stream-riffle sites with different periphyton communities were passed through a geometric series of 9 sieves (0.063–16 mm mesh) to quantify sieve retention probabilities of benthic invertebrates. Sieves retained all organisms with a body length >10× the mesh size. Logistic regression models were estimated to describe retention probabilities as functions of body length and mesh size. Retention probability functions differed slightly but significantly among sites, operators, and taxa. Retention probabilities were higher for samples containing filamentous algae, which entangled invertebrates. Size distributions of straight, elongate invertebrates (e.g., oligochaetes, midge larvae) retained by sieves were more variable than distributions of more spherical organisms (e.g., gastropods). On average, the 1-mm sieve retained >90% of invertebrate biomass but <33% of individuals retained by a 63-μm sieve. Size distribution of organisms retained by coarse sieves (≥1 mm), combined with logistic functions predicting retention probability, can be used to describe abundance, biomass, and size distribution of organisms retained by fine sieves. These results suggest that unbiased descriptions of benthic communities can be obtained with relatively little effort by using a minimum sieve mesh size of 1 mm.

Intensity and Importance of Abiotic Control and Inferred Competition on Biomass Distribution Patterns of Simuliidae and Hydropsychidae in Southern Québec Streams

Biomass of Simuliidae larvae was measured on individual rocks in four surveys of southern Québec streams to quantify the intensity and importance of relationships between biomass of black flies and microhabitat features (depth, current velocity, rock size), site properties (distance from the lake, seston quality), and biomass of Hydropsychidae. Microhabitat features and site properties accounted for 14-47% of the variability in biomass of hydropsychids and simuliids on individual rocks, and between 30 and 67% of variability in mean biomass across sites. Addition of biomass of potential competitors did not significantly improve the proportion of variability accounted for by the multiple regression models. Within most sampling sites, biomasses of simuliids and of hydropsychids were positively correlated. However, analysis of microdistribution patterns across sites suggests a small but significant negative partial correlation between hydropsychids and simuliids in summer and a positive partial correlation in winter. Analysis of covariance revealed strong site effects that could not be explained by microhabitat features or biomass of potential competitors. These results suggest that hydropsychids have a negative effect on biomass of simuliids in summer, but that biomass of simuliids in different sites is more influenced by microhabitat features and site properties than by the biomass of hydropsychids. Reanalysis of published work on interactions between simuliids and caddisflies suggests that intensity of competition is relatively constant, but that its importance is overestimated by controlled experiments. Quantitative models describing the distribution of filter feeders across sites should include better correlates of site suitability rather than considering abundance of competing groups.

Inadequacy of size distributions of stream benthic diatoms for environmental monitoring

Abstract Benthic diatom assemblages respond to changes in water quality, and this response is evidenced by shifts in taxonomic composition. As a result, several taxon-based indices have been developed for monitoring purposes. Some authors have suggested that diatom body size might provide a simpler method for bioassessment than taxonomy-based approaches. Moreover, current knowledge of algal ecology suggests that the slopes and intercepts of density size distributions should vary with environmental characteristics. However, results from studies of the relationship between algal size and trophic variables including P have been mixed. Our objectives were to examine normalized density size distributions and richness size distributions of benthic diatoms in streams along a gradient of agricultural land use to determine whether the size distributions changed in relation to environmental variables. Benthic diatoms from 29 streams in eastern Canada were identified and average body size measurements were obtained for each taxon. Normalized density size distributions and richness size distributions were plotted, and slopes and intercepts were compared among sites using general linear models (GLMs). Despite taxonomic differences in the assemblages, slopes of the normalized density size distributions did not differ among the 29 sites, and distribution shape and intercepts for the richness size distributions did not differ among sites. The intercepts of the normalized density size distributions were significantly different among sites, and this variation was attributed in part to the effect of several environmental variables including N and P. However, this difference in intercepts represented a change in density across all size classes, rather than a size-specific change. Environmental variables did not explain a significant amount of the variance in the shapes of density or richness size distributions. These results suggest that the slopes of lotic benthic diatom size distributions do not differ in response to environmental variables including P. Thus, benthic diatom body size should not be used as a proxy for nutrient status in environmental monitoring.

Interpreting ecological patterns generated through simple stochastic processes

The analysis of spatial patterns is fundamental to understanding ecological processes across geographic scales. Through an analysis of two simple, one-dimensional stochastic models, we develop a framework for identifying the scale of processes producing pattern. We show that for some simple model systems spectral analysis identifies exactly the scale of pattern formation. In other, more complicated systems, autocorrelation analysis appears to yield greater insight into the scale of the dynamics producing pattern; in these, the relative importance of processes at different scales can be determined directly from the change in slope of the autocorrelation function.In general, it is not possible to state which technique will be most useful in the analysis of pattern. Spectral analysis and autocorrelation analysis represent duals that can be extended and applied to more complex systems, potentially yielding insight into the nature of a wide variety of spatially determined ecological processes.

Reducing the cost of benthic sample processing by using sieve retention probability models

Estimation of abundance or biomass of benthic invertebrates requires considerable effort to process samples. Consequently, it has been suggested to process only organisms retained by a relatively coarse meshed sieve and apply size-specific correction factors based on the probability that a sieve retains individual organisms. Benthic samples were collected from 10 sites in 2 regions and processed to validate an existing empirical model predicting sieve retention probabilities, to test whether periphyton biomass affects probability of retention, and to determine the optimal strategy that minimizes both cost and variability of estimates. The existing model predicting sieve retention probabilities corrected for organisms lost through sieves and mostly corrected for underestimation of biomass, but this model lead to overestimates of the frequency of the smallest organisms. Inclusion of algal biomass improved slightly the proportion of correct predictions (whether an organism is retained or not by a sieve) by 0.6% relative to the existing model (from 90.8% to 91.4%), and removed the bias. Density and biomass estimates obtained by only processing organisms retained by 1- or 2-mm sieves and applying correction factors derived from the predicted retention probabilities were accurate and only marginally less precise than estimates obtained by processing all organisms. The reduced precision of estimates from subsets of organisms could be compensated by increasing sample size and still lead to a reduction of 40–60% of the number of organisms processed. Even though the use of subsets introduces additional analytical variability, this variability is relatively small compared to the natural spatial variability among replicates.