Experimental and environmental factors affect spurious detection of ecological thresholds

Ecological threshold of phosphorus load in Baiyangdian Lake based on a PCLake model and ecological network analysis

Ecological thresholds are a useful indicator for implementing ecological management. Many studies determine the thresholds for nutrient loads in lakes based on the maximum allowable concentration of chlorophyll a (Chla), although this neglects the overall performance of the ecosystem. A PCLake model of Baiyangdian (BYD) Lake in northern China was constructed with six ecological network analysis (ENA) indicators that characterized the ecosystem function, system maturity, and food web structure to quantify the overall status of the BYD ecosystem. To my knowledge, this is the first study on the system level responses of the BYD Lake to phosphorus load interference. Different phosphorus load scenarios were designed to simulate the ecological responses of BYD Lake. The simulated results were employed to calculate the ENA indicators. Ecological thresholds were determined through the driving response relationship between the phosphorus load gradient and the ENA indicators. The results show a non-linear transition response of ENA indicator under phosphorus load gradient. As phosphorus load increases, D/H, SOI, and FCI decreases while A/DC, TPP/TR, and TPP/TB increases. This indicates that the overall structure and function of the ecosystem will deteriorate if phosphorus load increases. The phosphorus load thresholds for the overall performance of BYD Lake were 0.50-1.32 mg m-2 d-1, slightly wider than that of Chla (0.53-1.26 mg m-2 d-1). The model results clearly indicate that there is a time-lag phenomenon at the switch points in the response of ENA indicators compared to that of single functional group. In addition, the A/DC, TPP/TR, SOI, and FCI present more time-lag than that of other ENA indicators. These time-lag effects provide a particular opportunity for biodiversity conservation. Therefore, a possible management strategy is proposed to combine system-level and function group-level thresholds, with the ENA-based threshold as the bottom line and the phytoplankton's threshold as the early-warning indicator. This design is expected to be more precise and efficient, by exploiting the advantages of two thresholds, and may benefit for ecological management practices.

Impact of different diets on the survival, pupation, and adult emergence of Culex pipiens biotype molestus larvae, and infectability with the insect-specific Culex Y virus

The current rapidly advancing climate change will affect the transmission of arthropod-borne viruses (arboviruses), mainly through changes in vector populations. Mosquitos of the Culex pipiens complex play a particularly prominent role in virus transmission in central Europe. Factors that contribute to the vector population density and the ability of those vectors to transmit viral pathogens (vector competence) can include nutrition during the larval stages. To test the influence of larval diet on larval survival and adult emergence, as well as vector competence, several diets varying in their nutritional composition were compared using a newly established assay. We tested the effects of 17 diets or diet combinations on the fitness of third-instar larvae of Culex pipiens biotype molestus. Larval survival rates at day 7 ranged from 43.33% to 94.44%. We then selected 3 of the 17 diets (Tetra Pleco, as the routine feed; JBL NovoTab, as the significantly inferior feed; and KG, as the significantly superior feed) and tested the effect of these diets, in combination with Culex Y virus infection, on larval survival rate. All Culex Y virus-infected larvae showed significantly lower larval survival, as well as low pupation and adult emergence rates. However, none of the tested diets in our study had a significant impact on larval survival in combination with viral infection. Furthermore, we were able to correlate several water quality parameters, such as phosphate, nitrate, and ammonium concentration, electrical conductivity, and low O2 saturations, with reduced larval survival. Thus, we were able to demonstrate that Culex Y virus could be a suitable agent to reduce mosquito population density by reducing larval density, pupation rate, and adult emergence rate. When combined with certain water quality parameters, these effects can be further enhanced, leading to a reduced mosquito population density, and reduce the cycle of transmission. Furthermore, we demonstrate, for the first time, the infection of larvae of the mosquito Culex pipiens biotype molestus with a viral pathogen.

Ecological models for estimating breakpoints and prediction intervals

The relationships between an environmental variable and an ecological response are usually estimated by models fitted through the conditional mean of the response given environmental stress. For example, nonparametric loess and parametric piecewise linear regression model (PLRM) are often used to represent simple to complex nonlinear relationships. In contrast, piecewise linear quantile regression models (PQRM) fitted across various quantiles of the response can reveal nonlinearities in its range of variation across the explanatory variable.We assess the number and positions of candidate breakpoints using loess and compare the relative efficiencies of PLRM and PQRM to quantitatively determine the breakpoints' location and precision. We propose a nonparametric method to generate bootstrap confidence intervals for breakpoints using PQRM and prediction bands for loess and PQRM. We illustrated the applications using data from two aquatic studies suspected to exhibit multiple environmental breakpoints: relating a fish multimetric index of community health (MMI) to agricultural activity in wetlands' adjacent drainage basins; and relating cyanobacterial biomass to total phosphorus concentration in Canadian lakes.Two statistically significant breakpoints were detected in each dataset, demarcating boundaries of three linear segments, each with markedly different slopes. PQRM generated less biased, more accurate, and narrower confidence intervals for the breakpoints and narrower prediction bands than PLRM, especially for small samples and large error variability. In both applications, the relationship between the response and environmental variables was weak/nonsignificant below the lower threshold, strong through the midrange of the environmental gradient, and weak/nonsignificant beyond the upper threshold.We describe several advantages of PQRM over PLRM in characterizing environmental relationships where the scatter of points represents natural environmental variation rather than measurement error. The proposed methodology will be useful for detecting multiple breakpoints in ecological applications where the limits of variation are as important as the conditional mean of a function.

Comparison of two commonly used methods for identifying water quality thresholds in freshwater ecosystems using field and synthetic data

Defining ecological thresholds has become increasingly relevant for water resource management. Despite the fact that there has been a rapid expansion in methods to evaluate ecological threshold responses to environmental stressors, evaluation of the relative benefits of various methods has received less attention. This study compares the performance of Gradient Forest (GF) and Threshold Indicator Taxa Analysis (TITAN) for identifying water quality thresholds in both field and synthetic data. Analysis of 14 years of macroinvertebrates data from the Mediterranean catchments of the Torrens and Onkaparinga Rivers, South-Australia, identified electrical conductivity (EC) and total phosphorus (TP) as the most important water quality variables affecting macroinvertebrates. Water quality thresholds for macroinvertebrates identified by both methods largely corresponded at low EC (GF: 400-900 μS cm^-1 vs. TITAN: 407-951 μScm^-1), total phosphorus (TP) (GF: 0.02-0.18 mg L^-1 vs. TITAN: 0.02-0.04 mg L^-1) and total nitrogen (TN) (GF: 0.2 mg L^-1 vs. TITAN: 0.28-0.67 mg L^-1) concentrations. However, multiple GF-derived thresholds, particularly at high stressor concentrations, were representative of low data distribution, and thus need to be considered with caution. In another case study of South Australian diatom data, there were marked differences in GF and TITAN identified thresholds for EC (GF: 5000 μScm^-1 vs. TITAN 1004-2440 μS cm^-1) and TP (GF: 250-500 μg L^-1 vs. TITAN: 11-329 μg L^-1). These differences were due to the fact that while TITAN parsed species responses into negative and positive taxa, GF overestimated thresholds by aggregating the response of taxa that increase and decrease along environmental gradients. Given these findings, we also evaluated the methods' performance using different distributions of synthetic data i.e. with both skewed and uniform distribution of samples and species responses. Both methods identified similar change-points in the case of a uniform environmental gradient, except when species optima were simulated at centre of the gradient. Here GF detected the change-points but TITAN failed to do so. GF also outperformed TITAN when four simulated species change-points were present. Thus, the distribution of species responses and optima and the evenness of the environment gradient can affect the models' performance. This study has shown that both methods are robust in identifying change in species response but threshold identification differs depending both on the analysis used and the nature of ecological data. We recommend the careful application of GF and TITAN, noting these differences in performance, will improve their application for water resource management.

Comparing statistical analyses to estimate thresholds in ecotoxicology

Different methods are used in ecotoxicology to estimate thresholds in survival data. This paper uses Monte Carlo simulations to evaluate the accuracy of three methods (maximum likelihood (MLE) and Markov Chain Monte Carlo estimates (Bayesian) of the no-effect concentration (NEC) model and Piecewise regression) in estimating true and apparent thresholds in survival experiments with datasets having different slopes, background mortalities, and experimental designs. Datasets were generated with models that include a threshold parameter (NEC) or not (log-logistic). Accuracy was estimated using root-mean square errors (RMSEs), and RMSE ratios were used to estimate the relative improvement in accuracy by each design and method. All methods had poor performances in shallow and intermediate curves, and accuracy increased with the slope of the curve. The EC5 was generally the most accurate method to estimate true and apparent thresholds, except for steep curves with a true threshold. In that case, the EC5 underestimated the threshold, and MLE and Bayesian estimates were more accurate. In most cases, information criteria weights did not provide strong evidence in support of the true model, suggesting that identifying the true model is a difficult task. Piecewise regression was the only method where the information criteria weights had high support for the threshold model; however, the rate of spurious threshold model selection was also high. Even though thresholds are an attractive concept from a regulatory and practical point of view, threshold estimates, under the experimental conditions evaluated in this work, should be carefully used in survival analysis or when there are any biological reasons to support the existence of a threshold.

The prevalence of nonlinearity and detection of ecological breakpoints across a land use gradient in streams

Human activities can alter aquatic ecosystems through the input of nutrients and carbon, but there is increasing evidence that these pressures induce nonlinear ecological responses. Nonlinear relationships can contain breakpoints where there is an unexpected change in an ecological response to an environmental driver, which may result in ecological regime shifts. We investigated the occurrence of nonlinearity and breakpoints in relationships between total dissolved nitrogen (TDN), total dissolved phosphorus (TDP), and total dissolved carbon (DOC) concentrations and ecological responses in streams with varying land uses. We calculated breakpoints using piecewise regression, two dimensional Kolmogorov-Smirnov (2DKS), and significant zero crossings (SiZer) methods. We found nonlinearity was common, occurring in half of all analyses, with some evidence of multiple breakpoints. Linearity, by contrast, occurred in less than 14% of cases, on average. Breakpoints were related to land use gradients, with 34–43% agricultural cover associated with DOC and TDN breakpoints, and 15% wetland and 9.5% urban land associated with DOC and nutrient breakpoints, respectively. While these breakpoints are likely specific to our study area, our study contributes to the growing literature of the prevalence and location of ecological breakpoints in streams, providing watershed managers potential criteria for catchment land use thresholds.

Identification of ecological thresholds from variations in phytoplankton communities among lakes: contribution to the definition of environmental standards

In aquatic ecosystems, the identification of ecological thresholds may be useful for managers as it can help to diagnose ecosystem health and to identify key levers to enable the success of preservation and restoration measures. A recent statistical method, gradient forest, based on random forests, was used to detect thresholds of phytoplankton community change in lakes along different environmental gradients. It performs exploratory analyses of multivariate biological and environmental data to estimate the location and importance of community thresholds along gradients. The method was applied to a data set of 224 French lakes which were characterized by 29 environmental variables and the mean abundances of 196 phytoplankton species. Results showed the high importance of geographic variables for the prediction of species abundances at the scale of the study. A second analysis was performed on a subset of lakes defined by geographic thresholds and presenting a higher biological homogeneity. Community thresholds were identified for the most important physico-chemical variables including water transparency, total phosphorus, ammonia, nitrates, and dissolved organic carbon. Gradient forest appeared as a powerful method at a first exploratory step, to detect ecological thresholds at large spatial scale. The thresholds that were identified here must be reinforced by the separate analysis of other aquatic communities and may be used then to set protective environmental standards after consideration of natural variability among lakes.

Field data reveal low critical chemical concentrations for river benthic invertebrates

River ecosystems are of immense ecological and social importance. Despite the introduction of wastewater treatment plants and advanced chemical authorization procedures in Europe, chemical pollution is still a major threat to freshwater ecosystems. Here, large-scale monitoring data was exploited to identify taxon-specific chemical concentrations beyond which benthic invertebrate taxa are unlikely to occur using Threshold Indicator Taxa Analysis (TITAN). 365 invertebrate taxa and 25 organic chemicals including pesticides, pharmaceuticals, plasticisers, flame retardants, complexing agents, a surfactant and poly- and monocyclic aromatic hydrocarbons from a total of 399 sites were analysed. The number of taxa that responded to each of these chemicals varied between 0% and 21%. These sensitive taxa belonged predominantly to the groups Plecoptera, Coleoptera, Trichoptera, Ephemeroptera, Turbellaria, Megaloptera, Crustacea, and Diptera. Strong effects were observed in response to wastewater-associated compounds, confirming that wastewater is an important cause of biological degradation. The majority of change points identified for each compound were well below predicted no-effect concentrations derived from laboratory toxicity studies. Thus, the results show that chemicals are likely to induce effects in the environment at concentrations much lower than expected based on laboratory experiments. Overall, it is confirmed that chemical pollution is still an important factor shaping the distribution of invertebrate taxa, suggesting the need for continued efforts to reduce chemical loads in rivers.

Metamorphosis enhances the effects of metal exposure on the mayfly, Centroptilum triangulifer

The response of larval aquatic insects to stressors such as metals is used to assess the ecological condition of streams worldwide. However, nearly all larval insects metamorphose from aquatic larvae to winged adults, and recent surveys indicate that adults may be a more sensitive indicator of stream metal toxicity than larvae. One hypothesis to explain this pattern is that insects exposed to elevated metal in their larval stages have a reduced ability to successfully complete metamorphosis. To test this hypothesis we exposed late-instar larvae of the mayfly, Centroptilum triangulifer, to an aqueous Zn gradient (32-476 μg/L) in the laboratory. After 6 days of exposure, when metamorphosis began, larval survival was unaffected by zinc. However, Zn reduced wingpad development at concentrations above 139 μg/L. In contrast, emergence of subimagos and imagos tended to decline with any increase in Zn. At Zn concentrations below 105 μg/L (hardness-adjusted aquatic life criterion), survival between the wingpad and subimago stages declined 5-fold across the Zn gradient. These results support the hypothesis that metamorphosis may be a survival bottleneck, particularly in contaminated streams. Thus, death during metamorphosis may be a key mechanism explaining how stream metal contamination can impact terrestrial communities by reducing aquatic insect emergence.

Ecological assessments with algae: a review and synthesis

Algae have been used for a century in environmental assessments of water bodies and are now used in countries around the world. This review synthesizes recent advances in the field around a framework for environmental assessment and management that can guide design of assessments, applications of phycology in assessments, and refinements of those applications to better support management decisions. Algae are critical parts of aquatic ecosystems that power food webs and biogeochemical cycling. Algae are also major sources of problems that threaten many ecosystems goods and services when abundances of nuisance and toxic taxa are high. Thus, algae can be used to indicate ecosystem goods and services, which complements how algal indicators are also used to assess levels of contaminants and habitat alterations (stressors). Understanding environmental managers' use of algal ecology, taxonomy, and physiology can guide our research and improve its application. Environmental assessments involve characterizing ecological condition and diagnosing causes and threats to ecosystems goods and services. Recent advances in characterizing condition include site-specific models that account for natural variability among habitats to better estimate effects of humans. Relationships between algal assemblages and stressors caused by humans help diagnose stressors and establish targets for protection and restoration. Many algal responses to stressors have thresholds that are particularly important for developing stakeholder consensus for stressor management targets. Future research on the regional-scale resilience of algal assemblages, the ecosystem goods and services they provide, and methods for monitoring and forecasting change will improve water resource management.

Emergence flux declines disproportionately to larval density along a stream metals gradient

Effects of contaminants on adult aquatic insect emergence are less well understood than effects on insect larvae. We compared responses of larval density and adult emergence along a metal contamination gradient. Nonlinear threshold responses were generally observed for larvae and emergers. Larval densities decreased significantly at low metal concentrations but precipitously at concentrations of metal mixtures above aquatic life criteria (cumulative criterion accumulation ratio (CCAR) ≥ 1). In contrast, adult emergence declined precipitously at low metal concentrations (CCAR ≤ 1), followed by a modest decline above this threshold. Adult emergence was a more sensitive indicator of the effect of low metals concentrations on aquatic insect communities compared to larvae, presumably because emergence is limited by a combination of larval survival and other factors limiting successful emergence. Thus effects of exposure to larvae are not manifest until later in life (during metamorphosis and emergence). This loss in emergence reduces prey subsidies to riparian communities at concentrations considered safe for aquatic life. Our results also challenge the widely held assumption that adult emergence is a constant proportion of larval densities in all streams.