Comparing copepod time-series in the north of Spain: Spatial autocorrelation of community composition

Four time-series of copepod species biomass in the north of Spain were contrasted to demonstrate spatial autocorrelation of local communities and their responses to short-term local and regional variability in oceanographic conditions. The series represented coastal and oceanic environments along a marked gradient of influence of seasonal upwelling from Galicia to the Mar Cantábrico (S Bay of Biscay), and each one included at least 10 years of continuous data collected at monthly frequency. Community composition (i.e. species number and diversity) was very consistent through the region, but local variations in the presence of new species and the relative proportions of common species allowed for the characterisation of the response to the environment at each site. Small-sized species were more frequent near the coast. A few species, however, captured the main patterns of variability in all series. Calanus helgolandicus and Acartia (mainly Acartia clausi) were generally the main contributors to total biomass, while other species as Paracalanus parvus and Clausocalanus spp. were important only at some locations. Most copepod indices were positively correlated with upwelling, either considering the whole community (biomass, species richness and diversity) or individual species, but only in the coastal series analysed since 1991. Copepods in the nearby ocean, however, showed negative correlations with upwelling in the period 1960–1986. The effects of upwelling may have been modulated by local factors, as showed by the increases in biomass, number of species and diversity in associations with increases in sea surface temperature in Galicia, while in the Mar Cantábrico only the warming-tolerant species increased and those typical of upwelling decreased. Density stratification of the water column was associated with decreases in total copepod biomass in Galicia, while it favoured the increase in species richness in the Mar Cantábrico. Nearly all significant responses of copepods to environmental variability were delayed by up to 5 months, showing the importance of considering time-lags in the analysis of temporal responses of zooplankton.

Spatial autocorrelation in the response of soft-bottom marine benthos to gas extraction activities: The case of amphipods in the Ionian Sea

The spatial distributions of marine fauna and of pollution are both highly structured, and thus the resulting high levels of autocorrelation may invalidate conclusions based on classical statistical approaches. Here we analyse the close correlation observed between proxies for the disturbance associated with gas extraction activities and amphipod distribution patterns around four hydrocarbon platforms. We quantified the amount of variation independently accounted for by natural environmental variables, proxies for the disturbance caused by platforms, and spatial autocorrelation. This allowed us to demonstrate how each of these three factors significantly affects the community structure of amphipods. Sophisticated statistical techniques are required when taking into account spatial autocorrelation: nevertheless our data demonstrate that this approach not only enables the formulation of robust statistical inferences but also provides a much deeper understanding of the subtle interactions between human disturbance and natural factors affecting the structure of marine invertebrates communities. (C) 2012 Elsevier Ltd. All rights reserved.

Role of Predalors, Habitat Attributes, and Spatial Autocorrelation on the Distribution of Eggs in the Northern Spectacled Salamander (Salamandrina perspicillata)

We predicted that the probability of egg occurrence of salamander Salamandrina perspicillata depended on stream features and predation by native crayfish Austropotamobius fulcisianus and the introduced trout Salmo trutta. We assessed the presence of S. perspicillata at 54 sites within a natural reserve of southern Tuscany, Italy. Generalized linear models with binomial errors were constructed using egg presence/absence and altitude, stream mean size and slope, electrical conductivity, water pH and temperature, and a predation factor, defined according to the presence/absence of crayfish and trout. Some competing models also included an autocovariate term, which estimated how much the response variable at any one sampling point reflected response values at surrounding points. The resulting models were compared using Akaike's information criterion. Model selection led to a subset of 14 models with Delta AIC(c) <7 (i.e., models ranging from substantial support to considerably less support), and all but one of these included an effect of predation. Models with the autocovariate term had considerably more support than those without the term. According to multimodel inference, the presence of trout and crayfish reduced the probability of egg occurrence from a mean level of 0.90 (SE limits: 0.98-0.55) to 0.12 (SE limits: 0.34-0.04). The presence of crayfish alone had no detectable effects (SE limits: 0.86-0.39). The results suggest that introduced trout have a detrimental effect on the reproductive output of S. perspicillata and confirm the fundamental importance of distinguishing the roles of endogenous and exogenous forces that act on population distribution.

Power properties if invariant tests for spatial autocorrelation in linear regression

This paper derives some exact power properties of tests for spatial autocorrelation in the context of a linear regression model. In particular, we characterize the circumstances in which the power vanishes as the autocorrelation increases, thus extending the work of Krämer (2005). More generally, the analysis in the paper sheds new light on how the power of tests for spatial autocorrelation is affected by the matrix of regressors and by the spatial structure. We mainly focus on the problem of residual spatial autocorrelation, in which case it is appropriate to restrict attention to the class of invariant tests, but we also consider the case when the autocorrelation is due to the presence of a spatially lagged dependent variable among the regressors. A numerical study aimed at assessing the practical relevance of the theoretical results is included

Testing for spatial autocorrelation: the regressors that make the power disappear

We show that for any sample size, any size of the test, and any weights matrix outside a small class of exceptions, there exists a positive measure set of regression spaces such that the power of the Cli-Ord test vanishes as the autocorrelation increases in a spatial error model. This result extends to the tests that dene the Gaussian power envelope of all invariant tests for residual spatial autocorrelation. In most cases, the regression spaces such that the problem occurs depend on the size of the test, but there also exist regression spaces such that the power vanishes regardless of the size. A characterization of such particularly hostile regression spaces is provided.

Disease mapping and regression with count data in the presence of overdispersion and spatial autocorrelation: a Bayesian model averaging approach

This paper applies the generalised linear model for modelling geographical variation to esophageal cancer incidence data in the Caspian region of Iran. The data have a complex and hierarchical structure that makes them suitable for hierarchical analysis using Bayesian techniques, but with care required to deal with problems arising from counts of events observed in small geographical areas when overdispersion and residual spatial autocorrelation are present. These considerations lead to nine regression models derived from using three probability distributions for count data: Poisson, generalised Poisson and negative binomial, and three different autocorrelation structures. We employ the framework of Bayesian variable selection and a Gibbs sampling based technique to identify significant cancer risk factors. The framework deals with situations where the number of possible models based on different combinations of candidate explanatory variables is large enough such that calculation of posterior probabilities for all models is difficult or infeasible. The evidence from applying the modelling methodology suggests that modelling strategies based on the use of generalised Poisson and negative binomial with spatial autocorrelation work well and provide a robust basis for inference.

SPATIAL AUTOCORRELATION OF MORPHOMETRIC VARIATION IN LUTOSA-BRASILIENSIS (BRUNNER VONWATTENWYL, 1888) (ORTHOPTERA, GRYLLACRIDOIDEA, HENICIDAE)

Geographic differentiation and sexual dimorphism in eighteen morphometric characters of Lutosa brasiliensis (Orthoptera: Henicidae) collected in eight localities of the State of São Paulo (Brazil) were analysed. A two-way Multivariate Analysis of Variance (MONOVA) was used to assess simultaneously the effects of sex and geographic location (plus their interaction) on morphometric variability. The spatial patterns of variation were analysed by Factor and Spatial Autocorrelation Analyses (Moran's I coefficient in four distance classes). Both indicate that the main direction of variation is, for males and females, a north-south cline in overall body size. In females, however, ovipositor length is not correlated with overall body size and displays a different pattern of variation over geographic space, indicating that distinct evolutionary forces produced the geographic differentiation in the species.

GEOGRAPHIC-VARIATION IN APIS-CERANA-INDICA F - A SPATIAL AUTOCORRELATION ANALYSIS OF MORPHOMETRIC PATTERNS

Spatial patterns of morphometric variation in Apis cerana indica were analysed. Factor and spatial autocorrelation analyses were applied to 29 characters, measured in 17 populations in India. Correlograms showed that 15 characters are patterned geographically, and 13 of them are related to overall size. These characters are distributed as a north-south cline, probably reflecting adaptations to environmental conditions. However, the great number of characteristics without geographical pattern suggests that part of the morphometric variability is due to local stochastic divergences.

Measurement of genetic structure within populations using Moran's spatial autocorrelation statistics.

Spatial structure of genetic variation within populations, an important interacting influence on evolutionary and ecological processes, can be analyzed in detail by using spatial autocorrelation statistics. This paper characterizes the statistical properties of spatial autocorrelation statistics in this context and develops estimators of gene dispersal based on data on standing patterns of genetic variation. Large numbers of Monte Carlo simulations and a wide variety of sampling strategies are utilized. The results show that spatial autocorrelation statistics are highly predictable and informative. Thus, strong hypothesis tests for neutral theory can be formulated. Most strikingly, robust estimators of gene dispersal can be obtained with practical sample sizes. Details about optimal sampling strategies are also described.

Strong inference from transect sign surveys : combining spatial autocorrelation and misclassification occupancy models to quantify the detectability of a recovering carnivore

Funding — Forest Enterprise Scotland and the University of Aberdeen provided funding for the project. The Carnegie Trust supported the lead author, E. McHenry, in this research through the award of a tuition fees bursary.

Spatial patterns and autocorrelation in the response of microarthropods to soil pollutants: The example of oribatid mites in an abandoned mining and smelting area

Although exogenous factors such as pollutants can act on endogenous drivers (e.g. dispersion) of populations and create spatially autocorrelated distributions, most statistical techniques assume independence of error terms. As there are no studies on metal soil pollutants and microarthropods that explicitly analyse this key issue, we completed a field study of the correlation between Oribatida and metal concentrations in litter, organic matter and soil in an attempt to account for spatial patterns of both metals and mites. The 50-m wide study area had homogenous macroscopic features, steep Pb and Cu gradients and high levels of Zn and Cd. Spatial models failed to detect metal-oribatid relationships because the observed latitudinal and longitudinal gradients in oribatid assemblages were independent of the collinear gradients in the concentration of metals. It is therefore hypothesised that other spatially variable factors (e.g. fungi, reduced macrofauna) affect oribatid assemblages, which may be influenced by metals only indirectly. (C) 2009 Elsevier Ltd. All rights reserved.

Spatial analysis of notified dengue fever infections

This study aimed to investigate the spatial clustering and dynamic dispersion of dengue incidence in Queensland, Australia. We used Moran’s I statistic to assess the spatial autocorrelation of reported dengue cases. Spatial empirical Bayes smoothing estimates were used to display the spatial distribution of dengue in postal areas throughout Queensland. Local indicators of spatial association (LISA) maps and logistic regression models were used to identify spatial clusters and examine the spatio-temporal patterns of the spread of dengue. The results indicate that the spatial distribution of dengue was clustered during each of the three periods of 1993–1996, 1997–2000 and 2001–2004. The high-incidence clusters of dengue were primarily concentrated in the north of Queensland and low-incidence clusters occurred in the south-east of Queensland. The study concludes that the geographical range of notified dengue cases has significantly expanded in Queensland over recent years.