Misleading results from conventional gap analysis – Messages from the warming north
Highlights
► We compare gap analysis results to indicators of persistence and find contradicting patterns. ► Northern species have declining population trends despite of being well protected. ► Many southern species are thriving although being identified as gaps or partial gaps. ► The mismatch seems to follow expected distribution changes caused by climate change. ► Gap analysis based on static distribution data may give misleading conservation guidance.
Introduction
Gap analysis is a conservation tool designed to assess the representativeness of existing protected area networks and to identify conservation priorities (Jennings, 2000, Margules and Pressey, 2000, Scott et al., 1993). It has been used at a variety of scales (e.g. Rodrigues et al., 2004a, Sowa et al., 2007), gaining popularity in scientific studies (e.g. 220 publications since 1992), as well as in practical assessments (GAP, 2010, Langhammer et al., 2007). Gap analysis is essentially a comparison of the distributions of species (or any other feature of conservation interest) with that of protected areas, used to define the degree to which species are represented in the protected areas, and to compare the representations to prescribed targets (Margules and Pressey, 2000, Scott et al., 1993). Species can then be classified as true gap species – species not represented in any of the protected areas; partial gap species – species underrepresented in protected areas, thus not achieving the targets set for them; and covered species – species that are represented in the protected areas and that achieve their targets (e.g. Rodrigues et al., 2004a, Rodrigues et al., 2004b). The main outcome of gap analysis is to identify the true or partial gap species which need further protection. Species identified as covered are assumed to be well protected. Although this information can be used to guide the selection of new protected areas, gap analysis per se does not prescribe methods for protected area design and is not a primary tool for selecting new areas for conservation. One fundamental assumption behind gap analysis is proactive conservation, i.e. not to focus only on rare species but also on common species. This is believed to be more cost-efficient (Scott et al., 1987) and exhibit a higher probability of success (Tear et al., 1993) than trying to save species when they are on the brink of extinction.
Gap analysis has known shortcomings. Like any other conservation planning analysis, the coarseness and/or correctness of data can influence the results (Hulbert and Jetz, 2007, Rondinini et al., 2006). A further source of uncertainty can arise from mismatching resolutions of species and protected areas data (Araújo, 2004). Thus, the true representation of biodiversity features in protected areas often cannot be guaranteed (Jennings, 2000) and more detailed surveys are needed for fine-tuning the results before prioritization can be made (Jennings, 2000, Scott et al., 1993). This is closely followed by the question of what is an adequate representation level and whether it is even possible to determine one (Jennings, 2000, Rodrigues et al., 2004b). By focusing exclusively on species presence or absence in protected areas, gap analysis does not explicitly account for future threats nor does it assess the long-term persistence of biodiversity in protected areas (Cabeza and Moilanen, 2001). However there is no reason why gap analysis cannot incorporate assessments of the long-term changes in species distributions and studies have already been undertaken by looking at expected species distribution shifts under climate change (e.g. Dockerty et al., 2003, Hannah et al., 2007). Despite these shortcomings, gap analysis is widely used because it offers a simple, quantitative, and standardized method for evaluating the representativeness of protected area networks. But in a rapidly changing world, what conclusions could be drawn from gap analysis based on observed species distribution data and how should it be used when evaluating priorities for conservation?
In this paper, we evaluate existing protected areas in Finland and demonstrate how conventional gap analysis using distribution data for breeding birds results in potentially misleading conservation guidelines. We do a gap analysis to evaluate representation of Finnish breeding birds, and compare the results to two indicators of persistence: recent population trends from the past ca. 25 years and projections of distributional shifts under climate change scenarios. Population trends are a dynamic measure of the status of species, which correlate strongly with extinction risk (O’Grady et al., 2004) and can act as an indicator of conservation success (Donald et al., 2007). Future changes are more difficult to anticipate, but out of all processes that can negatively affect biodiversity, climate change will very likely take place regardless of our current actions (IPCC, 2007). Several techniques are available to project some of the impacts of climate change, especially for projecting species potential distributional shifts (e.g. Guisan and Thuiller, 2005, Heikkinen et al., 2006). We thus evaluate the potential future impacts of climate change on bird distributions, using projections from bioclimatic envelope models for 2050.
These analyses allow us to evaluate (i) what are the representation gaps in the current protected area network; and (ii) how does the representation in current protected areas correspond to short term persistence (i.e. population trends) or to (iii) long term persistence (i.e. projections).
Section snippets
Gap analysis
The bird data are based on the combined information of the first and second Finnish Bird Atlases (Hyytiä et al., 1983, Väisänen et al., 1998), which have been compiled from bird surveys done during 1974–1979 and 1986–1989. The combined atlas contains an index of breeding probability (ranging from 0 = not found; to 4 = confirmed breeding) of 248 bird species on a 10 km × 10 km uniform grid that covers nearly the entire area of Finland (totaling 3813 grid cells). The species fall into 10 groups according
Results
The gap analysis identified 20 true gap species (8% of all species analyzed) with no coverage in protected areas when using the 20% threshold to differentiate between protected and unprotected cells. A further 117–132 species (approximately 56–63%), did not achieve the assigned representation depending on the target setting scheme and were identified as partial gaps. The analysis also revealed a clear difference in the level of representation of species from different habitat types (Table 1 and
Discussion
The comparison between population trends and the conservation priorities indicated by the gap analysis reveals a contradicting pattern: in terms of representation, the north is doing better than the south, but looking at population trends, the situation is the opposite. When expanding the time scale to include projected changes in species distributions under climate change, the difference becomes even more contrasting. As the poorly represented southern species will have increasing climatic
Acknowledgements
We thank the numerous volunteers, whose valuable efforts in collecting data made this study possible; R.A. Väisänen (FMNH) and R. Heikkinen (SYKE) for providing Finnish Bird Atlas and protected area data; L. Jäättelä, I. Pozo, and E. Meyke helped with data processing; R. Virkkala, A. Hurlbert, E. Gurarie and three anonymous reviewers for valuable comments on the earlier versions of this manuscript. HK acknowledges LUOVA Graduate School for funding. WT and MBA acknowledge support from the EC FP6
References (68)
Matching species with reserves – uncertainties from using data at different resolutions
Biol. Conser.
(2004)- et al.
Ensemble forecasting of species distributions
Trends Ecol. Evol.
(2007) - et al.
Design of reserve networks and the persistence of biodiversity
Trends Ecol. Evol.
(2001) - et al.
Climate change and nature reserves: examining the potential impacts, with examples from Great Britain
Glob. Environ. Change
(2003) - et al.
What are the best correlates of predicted extinction risk?
Biol. Conserv.
(2004) Population trends of forest birds in a Finnish Lapland landscape of large habitat blocks: Consequences of stochastic environmental variation or regional habitat alteration?
Biol. Conserv.
(1991)- et al.
Uneven regional distribution of protected areas in Finland: consequences for boreal forest bird populations
Biol. Conserv.
(2007) - et al.
Projected large-scale range reductions of northern-boreal land bird species due to climate change
Biol. Conserv.
(2008) - et al.
Validation of species–climate impact models under climate change
Glob. Change Biol.
(2005) - et al.
Climate warming and the decline of amphibians and reptiles in Europe
J. Biogeo.
(2006)
Uncertainty in ensemble forecasting of species distribution
Glob. Change Biol.
Making mistakes when predicting shifts in species range in response to global warming
Nature
Beyond predictions: biodiversity conservation in a changing climate
Science
Birds are tracking climate warming, but not fast enough
Proc. Roy. Soc. B: Biol. Sci.
Partitioning and mapping uncertainties in ensembles of forecasts of species turnover under climate change
Ecography
International conservation policy delivers benefits for birds in Europe
Science
A review of methods for the assessment of prediction errors in conservation presence/absence models
Environ. Conserv.
Interspecific abundance-range size relationships: an appraisal of mechanisms
J. Anim. Ecol.
An indicator of the impact of climatic change on european bird populations
PLoS ONE
Predicting species distribution: offering more than simple habitat models
Ecol. Lett.
Conservation of mammals in eastern North American wildlife reserves: how small is too small?
Conserv. Biol.
The EBCC Atlas of European Breeding Birds, Their Distribution and Abundance
Bioclimate envelope models: what they detect and what they hide
Glob. Ecol. Biogeo.
Protected area needs in a changing climate
Front. Ecol. Environ.
Tiirojen, sotkien, naurulokin ja haahkan kannankehitys rannikoilla 1986–2006
Haahkan ja lokkien kannakehitys rannikoilla 2005–2006
Methods and uncertainties in bioclimatic envelope modelling under climate change
Progr. Phys. Geo.
The distributions of a wide range of taxonomic groups are expanding polewards
Glob. Change Biol.
Species richness, hotspots, and the scale dependence or range maps in ecology and conservation
Proc. Nat. Acad. Sci. USA
Gap analysis: concepts, methods, and recent results
Landsc. Ecol.
Cited by (37)
Assessing the effectiveness of protected areas for panda conservation under future climate and land use change scenarios
2023, Journal of Environmental ManagementThe European Union can afford greater ambition in the conservation of its threatened plants
2021, Biological ConservationCitation Excerpt :Furthermore, it will provide the opportunity to incorporate impacts such as those caused by alien species, land use or climate changes into European conservation policy, factors not foreseen at the time the Directives were approved but which are already having a deep effect on the status of many European listed species (Araújo et al., 2011). By means of modelling the future distributions of habitats and species from the EU Nature Directives, their range shifts could be anticipated and taken into account for the coverage and design of any expansion of the N2000 network in the future (Kujala et al., 2011). Based on bryophytes and vascular plants, we confirm previous results showing that N2000 could efficiently house species of interest beyond those included at present in HD annexes (Hermoso et al., 2019b).
An assessment of the efficiency and ecological representativity of existing marine reserve networks in Wales, UK
2017, Ocean and Coastal ManagementCitation Excerpt :Data were obtained by importing best results from the output of each Marxan scenario into ArcGIS, while outcomes of best and summed solutions were overlaid to illustrate the frequency, in a total of 100/1000 runs, with which each planning unit was included in the best result. A gap analysis was used to determine under-represented habitats within the existing MPA network, and that recognised the need for data on species status and trends (Kujala et al., 2011). This was extended by developing a Marxan assessed scenario in which the only areas available for selection were within the existing network, the purpose being to determine which areas and habitats were over- and/or under-represented (Stewart et al., 2003; Evans et al., 2015) (Table 4).
Using null models to identify under-represented species in protected areas: A case study using European amphibians and reptiles
2015, Biological ConservationCitation Excerpt :On the contrary, considering the spatial overlap between the entire distribution of a species within the study area and PAs, we can detect cases of species that would be identified as protected for almost any threshold used in traditional gap analysis and whose viability would be at least questionable. It could be the case of a hypothetical species with only a few (two or three) populations appearing in PU with a high percentage of PA and the remaining distribution being almost totally unprotected (e.g. the frog Rana latastei in NPAs), especially if its long term persistence in these few “protected” occurrences is not considered (Kujala et al., 2011; Sánchez-Fernández et al., 2013). Similarly, and also in contrast to traditional gap analyses, our approach does not require the choice of a subjectively-defined threshold to assign planning units to PAs (i.e. to match protected area polygons with coarse grid-cell data for species).
Population trends in boreal birds: Continuing declines in agricultural, northern, and long-distance migrant species
2013, Biological ConservationCitation Excerpt :In the temperate and boreal zones, climate change may be disproportionately threatening species with northern distributions, which appear to be in decline, while the populations of those with primarily southern distributions seem to be increasing (Virkkala and Rajasärkkä, 2011a,b). These findings concur with the observation that northern species are retracting the southern limit of their ranges at the same time as southern species appear to be expanding their distributions northwards (Brommer et al., 2012; Kujala et al., 2013, 2011; Virkkala and Rajasärkkä, 2011b). These changes may be due to differences in the thermal tolerances of species (Jiguet et al., 2010), and they are predicted to continue (Barbet-Massin et al., 2012).