Confronting collinearity: comparing methods for disentangling the effects of habitat loss and fragmentation

Estimating the relative importance of habitat loss and fragmentation is necessary to estimate the potential benefits of specific management actions and to ensure that limited conservation resources are used efficiently. However, estimating relative effects is complicated because the two processes are highly correlated. Previous studies have used a wide variety of statistical methods to separate their effects and we speculated that the published results may have been influenced by the methods used. We used simulations to determine whether, under identical conditions, the following 7 methods generate different estimates of relative importance for realistically correlated landscape predictors: residual regression, model or variable selection, averaged coefficients from all supported models, summed Akaike weights, classical variance partitioning, hierarchical variance partitioning, and a multiple regression model with no adjustments for collinearity. We found that different methods generated different rankings of the predictors and that some metrics were strongly biased. Residual regression and variance partitioning were highly biased by correlations among predictors and the bias depended on the direction of a predictor’s effect (positive vs. negative). Our results suggest that many efforts to deal with the correlation between amount and fragmentation may have done more harm than good. If confounding effects are controlled and adequate thought is given to the ecological mechanisms behind modeled predictors, then standardized partial regression coefficients are unbiased estimates of the relative importance of amount and fragmentation, even when predictors are highly correlated.     

Loss and fragmentation of mature woodland reduce the habitat niche breadth of forest birds

Context

Habitat loss and fragmentation may alter habitat occupancy patterns, for example through a reduction in regional abundance or in functional connectivity, which in turn may reduce the number of dispersers or their ability to prospect for territories. Yet, the relationship between landscape structure and habitat niche remains poorly known.

Objectives

We hypothesized that changes in landscape structure associated with habitat loss and fragmentation will reduce the habitat niche breadth of forest birds, either through a reduction in density-dependent spillover from optimal habitat or by impeding the colonization of patches.

Methods

We surveyed forest birds with point counts in eastern Ontario, Canada, and analyzed their response to loss and fragmentation of mature woodland. We selected 62 landscapes varying in both forest cover (15–45%) and its degree of fragmentation, and classified them into two categories (high versus low levels of loss and fragmentation). We determined the habitat niche breadth of 12 focal species as a function of 8 habitat structure variables for each landscape category.

Results

Habitat niche breadth was narrower in landscapes with high versus low levels of loss and fragmentation of forest cover. The relative occupancy of marginal habitat appeared to drive this relationship. Species sensitivity to mature forest cover had no apparent influence on relative niche breadth.

Conclusions

Regional abundance and, in turn, density-dependent spillover into suboptimal habitat appeared to be determinants of habitat niche breadth. For a given proportion of forest cover, fragmentation also appeared to alter habitat use, which could exacerbate its other negative effects unless functional connectivity is high enough to allow individuals to saturate optimal habitat.

     

Correlations and variance among species traits explain contrasting impacts of fragmentation and habitat loss on functional diversity

Landscape Ecology - Understanding how landscape fragmentation affects functional diversity, defined as the distribution of functional traits in an assemblage, is critical for managing landscapes...     

Metapopulation theory and habitat fragmentation: a review of holarctic breeding bird studies

Metapopulations are conceived as spatially structured populations consisting of distinct units (subpopulations), separated by space or barriers, and connected by dispersal movements. Metapopulations characteristically demonstrate a turnover of local populations going extinct and becoming re-established, resulting in a distribution pattern that shifts over time. Metapopulation theory is used to analyse the effects of habitat fragmentation on birds in the temperate zone, integrating various explanations for the paucity of species in isolated ecotopes.There is some evidence that turnover of local populations occurs in fragmented systems. A few studies based on time series demonstrate the local extinction rate to be related to the size of the habitat fragment, whereas the recolonization rate depends on the degree of isolation. Most evidence comes from short-term pattern studies in which the probability of occurrence was found to depend on the size of habitat fragments, on their relative position in the landscape and on the density of corridors lowering the landscape resistance. These data are consistent with predictions from metapopulation theory. However, almost all investigations consider wood fragmentation in agricultural landscapes, and there is a great need for studies in naturally fragmented landscapes as well as for studies focussing on other, less predictable, habitat types.     

The risk of ignoring fear: underestimating the effects of habitat loss and fragmentation on biodiversity

Landscape Ecology - Habitat loss and fragmentation threaten species not only through structural landscape changes and resource reduction, but also through modifications to species’...     

Effect of fragmentation, habitat loss and within-patch habitat characteristics on ant assemblages in semi-arid woodlands of eastern Australia

The reliability of ants as bioindicators of ecosystem condition is dependent on the consistency of their response to localised habitat characteristics, which may be modified by larger-scale effects of habitat fragmentation and loss. We assessed the relative contribution of habitat fragmentation, habitat loss and within-patch habitat characteristics in determining ant assemblages in semi-arid woodland in Queensland, Australia. Species and functional group abundance were recorded using pitfall traps across 20 woodland patches in landscapes that exhibited a range of fragmentation states. Of fragmentation measures, changes in patch area and patch edge contrast exerted the greatest influence on species assemblages, after accounting for differences in habitat loss. However, 35% of fragmentation effects on species were confounded by the effects of habitat characteristics and habitat loss. Within-patch habitat characteristics explained more than twice the amount of species variation attributable to fragmentation and four times the variation explained by habitat loss. The study indicates that within-patch habitat characteristics are the predominant drivers of ant composition. We suggest that caution should be exercised in interpreting the independent effects of habitat fragmentation and loss on ant assemblages without jointly considering localised habitat attributes and associated joint effects. The State of Queensland's right to retain a non-exclusive, royalty free license in and to any copyright is acknowledged.     

An experimental test of the habitat amount hypothesis reveals little effect of habitat area but transient or indirect effects of fragmentation on local species richness

Landscape Ecology - The habitat amount hypothesis (HAH) posits that local species richness is driven more by the amount of habitat in the surrounding landscape than by local patch size or habitat...     

Residuals cannot distinguish between ecological effects of habitat amount and fragmentation: implications for the debate

Habitat amount and fragmentation usually covary in natural and simulated landscapes. A common way of distinguishing between their effects is to take the residuals of the fragmentation index or indices regressed on habitat amount, as the index of habitat fragmentation. We used data on prairie songbird relative abundances from southern Alberta, Canada to compare this approach with the reverse: taking the residuals of habitat amount regressed on habitat fragmentation as the index of habitat amount. We used generalized additive models (GAMs) to derive residuals, and modeled relative abundances using linear mixed-effects models. The modeling approach used strongly influenced the statistical results. Using residuals as an index of fragmentation resulted in an apparently stronger effect of habitat amount relative to habitat fragmentation. In contrast, habitat fragmentation appeared more influential than habitat amount when residuals were used as an index of habitat amount. Regression of residuals may eliminate statistical collinearity, but cannot distinguish between the ecological effects of habitat amount and fragmentation. Habitat fragmentation may therefore have a larger effect on species than previously studies have shown, but experimental manipulations of underlying mechanisms are ultimately required to address this debate.     

Accessible habitat: an improved measure of the effects of habitat loss and roads on wildlife populations

Habitat loss is known to be the main cause of the current global decline in biodiversity, and roads are thought to affect the persistence of many species by restricting movement between habitat patches. However, measuring the effects of roads and habitat loss separately means that the configuration of habitat relative to roads is not considered. We present a new measure of the combined effects of roads and habitat amount: accessible habitat. We define accessible habitat as the amount of habitat that can be reached from a focal habitat patch without crossing a road, and make available a GIS tool to calculate accessible habitat. We hypothesize that accessible habitat will be the best predictor of the effects of habitat loss and roads for any species for which roads are a major barrier to movement. We conducted a case study of the utility of the accessible habitat concept using a data set of anuran species richness from 27 ponds near a motorway. We defined habitat as forest in this example. We found that accessible habitat was not only a better predictor of species richness than total habitat in the landscape or distance to the motorway, but also that by failing to consider accessible habitat we would have incorrectly concluded that there was no effect of habitat amount on species richness.     

Separating the effects of habitat amount and fragmentation on invertebrate abundance using a multi-scale framework

Context

Herbicide treatments in viticulture can generate highly contrasting mosaics of vegetated and bare vineyards, of which vegetated fields often provide better conditions for biodiversity. In southern Switzerland, where herbicides are applied at large scales, vegetated vineyards are limited in extent and isolated from one another, potentially limiting the distribution and dispersal ability of organisms.

Objectives

We tested the separate and interactive effects of habitat amount and fragmentation on invertebrate abundance using a multi-scale framework, along with additional environmental factors. We identified which variables at which scales were most important in predicting patterns of invertebrate abundance.

Methods

We used a factorial design to sample across a gradient of habitat amount (area of vegetated vineyards, measured as percentage of landscape PLAND) and fragmentation (number of vegetated patches, measured as patch density PD). Using 10 different spatial scales, we identified the factors and scales that most strongly predicted invertebrate abundance and tested potential interactions between habitat amount and fragmentation.

Results

Habitat amount (PLAND index) was most important in predicting invertebrate numbers at a field scale (50 m radius). In contrast, we found a negative effect of fragmentation (PD) at a broad scale of 450 m radius, but no interactive effect between the two.

Conclusions

The spatial scales at which habitat amount and fragmentation affect invertebrates differ, underpinning the importance of spatially explicit study designs in disentangling the effects between habitat amount and configuration. We showed that the amount of vegetated vineyards has more influence on invertebrate abundance, but that fragmentation also contributed substantially. This suggests that efforts for augmenting the area of vegetated vineyards is more beneficial for invertebrate numbers than attempts to connect them.

     

Dissimilar effects of human and elephant disturbance on woodland structure and functional bird diversity in the mopane woodlands of Zambia

Landscape Ecology - Humans and elephants are major disturbance agents in the African savanna woodlands. While both species remove trees, humans selectively harvest larger stems, which are less...     

Consequences of a large-scale fragmentation experiment for Neotropical bats: disentangling the relative importance of local and landscape-scale effects

Context

Habitat loss, fragmentation and degradation are widespread drivers of biodiversity decline. Understanding how habitat quality interacts with landscape context, and how they jointly affect species in human-modified landscapes, is of great importance for informing conservation and management.

Objectives

We used a whole-ecosystem manipulation experiment in the Brazilian Amazon to investigate the relative roles of local and landscape attributes in affecting bat assemblages at an interior-edge-matrix disturbance gradient.

Methods

We surveyed bats in 39 sites, comprising continuous forest (CF), fragments, forest edges and intervening secondary regrowth. For each site, we assessed vegetation structure (local-scale variable) and, for five focal scales, quantified habitat amount and four landscape configuration metrics.

Results

Smaller fragments, edges and regrowth sites had fewer species and higher levels of dominance than CF. Regardless of the landscape scale analysed, species richness and evenness were mostly related to the amount of forest cover. Vegetation structure and configurational metrics were important predictors of abundance, whereby the magnitude and direction of response to configurational metrics were scale-dependent. Responses were ensemble-specific with local-scale vegetation structure being more important for frugivorous than for gleaning animalivorous bats.

Conclusions

Our study indicates that scale-sensitive measures of landscape structure are needed for a more comprehensive understanding of the effects of fragmentation on tropical biota. Although forest fragments and regrowth habitats can be of conservation significance for tropical bats our results further emphasize that primary forest is of irreplaceable value, underlining that their conservation can only be achieved by the preservation of large expanses of pristine habitat.

     

Testing landscape metrics as indicators of habitat loss and fragmentation in continuous eucalypt forests (Queensland,Australia)

Landscape metrics are widely applied in landscape ecology to quantify landscape structure. However, many are poorly tested and require rigorous validation if they are to serve as reliable indicators of habitat loss and frag-mentation, such as Montreal Process Indicator 1.1e. We apply a landscape ecology theory, supported by explor-atory and confirmatory statistical techniques, to empirically test landscape metrics for reporting Montreal Process Indicator 1.1e in continuous dry eucalypt forests of sub-tropical Queensland, Australia. Target biota examined included: the Yellow-bellied Glider (Petaurus australis); the diversity of nectar and sap feeding glider species including P. australis, the Sugar Glider P. breviceps, the Squirrel Glider P. norfolcensis, and the Feathertail GliderAcrobates pygmaeus; six diurnal forest birds species; total diurnal bird species diversity; and the density of nec-tar-feeding diurnal bird species. Two scales of influence were considered: the stand-scale (2 ha), and a series of radial landscape extents (500 m –2 km;78–1250 ha) surrounding each fauna transect. For all biota, stand-scale structural and compositional attributes were found to be more influential than landscape metrics. For the Yellow-belliedGlider, the proportion of trace habitats with a residual element of old spotted-gum/ironbark eucalypt trees was a significant landscape metric at the 2 km landscape extent. This is a measure of habitat loss rather than habitat fragmentation. For the diversity of nectar and sap feeding glider species, the proportion of trace habitats with a high coefficient of variation in patch size at the 750 m extent was a significant landscape metric. None of the landscape metrics tested was important for diurnal forest birds. We conclude that no single landscape metricadequately captures the response of the regions forest biota per se. This poses a major challenge to regional reporting of Montreal Process Indicator 1.1e, fragmentation of forest types.This revised version was published online in May 2005 with corrections to the Cover Date.     

Separating the effects of habitat area,fragmentation and matrix resistance on genetic differentiation in complex landscapes

Little is known about how variation in landscape mosaics affects genetic differentiation. The goal of this paper is to quantify the relative importance of habitat area and configuration, as well as the contrast in resistance between habitat and non-habitat, on genetic differentiation. We hypothesized that habitat configuration would be more influential than habitat area in influencing genetic differentiation. Population size is positively related to habitat area, and therefore habitat area should affect genetic drift, but not gene flow. In contrast, differential rates and patterns of gene flow across a landscape should be related to habitat configuration. Using spatially explicit, individual-based simulation modeling, we found that habitat configuration had stronger relationships with genetic differentiation than did habitat area, but there was a high degree of confounding between the effects of habitat area and configuration. We evaluated the predictive ability of six widely used landscape metrics and found that patch cohesion and correlation length of habitat are among the strongest individual predictors of genetic differentiation. Correlation length, patch density and clumpy are the most parsimonious set of variables to predict the magnitude of genetic differentiation in complex landscapes.     

Effects of structural and functional habitat gaps on breeding woodland birds: working harder for less

The effects of habitat gaps on breeding success and parental daily energy expenditure (DEE) were investigated in great tits (Parus major) and blue tits (Cyanistes caeruleus) in urban parkland (Cardiff, UK) compared with birds in deciduous woodland (eastern England, UK). Tree canopy height, the percentage of gap in the canopy and the percentage of oak (in the wood only) within a 30 m radius of nest boxes were obtained from airborne remote-sensed data. Breeding success was monitored and parental DEE (great tits: both habitats; blue tits: park only) was measured using doubly labelled water in birds feeding young. In the park, mean (±SD) tree height (7.5 ± 4.7 m) was less than in the wood (10.6 ± 4.5 m), but the incidence of gaps (32.7 ± 22.6%) was greater (9.2 ± 14.7%). Great tits and blue tits both reared fewer young in the park and chick body mass was also reduced in park-reared great tits. Park great tits had a higher DEE (86.3 ± 12.3 kJ day⁻¹) than those in the wood (78.0 ± 11.7 kJ day⁻¹) and, because of smaller brood sizes, worked about 64% harder for each chick reared. Tits in the park with more than about 35% gap around their boxes had higher DEEs than the average for the habitat. In the wood, great tits with less oak around their boxes worked harder than average. Thus structural gaps, and functional gaps generated by variation in the quality of foraging habitat, increased the costs of rearing young. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Characterising the relationship between suitable habitat and gene flow for Myotis bechsteinii and Eptesicus serotinus in Britain

Landscape Ecology - Habitat suitability models (HSM) have been used to understand the impacts of landscape-scale habitat connectivity and gene flow mostly by assuming a regular decrease in the cost...     

Relative importance of habitat area and isolation for bird occurrence patterns in a naturally patchy landscape

There is debate among ecologists about whether total habitat area or patch arrangement contributes most to population and/or community responses to fragmented or patchy landscapes. We tested the relative effects of patch area and isolation for predicting bird occurrence in a naturally patchy landscape in the Bear River Mountains of Northern Utah, USA. We selected focal patches (mountain meadows) ranging in elevation from 1,920 to 2,860 m and in size from 0.6 to 182 ha. Breeding birds were sampled in each focal meadow during the summers of 2003 and 2004 using variable-distance point transects. Logistic regression and likelihood-based model selection were used to determine the relationship between likelihood of occurrence of three bird species (Brewer’s sparrow, vesper sparrow, and white-crowned sparrow) and area, isolation, and proximity metrics. We used model weights and model-averaged confidence intervals to assess the importance of each predictor variable. Plots of area versus isolation were used to evaluate complex relationships between the variables. We found that meadow area was the most important variable for explaining occurrence for two species, and that isolation was the most important for the other. We also found that the absolute distance was more appropriate for evaluating isolation responses than was the species-specific proximity metric. Our findings add clarity to the debate between ecologists regarding the relative importance of area and isolation in species responses to patchy landscapes.     

Islands in the forest: effects of patch size and isolation on farmland bird species richness and community composition of farmland patches in forest landscapes

Landscape Ecology - Theory predicts that species diversity of isolated habitat patches depends on patch size and isolation. However, there are few previous studies of how patch size and isolation...     

Models to distinguish effects of landscape patterns and human population pressures associated with species loss in Canadian national parks

It is widely accepted that large protected areas are required to effectively conserve historical species composition. However, recent analyses of mammal species loss in Canadian and African national parks contradict earlier conclusions that extent of local extinctions (i.e., extirpations) is strongly inversely related to park size, suggesting that park size alone is inadequate to predict reserve designs that may sustain biodiversity. To plan protected areas that will meet conservation goals, reserve-design models that incorporate other landscape-scale factors in addition to reserve area are needed; potential factors include the types and intensity of land use and habitat change, together with land cover types, in and around parks. Additionally, human population size around parks, and visitor density in parks may affect species loss. We quantified land use, land cover, and human population in and around 24 Canadian national parks to model effects of human disturbance and changes in natural habitats on known mammal extirpations.Multiple regression models were compared using the Akaike Information Criterion (AIC_c). The most parsimonious model (AIC_c weighting w _i = 0.5391) emphasized effective habitat area in and around parks and not visitor numbers nor human population size around parks. Our model suggests that parks with as little as 3140 km² of effective habitat area inside may be large enough to conserve historical mammal species composition if they are also surrounded by at least 18 000 km² of effective habitat within 50 km of park boundaries.