A multi-model approach to evaluate the role of environmental variability and fishing pressure in sardine fisheries

Understanding the fluctuations in population abundance is a central question in fisheries. Sardine fisheries is of great importance to Portugal and is data-rich and of primary concern to fisheries managers. In Portugal, sub-stocks of Sardina pilchardus (sardine) are found in different regions: the Northwest (IXaCN), Southwest (IXaCS) and the South coast (IXaS-Algarve). Each of these sardine sub-stocks is affected differently by a unique set of climate and ocean conditions, mainly during larval development and recruitment, which will consequently affect sardine fisheries in the short term. Taking this hypothesis into consideration we examined the effects of hydrographic (river discharge), sea surface temperature, wind driven phenomena, upwelling, climatic (North Atlantic Oscillation) and fisheries variables (fishing effort) on S. pilchardus catch rates (landings per unit effort, LPUE, as a proxy for sardine biomass). A 20-year time series (1989-2009) was used, for the different subdivisions of the Portuguese coast (sardine sub-stocks). For the purpose of this analysis a multi-model approach was used, applying different time series models for data fitting (Dynamic Factor Analysis, Generalised Least Squares), forecasting (Autoregressive Integrated Moving Average), as well as Surplus Production stock assessment models. The different models were evaluated, compared and the most important variables explaining changes in LPUE were identified. The type of relationship between catch rates of sardine and environmental variables varied across regional scales due to region-specific recruitment responses. Seasonality plays an important role in sardine variability within the three study regions. In IXaCN autumn (season with minimum spawning activity, larvae and egg concentrations) SST, northerly wind and wind magnitude were negatively related with LPUE. In IXaCS none of the explanatory variables tested was clearly related with LPUE. In IXaS-Algarve (South Portugal) both spring (period when large abundances of larvae are found) northerly wind and wind magnitude were negatively related with LPUE, revealing that environmental effects match with the regional peak in spawning time. Overall, results suggest that management of small, short-lived pelagic species, such as sardine quotas/sustainable yields, should be adapted to a regional scale because of regional environmental variability.

Climate drives temporal replacement and nested‐resultant richness patterns of Scottish coastal vegetation

Beta diversity quantifies spatial and/or temporal variation in species composition. It is comprised of two distinct components, species replacement and nestedness, which derive from opposing ecological processes. Using Scotland as a case study and a β-diversity partitioning framework, we investigate temporal replacement and nestedness patterns of coastal grassland species over a 34-yr time period. We aim to 1) understand the influence of two potentially pivotal processes (climate and land-use changes) on landscape-scale (5 × 5 km) temporal replacement and nestedness patterns, and 2) investigate whether patterns from one β-diversity component can mask observable patterns in the other.

We summarised key aspects of climate driven macro-ecological variation as measures of variance, long-term trends, between-year similarity and extremes, for three important climatic predictors (minimum temperature, water-balance and growing degree-days). Shifts in landscape-scale heterogeneity, a proxy of land-use change, was summarised as a spatial multiple-site dissimilarity measure. Together, these climatic and spatial predictors were used in a multi-model inference framework to gauge the relative contribution of each on temporal replacement and nestedness patterns.

Temporal β-diversity patterns were reasonably well explained by climate change but weakly explained by changes in landscape-scale heterogeneity. Climate was shown to have a greater influence on temporal nestedness than replacement patterns over our study period, linking nestedness patterns, as a result of imbalanced gains and losses, to climatic warming and extremes respectively. Important climatic predictors (i.e. growing degree-days) of temporal β-diversity were also identified, and contrasting patterns between the two β-diversity components revealed.

Results suggest climate influences plant species recruitment and establishment processes of Scotland's coastal grasslands, and while species extinctions take time, they are likely to be facilitated by climatic perturbations. Our findings also highlight the importance of distinguishing between different components of β-diversity, disentangling contrasting patterns than can mask one another.

Using life strategies to explore the vulnerability of ecosystem services to invasion by alien plants

Invasive plants can have different effects of ecosystem functioning and on the provision of ecosystem services, from strongly deleterious impacts to positive effects. The nature and intensity of such effects will depend on the service and ecosystem being considered, but also on features of life strategies of invaders that influence their invasiveness as well as their influence of key processes of receiving ecosystems. To address the combined effect of these various factors we developed a robust and efficient methodological framework that allows to identify areas of possible conflict between ecosystem services and alien invasive plants, considering interactions between landscape invasibility and species invasiveness. Our framework combines the statistical robustness of multi-model inference, efficient techniques to map ecosystem services, and life strategies as a functional link between invasion, functional changes and potential provision of services by invaded ecosystems. The framework was applied to a test region in Portugal, for which we could successfully predict the current patterns of plant invasion, of ecosystem service provision, and finally of probable conflict (expressing concern for negative impacts, and value for positive impacts on services) between alien species richness (total and per plant life strategy) and the potential provision of selected services. Potential conflicts were identified for all combinations of plant strategy and ecosystem service, with an emphasis for those concerning conflicts with carbon sequestration, water regulation and wood production. Lower levels of conflict were obtained between invasive plant strategies and the habitat for biodiversity supporting service. The added value of the proposed framework in the context of landscape management and planning is discussed in perspective of anticipation of conflicts, mitigation of negative impacts, and potentiation of positive effects of plant invasions on ecosystems and their services.

Big Decisions and Sparse Data: Adapting Scientific Publishing to the Needs of Practical Conservation

The biggest challenge in conservation biology is breaking down the gap between research and practical management. A major obstacle is the fact that many researchers are unwilling to tackle projects likely to produce sparse or messy data because the results would be difficult to publish in refereed journals. The obvious solution to sparse data is to build up results from multiple studies. Consequently, we suggest that there needs to be greater emphasis in conservation biology on publishing papers that can be built on by subsequent research rather than on papers that produce clear results individually. This building approach requires: (1) a stronger theoretical framework, in which researchers attempt to anticipate models that will be relevant in future studies and incorporate expected differences among studies into those models; (2) use of modern methods for model selection and multi-model inference, and publication of parameter estimates under a range of plausible models; (3) explicit incorporation of prior information into each case study; and (4) planning management treatments in an adaptive framework that considers treatments applied in other studies. We encourage journals to publish papers that promote this building approach rather than expecting papers to conform to traditional standards of rigor as stand-alone papers, and believe that this shift in publishing philosophy would better encourage researchers to tackle the most urgent conservation problems.

Landscape characteristics associated with species richness and occurrence of small native mammals inhabiting a coastal heathland: a spatial modelling approach

In the coastal region of south-western Victoria, Australia, populations of native small mammal species are restricted to patches of suitable habitat in a highly fragmented landscape. The size and spatial arrangement of these patches is likely to influence both the occupancy and richness of species at a location. Geographic Information System (GIS)-based habitat models of the species richness of native small mammals, and individual species  occurrences, were developed to produce maps displaying the spatial  configuration of suitable habitat. Models were generated using either generalised linear Poisson regression (for species richness) or logistic regression (for species occurrences) with species richness or  presence/absence as the dependent variable and landscape variables, extracted from both GIS data layers and multi-spectral digital imagery, as the predictor variables. A multi-model inference approach based on the Akaike Information Criterion was used and the resulting model was applied in a GIS framework to extrapolate predicted richness/likelihood of occurrence across the entire area of the study. A negative association between species  richness and elevation, habitat complexity and sun index indicated that richness within the study area decreases with increasing altitude, vertical vegetation structure and exposure to solar radiation. Landform  characteristics were important (to varying degrees) in determining habitat occupancy for all of the species examined, while the influence of habitat complexity was important for only one of the species. Performance of all but one of the models generated using presence/absence data was high, as indicated by the area under the curve of a receiver-operating characteristic plot. The effective conservation of the small mammal species in the area of concern is likely to depend on management actions that promote the protection of the critical habitats identified in the models.

Environment and dispersal paths override life strategies and residence time in determining regional patterns of invasion by alien plants

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Allometric growth of a common Nassariidae (Gastropoda) in south-east Brazil

The allometric growth of two groups of Nassarius vibex on beds of the bivalve Mytella charruana on the northern coast of the State of Sao Paulo, was evaluated between September 2006 and February 2007 in the bed on Camaroeiro Beach, and from March 2007 to June 2007 at Cidade Beach. The shells from Camaroeiro were longer and wider and had a smaller shell aperture than those from Cidade; a principal components analysis also confirmed different morphometric patterns between the areas. The allometric growth of the two groups showed great variation in the development of individuals. The increase of shell width and height in relation to shell length did not differ between the two areas. Shell aperture showed a contrasting growth pattern, with individuals from Camaroeiro having smaller apertures. The methodology based on Kullback-Leibler information theory and the multi-model inference showed, for N. vibex, that the classic linear allometric growth was not the most suitable explanation for the observed morphometric relationships. The patterns of relative growth observed in the two groups of N. vibex may be a consequence of different growth and variation rates, which modifies the development of the individuals. Other factors such as food resource availability and environmental parameters, which might also differ between the two areas, should also be considered.

Dynamic multi-relational Chinese restaurant process for analyzing influences on users in social media

We study the problem of analyzing influence of various factors affecting individual messages posted in social media. The problem is challenging because of various types of influences propagating through the social media network that act simultaneously on any user. Additionally, the topic composition of the influencing factors and the susceptibility of users to these influences evolve over time. This problem has not been studied before, and off-the-shelf models are unsuitable for this purpose. To capture the complex interplay of these various factors, we propose a new non-parametric model called the Dynamic Multi-Relational Chinese Restaurant Process. This accounts for the user network for data generation and also allows the parameters to evolve over time. Designing inference algorithms for this model suited for large scale social-media data is another challenge. To this end, we propose a scalable and multi-threaded inference algorithm based on online Gibbs Sampling. Extensive evaluations on large-scale Twitter and Face book data show that the extracted topics when applied to authorship and commenting prediction outperform state-of-the-art baselines. More importantly, our model produces valuable insights on topic trends and user personality trends beyond the capability of existing approaches.

Matching image sets via adaptive multi convex hull

Traditional nearest points methods use all the samples in an image set to construct a single convex or affine hull model for classification. However, strong artificial features and noisy data may be generated from combinations of training samples when significant intra-class variations and/or noise occur in the image set. Existing multi-model approaches extract local models by clustering each image set individually only once, with fixed clusters used for matching with various image sets. This may not be optimal for discrimination, as undesirable environmental conditions (eg. illumination and pose variations) may result in the two closest clusters representing different characteristics of an object (eg. frontal face being compared to non-frontal face). To address the above problem, we propose a novel approach to enhance nearest points based methods by integrating affine/convex hull classification with an adapted multi-model approach. We first extract multiple local convex hulls from a query image set via maximum margin clustering to diminish the artificial variations and constrain the noise in local convex hulls. We then propose adaptive reference clustering (ARC) to constrain the clustering of each gallery image set by forcing the clusters to have resemblance to the clusters in the query image set. By applying ARC, noisy clusters in the query set can be discarded. Experiments on Honda, MoBo and ETH-80 datasets show that the proposed method outperforms single model approaches and other recent techniques, such as Sparse Approximated Nearest Points, Mutual Subspace Method and Manifold Discriminant Analysis.

The role of northern Arabian Sea surface temperature biases in CMIP5 model simulations and future projections of Indian summer monsoon rainfall

Many climate models have problems simulating Indian summer monsoon rainfall and its variability, resulting in considerable uncertainty in future projections. Problems may relate to many factors, such as local effects of the formulation of physical parametrisation schemes, while common model biases that develop elsewhere within the climate system may also be important. Here we examine the extent and impact of cold sea surface temperature (SST) biases developing in the northern Arabian Sea in the CMIP5 multi-model ensemble, where such SST biases are shown to be common. Such biases have previously been shown to reduce monsoon rainfall in the Met Office Unified Model (MetUM) by weakening moisture fluxes incident upon India. The Arabian Sea SST biases in CMIP5 models consistently develop in winter, via strengthening of the winter monsoon circulation, and persist into spring and summer. A clear relationship exists between Arabian Sea cold SST bias and weak monsoon rainfall in CMIP5 models, similar to effects in the MetUM. Part of this effect may also relate to other factors, such as forcing of the early monsoon by spring-time excessive equatorial precipitation. Atmosphere-only future time-slice experiments show that Arabian Sea cold SST biases have potential to weaken future monsoon rainfall increases by limiting moisture flux acceleration through non-linearity of the Clausius-Clapeyron relationship. Analysis of CMIP5 model future scenario simulations suggests that, while such effects are likely small compared to other sources of uncertainty, models with large Arabian Sea cold SST biases suppress the range of potential outcomes for changes to future early monsoon rainfall.

Reliability of regional climate model trends

A necessary condition for a good probabilistic forecast is that the forecast system is shown to be reliable: forecast probabilities should equal observed probabilities verified over a large number of cases. As climate change trends are now emerging from the natural variability, we can apply this concept to climate predictions and compute the reliability of simulated local and regional temperature and precipitation trends (1950–2011) in a recent multi-model ensemble of climate model simulations prepared for the Intergovernmental Panel on Climate Change (IPCC) fifth assessment report (AR5). With only a single verification time, the verification is over the spatial dimension. The local temperature trends appear to be reliable. However, when the global mean climate response is factored out, the ensemble is overconfident: the observed trend is outside the range of modelled trends in many more regions than would be expected by the model estimate of natural variability and model spread. Precipitation trends are overconfident for all trend definitions. This implies that for near-term local climate forecasts the CMIP5 ensemble cannot simply be used as a reliable probabilistic forecast.

Present state of global wetland extent and wetland methane modelling: methodology of a model inter-comparison project (WETCHIMP)

The Wetland and Wetland CH4 Intercomparison of Models Project (WETCHIMP) was created to evaluate our present ability to simulate large-scale wetland characteristics and corresponding methane (CH4) emissions. A multi-model comparison is essential to evaluate the key uncertainties in the mechanisms and parameters leading to methane emissions. Ten modelling groups joined WETCHIMP to run eight global and two regional models with a common experimental protocol using the same climate and atmospheric carbon dioxide (CO2) forcing datasets. We reported the main conclusions from the intercomparison effort in a companion paper (Melton et al., 2013). Here we provide technical details for the six experiments, which included an equilibrium, a transient, and an optimized run plus three sensitivity experiments (temperature, precipitation, and atmospheric CO2 concentration). The diversity of approaches used by the models is summarized through a series of conceptual figures, and is used to evaluate the wide range of wetland extent and CH4 fluxes predicted by the models in the equilibrium run. We discuss relationships among the various approaches and patterns in consistencies of these model predictions. Within this group of models, there are three broad classes of methods used to estimate wetland extent: prescribed based on wetland distribution maps, prognostic relationships between hydrological states based on satellite observations, and explicit hydrological mass balances. A larger variety of approaches was used to estimate the net CH4 fluxes from wetland systems. Even though modelling of wetland extent and CH4 emissions has progressed significantly over recent decades, large uncertainties still exist when estimating CH4 emissions: there is little consensus on model structure or complexity due to knowledge gaps, different aims of the models, and the range of temporal and spatial resolutions of the models.

Host model uncertainties in aerosol radiative forcing estimates: results from the AeroCom Prescribed intercomparison study

Simulated multi-model “diversity” in aerosol direct radiative forcing estimates is often perceived as a measure of aerosol uncertainty. However, current models used for aerosol radiative forcing calculations vary considerably in model components relevant for forcing calculations and the associated “host-model uncertainties” are generally convoluted with the actual aerosol uncertainty. In this AeroCom Prescribed intercomparison study we systematically isolate and quantify host model uncertainties on aerosol forcing experiments through prescription of identical aerosol radiative properties in twelve participating models. Even with prescribed aerosol radiative properties, simulated clear-sky and all-sky aerosol radiative forcings show significant diversity. For a purely scattering case with globally constant optical depth of 0.2, the global-mean all-sky top-of-atmosphere radiative forcing is −4.47Wm−2 and the inter-model standard deviation is 0.55Wm−2, corresponding to a relative standard deviation of 12 %. For a case with partially absorbing aerosol with an aerosol optical depth of 0.2 and single scattering albedo of 0.8, the forcing changes to 1.04Wm−2, and the standard deviation increases to 1.01W−2, corresponding to a significant relative standard deviation of 97 %. However, the top-of-atmosphere forcing variability owing to absorption (subtracting the scattering case from the case with scattering and absorption) is low, with absolute (relative) standard deviations of 0.45Wm−2 (8 %) clear-sky and 0.62Wm−2 (11 %) all-sky. Scaling the forcing standard deviation for a purely scattering case to match the sulfate radiative forcing in the Aero- Com Direct Effect experiment demonstrates that host model uncertainties could explain about 36% of the overall sulfate forcing diversity of 0.11Wm−2 in the AeroCom Direct Radiative Effect experiment.

Exploring the impact of CMIP5 model biases on the simulation of North Atlantic decadal variability

Instrumental observations, palaeo-proxies, and climate models suggest significant decadal variability within the North Atlantic subpolar gyre (NASPG). However, a poorly sampled observational record and a diversity of model behaviours mean that the precise nature and mechanisms of this variability are unclear. Here, we analyse an exceptionally large multi-model ensemble of 42 present-generation climate models to test whether NASPG mean state biases systematically affect the representation of decadal variability. Temperature and salinity biases in the Labrador Sea co-vary and influence whether density variability is controlled by temperature or salinity variations. Ocean horizontal resolution is a good predictor of the biases and the location of the dominant dynamical feedbacks within the NASPG. However, we find no link to the spectral characteristics of the variability. Our results suggest that the mean state and mechanisms of variability within the NASPG are not independent. This represents an important caveat for decadal predictions using anomaly-assimilation methods.

Inference in differences-in-differences with few treated groups and heteroskedasticity

Differences-in-Differences (DID) is one of the most widely used identification strategies in applied economics. However, how to draw inferences in DID models when there are few treated groups remains an open question. We show that the usual inference methods used in DID models might not perform well when there are few treated groups and errors are heteroskedastic. In particular, we show that when there is variation in the number of observations per group, inference methods designed to work when there are few treated groups tend to (under-) over-reject the null hypothesis when the treated groups are (large) small relative to the control groups. This happens because larger groups tend to have lower variance, generating heteroskedasticity in the group x time aggregate DID model. We provide evidence from Monte Carlo simulations and from placebo DID regressions with the American Community Survey (ACS) and the Current Population Survey (CPS) datasets to show that this problem is relevant even in datasets with large numbers of observations per group. We then derive an alternative inference method that provides accurate hypothesis testing in situations where there are few treated groups (or even just one) and many control groups in the presence of heteroskedasticity. Our method assumes that we know how the heteroskedasticity is generated, which is the case when it is generated by variation in the number of observations per group. With many pre-treatment periods, we show that this assumption can be relaxed. Instead, we provide an alternative application of our method that relies on assumptions about stationarity and convergence of the moments of the time series. Finally, we consider two recent alternatives to DID when there are many pre-treatment groups. We extend our inference method to linear factor models when there are few treated groups. We also propose a permutation test for the synthetic control estimator that provided a better heteroskedasticity correction in our simulations than the test suggested by Abadie et al. (2010).