Implementing comprehensiveness, adequacy and representativeness criteria (CAR) to indicate gaps in an existing reserve system: a case study from Victoria, Australia

Ecological Indicators 18 (2012) 342–352

Contents lists available at SciVerse ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Implementing comprehensiveness, adequacy and representativeness criteria (CAR) to indicate gaps in an existing reserve system: A case study from Victoria, Australia Seyedeh Mahdieh Sharafi a,∗ , Matt White b , Mark Burgman a a b

The Australian Centre of Excellence for Risk Analysis, The School of Botany, University of Melbourne, Victoria 3010, Australia The Arthur Rylah Institute for Environmental Research, Department of Sustainability and Environment, PO Box 137, Heidelberg 3084, Australia

a r t i c l e

i n f o

Article history: Received 6 April 2011 Received in revised form 18 November 2011 Accepted 21 November 2011 Keywords: Gap analysis Indicator Comprehensiveness Adequacy Representativeness Environmental space Systematic conservation planning

a b s t r a c t Land clearance and disturbance in human-dominated landscapes have significant impacts on biodiversity internationally. Conservation planning can play a role in mitigating the effects of human-related activities. One element of conservation planning involves the analysis of the spatial arrangement of areas in a landscape and identifying characteristics that are underrepresented in protected areas. In this study, the distribution of protected areas in Victoria, Australia is assessed in relation to environmental space. We use comprehensiveness, adequacy and representativeness (CAR) criteria as a framework to examine the efficiency of existing protected areas and to define gaps in the current reserve system, based on multivariate environmental space. Our method is based on the conversion of combinations of environmental variables to biodiversity features. The analysis provides a systematic, quantitative gap analysis. Our framework provides feasible tools for involving CAR principles in evaluating the efficiency of reserve areas. Our metrics are transparent, simple to interpret and easy to implement. It is also applicable as a decision support system in land use and conservation planning for analysing future land development scenarios. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction A key focus of conservation planning is the design of reserve systems that sample and protect biodiversity surrogates in a cost effective manner (C-Plan (NSW NPWS, 1999); Marxan (Ball and Possingham, 2000); Zonation (Moilanen, 2007)). However across much of the world, reserve systems are already in place. These reserve systems are rarely the consequence of reserve selection strategies and/or strictly rational criteria. Typically reserve systems emerge from a protracted sequence of decisions strongly influenced by land-use history and contemporaneous political, social and economic motivations. Typically, inexpensive, inaccessible or otherwise un-arable lands comprise a high proportion of reserved lands, resulting in unrepresentative networks of protected areas (Mark, 1985; Pressey, 1994). Analysing gaps in networks of protected areas is a high priority for planning, providing a measure of protection against trends of landscape changes, habitat loss, and the effects of climate change (Fairbanks and Benn, 2000; Mackey et al., 1988;

∗ Corresponding author. Tel.: +61 03 83443305; fax: +61 03 93481620. E-mail addresses: s.sharafi@pgrad.unimelb.edu.au (S.M. Sharafi), [email protected] (M. White), [email protected] (M. Burgman). 1470-160X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecolind.2011.11.023

Oldfield et al., 2004; Rodríguez et al., 2007). Gap analysis is a tool for assessing the distribution of protected areas in a landscape with the aim of identifying the places that are potentially useful additions to the reserve network. Prioritizing areas for conservation can be depicted by mapping under-represented areas. Gap analysis is considered to be an efficient, proactive conservation tool for assessing protection status and setting priorities for land management (Burley, 1988; Maxted and Dulloo, 2008; Rodríguez et al., 2007; Scott, 1993). Gap analysis aims to analyse the performance of a reserve system with respect to one or more objectives such as, to compensate for the impacts of habitat loss, maintain long-term habitat viability, reduce the number of endangered and threatened species (Angelstam et al., 2003; Pressey et al., 2002; Scott, 1993), provide basic cartographic information at different scales for decision support systems (Pearlstine et al., 2002; Scott, 1993) or achieve quantitative conservation targets and integrated conservation plans (Kirkpatrick, 1983; Margules and Pressey, 2000; Pressey et al., 2002; Scott, 1993). Many conservation planning tools use information on individual species to set priorities (Sarkar et al., 2006). However, most taxa are currently unknown or poorly known. For instance, of about 5–10 million extant species, about 1.4 million have been described (Lindenmayer and Burgman, 2005; Myers et al., 2000; Pimm et al.,

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

343

Fig. 1. The study area is located in the State of Victoria, Australia.

1995). There is very little information about the distribution and abundance of the vast majority of species. There is an urgent need to further develop indicators that serve to protect the majority of undiscovered and poorly known taxa. Reserve selection procedures and assessment of biodiversity values often involved the use of surrogates where species distribution data are inadequate (Martinez et al., 2006; Wessels et al., 1999). In other studies, a broad range of indicators have been used as surrogate measures of biodiversity; they include land facet (the simplest unit of a landscape with uniform slope, soil and hydrological conditions), climatic variables, landform vegetation classes, land systems, physicochemical variables and environmental variables (Faith and Walker, 1996; Faith, 2003; Margules et al., 1988; Margules and Pressey, 2000; Martinez et al., 2006; Sarkar and Margules, 2002; Snelder et al., 2007; Wessels et al., 1999). It is important that the selection of surrogate measures is linked directly to the distribution and abundance of the species and systems we intend to conserve (Austin et al., 1994). Our approach, outlined below, is to choose a broad spectrum of environmental parameters known to affect the distribution and abundance of many different species. The variables employed in a gap analysis act as broad surrogates for biodiversity at both the species and ecosystem level (Arponen et al., 2008; Faith et al., 2004; Faith and Walker, 1996; Faith, 2003, 2010; Ferrier et al., 2004; Martinez et al., 2006; Pyke and Fischer, 2005; Snelder et al., 2007). Faith and Walker’s environmental diversity approach uses pattern representation of the areas in a region. It assumes species abundances are concentrated around unique sets of environmental characteristics so that larger distances in environmental space indicate greater differences in species composition (Belbin, 1993; Faith and Walker, 1996; Faith et al., 2001; Ferrier, 2002; Kirkpatrick and Brown, 1994; Trakhtenbrot and Kadmon, 2005; Woinarski et al., 1996). There are many other approaches to the generalised problem of reserve design. The intention here is to outline a method that is effective in the absence of detailed data or where there is a need to compare data rich and data poor regions. Within this context of aiming to conserve undiscovered and poorly known taxa, our strategy is to satisfy socially mandated objectives specified in government regulations. We have framed our reserve system metrics within the context of concepts that have been developed as

policy objectives at the national level (Commonwealth of Australia, 1999; Natural Resource Management Ministerial Council, 2004). A conservation reserve system should be ‘comprehensive’, ‘adequate’ and ‘representative’ (CAR). These concepts have emerged from negotiations between State and Federal Governments in Australia with respect to the reservation of land for biodiversity conservation. The Natural Resource Management Ministerial Council defines these concepts as follows: 1. Comprehensive: the inclusion in the National Reserve System of examples of regional-scale ecosystems in each bioregion. 2. Adequate: the inclusion of sufficient levels of each ecosystem within the protected area network to provide ecological viability and to maintain the integrity of populations, species and communities. 3. Representative: the inclusion of areas at a finer scale, to encompass the variability of habitat within ecosystems. The CAR approach has been applied to both terrestrial and aquatic ecosystem conservation planning (Fitzsimons and Robertson, 2005; Pressey et al., 2002). We intend to assess how the reserve system of the State of Victoria, Australia, has performed in capturing the State’s environmental space, as defined by a suite of continuous and discrete biophysical variables and to identify candidate areas for inclusion in future planning activities. The state of Victoria, Australia (Fig. 1) covering an area of 22.7 million hectares, has an established and legally defined conservation reserve system within which, most resource exploitation has been excluded. State jurisdictions in Australia have the central responsibility for land use and development and each state has its own conservation and nature reserve management agencies. Victoria is the most densely settled of all Australian States and is the most highly modified in terms of intensive human land uses such as urban areas, cereal cropping, and grazing on improved pastures. Human settlement patterns within Victoria are a largely a consequence of the extent of arable land and government schemes to encourage and promote ‘close settlement’ and farming. With very few exceptions, the Victorian reserve system circa 2011 has been created on Government owned lands that were never alienated or

344

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Fig. 2. Victoria’s Nature Conservation Reserve System. Green regions denote reserves and grey areas circumscribe the State of Victoria, Australia.

that reverted to public ownership after they were abandoned by early unsuccessful European settlement. Prior to reservation, Government lands were used for a variety of often co-incident purposes, including rangeland grazing, mining, timber, water production, fuelwood harvesting and hunting. Further, extensive tracts of land were Government owned and controlled merely because there were no other interests – the land being either too poor for agriculture or animal husbandry or too remote from services or reliable water. The parks and nature reserve system has evolved over many years as the consequence of myriad formal and informal planning processes. Freehold land purchased for inclusion in the reserve system is a recent phenomenon and is negligible in terms of land area. The total area dedicated to nature conservation is approximately 17% of total area of the state (3.86 million hectares). The reserve system is shown in Fig. 2. 2. Methods Gap analysis aims to identify areas that are under-represented in a reserve network. By using multivariate environmental space as a surrogate for biodiversity (Arponen et al., 2008; Faith and Walker, 1996) potential conservation areas may be selected to maximise environmental diversity. The following flow diagram (Fig. 3) depicts the stages in the gap analysis. In order to analyse the status of a protected area, it is first necessary to identify which elements of environmental features are missing in the network of conservation areas. Gap analysis based on environmental variables depends on the selection of appropriate variables. 2.1. Developing ecosystem surrogates – delimiting the environmental space The computational intensity of this task necessitated significant data reduction. We reduced the complexity of the environmental space to a series of categorical landscape properties embodied in a suite of spatially explicit environmental variables that were assumed a priori to be useful in describing the terrestrial environmental space (see below). We assumed that the interplay of

climatic, edaphic and physiographic variables broadly captures the terrestrial ecosystem variation across the State. A wide range of environmental variables is employed for species modelling, including climatic, physical and biophysical data (Fairbanks and Benn, 2000; Estrada et al., 2008; Mackey et al., 2008; Austin, 1992, 2007; Barbosa et al., 2009; Franklin, 1998; Mackey et al., 1988; Newell et al., 2009). The relative importance and utility of variables is likely to vary between species, between regions and at different spatial scales (Peterson and Nakazawa, 2008). A minimum set of predictors used in species distribution modelling typically include climatic, physiographic and edaphic variables (Mackey and Lindenmayer, 2001). The interplay of such factors broadly reflect gross primary production. In this study, we selected from a large set of potential environmental variables with a view to describing the physical environment of our study area efficiently. We included annual net rainfall and annual temperature as our climate descriptors. We selected direct solar radiation, topographic roughness and terrain wetness index to describe physiographic context and also to confer some hydrological and geomorphological intelligence to our analysis. Finally geology and radiometric data (thorium: potassium ratio), were selected to reflect the general nature of the regolith and soil texture respectively. Previous studies have shown that Gamma ray data can provide information on soil properties including texture in depositional environments (including aeolian, colluvial and fluvial-lacustine deposits). In general, K and Th concentrations in surficial materials are associated with clays and sands are characterised by a low radioelement response (Cattle et al., 2003; Cook et al., 1996; Wilford et al., 1997). We used continuous Th:K ratio data in our analysis to represent soil texture. The details of the layers are provided in Table 1 and Appendices A and B.

2.2. GIS database 2.2.1. Data preparation We prepared GIS data to represent the selected environmental variables. Bioclimatic variables were derived from interpolated

Table 1 A summary of the data sets used in the analysis and subsequent geo-processing. Origin

Data format

Initial data resolution

Re-sampling method (re-sampled from original resolution to 500 m)

Units

Discretisation method

Terrain wetness index

Topographic wetness index – a compound terrain attribute (sensu Beven and Kirkby, 1979) implemented using the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model and TOPOCROP Version 2.1 (Schmidt, 2002) Vector ruggedness measure – a terrian heterogeneity measure (sensu Hobson, 1972) implemented using the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model and VRM script (see Sappington et al., 2007) Mean annual rainfall (mm) surface from ANUCLIM (Houlder et al., 2000) model less Mean Annual Evaporation (mm) model from ANUCLIM (Houlder et al., 2000) Mean annual temperature (◦ C). Developed using ANUCLIM (Houlder et al., 2000) Diffuse solar radiation developed. Derived from Shuttle Radar Topography Mission (SRTM) Digital Elevation Model using The Solar Analyst 1.0 (Fu and Rich, 2000) Radiometric data – ratio of the inverse radioelement count of thorium and the radioelement count of potassium. Sourced from Victorian Government, Department of Primary Industries Surface geology of Australia 1:1,000,000 scale, Victoria – 3rd edition. Source Geoscience Australia

Continuous grid (raster)

100 m

Bilinear

None

6 classes – derived from standard deviation

Continuous grid (raster)

100 m

Bilinear

None

19 classes – equal interval

Continuous grid (raster)

100 m

Bilinear

Millimetres

20 classes – equal interval

Continuous grid (raster)

100 m

Bilinear

Degrees centigrade

20 classes – equal interval

Continuous grid (raster)

100 m

Bilinear

w/m2 /yr

27 classes – equal interval

Continuous grid (raster)

50 m

Bilinear

Count

19 classes – quantile breaks

Discrete – Polygon (vector)

Largely from 1:250,000 scale mapping

Nearest neighbour

N/A

16 classes – expert reclassification

Vector ruggedness measure

Annual net rainfall

Annual temperature Radiation

Radiometric

Geology

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Environmental variable

345

346

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Delimit the environmental space

Develop ecosystem surrogates by choosing main environmental variables that determine the distribuon and abundance of important species

Prepare GIS data base

Create muldimensional space for the study area

Plot pixels (500 meter by 500 meter cells) as points, in spaces defined by each pair of environmental variables

Overlay the park and reserve map (Tenure map) on the plot of points in environmental space

Look for areas of environmental space that are not covered by parks and reserves according to comprehensiveness, adequacy and representaveness (CAR) criteria Fig. 3. Gap analysis procedure employed in this study.

surfaces estimated using the software package ANUCLIM 5.1 (Houlder et al., 2000) which uses thin plate smoothing splines fitted to long-term meteorological station data. All the spatial data were re-sampled from the native resolution to a standard 500 m × 500 m grid resolution (Table 1). This grid size was used effectively for systematic conservation planning in Maputaland, South Africa (Smith et al., 2006), an area with similar topography and ecology. The 500 m × 500 m grid was produced using the ET VectorGrid extension in ArcGIS 9.3 (ET GeoWizards, 2005). It should be noted that at this scale the loss of data is acute for categorical data. In particular, small reserves below the resolution of the data are eliminated from subsequent analysis.

2.2.2. Data reduction – variable discretisation To generate measures of the degree to which objects we want in our reserve system are represented within it, we discretised continuous environmental variables into classes (see Table 1). The variation of classes between environmental variables in discretisation is due to the amount of data on each layer. We used the pair wise combination of environmental variables because in 3 way interactions and higher order interactions every pixel has a unique characteristic which does not provide useful information for the purpose of this analysis. The variables were then combined pairwise (class by class) to generate a set of 5508 unique class pairs, many of which overlap in geographic space (see Table 2). Note that not all potential combinations between the variables are realised in the environmental space. Spatially explicit pair-wise combinations were generated in the GIS using the “combine” function in ArcGIS (combines multiple rasters so a unique output value is assigned to each unique combination of input values). How many continuous environmental variables should we choose and how many classes should we identify within them for the purpose of measuring the performance of a reserve system? A useful number of variables and classes should effectively and

efficiently characterise the environmental space. As we increase the number of classes within each variable and subsequently combine those variables we rapidly increase the number of classes. As all places are unique in environmental space and all loci are unique in geographic space, our metric needs to be somewhere between describing every place as a unique class (in such an analysis approximately 17% of the types are 100% sampled by the reserve system in Victoria and 83% of the classes are wholly unrepresented in the reserve system) and the entire study area as a single class (a summary at this scale suggests that 17% of Victoria is represented in the reserve system). As we approach either extreme, any metric will be increasingly less informative. Choosing between the two extremes of class heterogeneity and scale heterogeneity is somewhat arbitrary. As with all classifications and clustering, the real test of the derived classes is whether they are fit for their purpose, in this case, supplying some measure of the reach of the reserve system within the environmental space. 2.2.3. Aggregation of existing categorical data One of the variables used in the analysis was already a broad multivariate class – geology. This comprised many (some 8393) discrete classes based on lithology, age, and other descriptors. These geology classes were simplified while retaining, where possible, salient attributes relevant to ecosystem productivity and process. The simplified categories of geology are listed in Appendix A. 2.3. Definitions of CAR criteria In reserve design, comprehensive refers to the inclusion within protected areas of samples of each of the ecosystems discernible at the bioregional scale (Natural Resource Management Ministerial Council, 2004). In this study, comprehensiveness is interpreted as the inclusion within protected areas of each of the combinations of environmental variables. Hence, comprehensiveness analysis involves identifying which elements of environmental features are

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

347

Table 2 Pair wise combination of environmental variables. The numbers show the unique combinations. Pair wise combination

Terrain wetness index

Vector ruggedness measure

Annual net rainfall

Annual temperature

Solar radiation

Radiometric data

Geology

Terrain wetness index Vector ruggedness measure Annual net rainfall Annual temperature Solar radiation Radiometric data Geology

– – – – – – –

66 – – – – – –

89 254 – – – – –

92 266 153 – – – –

119 390 440 468 – – –

116 330 377 387 572 – –

82 220 177 181 378 351 –

missing from the network of protected areas. The following formula shows the equation for quantifying comprehensiveness in our reserve system; Ci =

N 

wij

0 ≤ Ci ≤ 21,

(1) Ri =

j−1

and wj = 0,1 where wj = 1 implies this pixel is not in the reserve but this combination is protected somewhere and wj = 0, implies not protected anywhere, N = total number of pair wise combinations (=21), and i = Pixel Index (1 < i < 908,344). Adequacy refers to how much of each ecosystem should be included within a protected area network to provide ecological viability and integrity of populations, species and communities (Natural Resource Management Ministerial Council, 2004). In this study, adequacy is measured by the proportion of each unique class (combination of environmental variables) that is represented in the reserve system (see below). The following formula quantifies the adequacy of areas in our reserve system; Ai =

N  

n n

j=1

n



q i=1 ij

q i=1 ij

+

manipulated such that they were at least 100,000 hectares but not greater than 3,000,000 hectares in size. The formula for representativeness is the same as comprehensiveness, computed at a different scale. In our reserve system,

n

,

r i=1 ij

(2)

n

q + r is the number of pixels in which the comwhere i=1 ij i=1 ij bination j occurs, i = Pixel Index (1 < i < 908,344), N = total number of pair wise combinations (=21), n = total number of pixels (=908,344), vj = 1, 0, where vj = 1, implies protected here and vj = 0 implies not protected here but protected somewhere, q = 1 if vj = 1, q = 0, otherwise, and r = 1 if vj = 0, r = 0, otherwise. Representativeness is comprehensiveness considered at a finer scale, and infers that the variability within ecosystems is sampled within the reserve system (Natural Resource Management Ministerial Council, 2004). Representativeness, like comprehensiveness, was measured by the presence or absence of environmental classes in the reserve system. The difference was that the representativeness was assessed within sub-regions of the geographic space, to account for local variation within ecosystem types and species turnover. We divided Victoria into 25 major river basins or watersheds. Surface flow watersheds were derived from the SRTM DEM using the HydroSHEDS Arcview extension (Lehner et al., 2006). To ensure that the study area was subdivided into regions of similar geographic extent, catchment size was

Q 

Cki

(3)

k=1

where Q = the total number of major river basins (=35), each discretised variable was combined pair-wise to create additional variables in the environmental space (Table 3). Consequently each 500 m grid cell supports a compliment of 21 classes (from (6 × 7)/2 combinations of variable pairs). Each unique class resulting from the pair wise combination of the 7 environmental variables was interrogated in the GIS system to determine: • Whether the unique class is represented in the reserve system or not (reflecting comprehensiveness, as defined above). This is reported to the relevant cells. To translate this multidimensional view into a 2D view (or map) we have determined those places (or grid cells) where one or more unique combinations of variables are not represented in the reserve system. • The area extent of each unique class across Victoria (this is the count of cells within each class). • The proportion of each class that is in the reserve system. This equates to the ratio of the number of cells within each unique environmental class that is within the reserve system with the number of cells within the same class that is currently outside the reserve system. • As the pair-wise classes overlap in the geographic space, each 500 m grid cell supports a combination of 21 classes. To build a visual summary of the degree of reservation of each class at every 500 m × 500 m locus, we assessed the proportion of each of the co-occurring 21 biophysical classes that is reserved. Of these 21 proportions, the minimum % reserved is reported for the pixel – this is our principle adequacy metric. • Whether each unique class is represented in the reserve system, assessed within the relevant cells to determine representativeness. To translate this multidimensional view into a 2D view we defined representativeness to mean those places (or grid cells) where one or more unique variables are not represented in the reserve system.

Table 3 Unique combinations of environmental variables within watersheds. Combinations

Terrain wetness index

Vector ruggedness measure

Annual net rainfall

Annual temperature

Solar radiation

Radiometric data

Geology

Terrain wetness index Vector ruggedness measure Annual net rainfall Annual temperature Solar radiation Radiometric data Geology

– – – – – – –

1202 – – – – – –

823 2504 – – – – –

956 3062 1034 – – – –

2306 7510 4287 5232 – – –

1960 4620 2833 3330 8298 – –

937 2006 1057 1207 3754 3150 –

348

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Fig. 4. Comprehensiveness of Victoria’s reserve system which is evaluated in environmental space and mapped in geographical space. Grey denotes regions where the interplay of environmental classes is represented in the reserve system. Blue denote regions where the interplay of environmental classes is unrepresented in the reserve system.

3. Result

3.3. Representativeness

3.1. Comprehensiveness

Each catchment was combined with the environmental classes to create 632,068 additional but strongly geographically delimited new discrete classes (Fig. 7). The representativeness analysis focuses on the gaps in the reserve system within the geographic space. At the catchment scale substantial regions of the environmental space are unrepresented.

Fig. 4 shows the status of comprehensiveness of Victoria’s reserve system in environmental space. Grey regions in the comprehensiveness map are those where all the biophysical classes that occur at that location (each 500 m2 pixel) have some representation in the reserve system. Black regions are those where there are biophysical classes that are not represented in the reserve system. Using our comprehensiveness metric, we discovered 94,000 ha across Victoria that coincides with regions of environmental space that are not sampled by the reserve system. Of this area, some 79,450 ha is on freehold land and the remainder is publically owned land. The analysis directs our attention to highly restricted and/or outlying occurrences of restricted geologies that have limited or no representation in the reserve system in the context of their cooccurrence with other variables. Of particular note is the general absence of intrusive mafic volcanics, lime stones and lateritic duricrust surface geologies in the reserve system. Also of note is the localised lack of representation of basaltic landscapes in upland and or low rainfall contexts.

3.2. Adequacy Adequacy was analysed using the two measures outlined above. First, the percentage of the least well reserved biophysical class in the reserve system in each 500 m cell was identified (Fig. 5). Then the percentage of the average class in each 500 m cell was identified (Fig. 6). We compare the adequacy results with land use history in Victoria. The results highlight the degree to which agriculture and to a lesser extent forestry and urban land uses compete with nature conservation. In general fertile lowlands in moderate to high rainfall zones typically have low adequacy scores, whereas uplands on widespread geologies and siliceous dune fields have higher adequacy scores.

4. Discussion The results of studies worldwide have shown that there are significant gaps in current reserve systems (Andelman and Willig, 2003; Estrada et al., 2008; Jennings, 1995, 2000; Kirkpatrick, 1983; Lacher, 1998; Rodrigues et al., 2004; Rodríguez et al., 2007; Scott et al., 2001; Strittholt and Boerner, 1995). More importantly, in some cases the protected areas are above the proposed conservation targets for setting aside a specific percentage of land for conservation (Oldfield et al., 2004; Tognelli et al., 2008). Typically, some places such as highlands are well represented when compared with low altitude and coastal areas (Oldfield et al., 2004; Tognelli et al., 2008). Distributing national targets across a country to maintain the viability of biodiversity and protect species efficiently is a serious challenge (Fitzsimons and Robertson, 2005; Oldfield et al., 2004; Tognelli et al., 2008). Setting aside a certain percentage of land in protected areas, without considering the efficiency of that area to protect species, will result in biodiversity loss and an increasing number of species becoming threatened and vulnerable (Tognelli et al., 2008). There is no clear answer to the question of how much land is enough for protecting biodiversity (Tear et al., 2005). Our analysis can help to improve the efficiency of reserve areas in achieving long term protection of biodiversity by considering comprehensiveness, adequateness and representativeness of the protected areas, no matter what extent or percentage of the total area is protected. Other gap analyses of this nature use discrete classes of spatially explicit objects such as mapped vegetation types (Scott,

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

349

Fig. 5. Map of adequacy based on the least well reserved class, evaluated in environmental space and mapped in geographical space.

1993; Strittholt and Boerner, 1995). In Victoria, vegetation has been mapped across the state. However, the types have been defined and mapped by many different individuals differentially applying and weighting various attributes and deploying subtly different approaches and biases (Keith and Bedward, 1999). Further to this, the resolution of mapping units varies markedly between tranches of mapping effort and between land tenures. The analysis in this study is repeatable and uses a uniform scale and definitions across tenures and regions. Our innovation in this gap analysis approaches, is to ensure that the results of the analysis are bound to regional

needs as specified in government regulations and its objectives to preserve land for biodiversity conservation. The comprehensiveness analysis highlights regions of deficiency in the Victorian reserve system. At 500 m resolution, the terrestrial reserve system does not sample the entire range of fertile lowlands with restricted geologies. Of particular note are lime stones, basalts, and laterised paleo-land-surfaces in the west of the state. The adequacy results draw attention to the competition between intensive human land use and the reserve system (Figs. 5 and 6). We also compare the result of CAR criteria with

Fig. 6. Adequacy based on the average of the reservation status of the environmental classes which evaluated in environmental space and mapped in geographical space.

350

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Fig. 7. Representativeness of the reserve system which is evaluated in environmental space and mapped in geographical space. Grey regions are those where all the biophysical classes that occur at that location (i.e. each 500 m2 pixel) have at least some representation in the reserve system. Blue regions are those locations where there are biophysical classes that are not represented in the reserve system.

the tenure map of Victoria, The comparison analysis highlight that these absolute deficiencies are largely on freehold land tenures. It is worth mentioning that some arbitrary decisions have been made in running the current analysis. They include the selection of environmental variables, the number of classes within each variable, scale of the analysis (500 m pixels versus coarser or finer scale) and breaking up the environmental space for the representativeness analysis (the watershed approach). Sensitivity analysis can be used to evaluate the importance of these decisions in determining outcomes, prior to finalising decisions about future investments in reserve systems. We conducted several such analyses (not shown here), exploring the effect of scale aggregation of environmental variables, and so on; none made a substantial difference to the results reported here. This approach provides a new frame-work for analysing gaps in the network of reserve areas. The result of this analysis will be valuable for prioritizing the design of future reserve areas and allow areas with low Adequate scores to be considered as first priority for future conservation actions. The proposed framework at the landscape scale could be used to guide conservation priorities and to enhance the landscape’s capacity to meet changes such as land use or climate change.

Table A1 The simplified categories of geology layer. Relict Alluvial Plains Colluvial and sub-lithic deposits Active Alluvial and Lacustrine depositional environments Basalt (Igneous extrusives) Calcareous dune fields Evaporite deposits Felsic Igneous rocks (i.e. granites and granodiorites) Relict fluvial braid plains Gnessic and schistic rocks Laterites and duricrusts Limestones Lunette dunes Intrusive mafic rocks Rhyolites Sandstones, mudstones, siltstones Siliceous dune fields

agreements prohibit us passing them onto other researchers in the absence of a similar such agreement. However, we would encourage those interested in obtaining the data to contact the authors. We also thank two anonymous reviewers for their valuable comments and suggestions.

Acknowledgments Appendix A. Geology data layer used in this work We thank the Victorian Department of Sustainability and Environment who supplied the original spatial data. Licensing

See Table A1.

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

351

Appendix B. Maps of classified environmental variables that have been used in this analysis

Scaling and precision: the values in ANUCLIM are multiplied by an appropriate power of 10 then rounded to the nearest integer to preserve the indicated precision. I.e. 1 decimal places requires scaling by 10, 2 decimal places requires scaling by 100 (Houlder et al., 2000). Parameter

Multiplier ◦

Temperature ( C) Rainfall (mm)

10 1

References Andelman, S.J., Willig, M.R., 2003. Present patterns and future prospects for biodiversity in the Western Hemisphere. Ecology Letters 6, 818–824. Angelstam, P., Mikusi ski, G., Rönnbäck, B.I., Östman, A., Lazdinis, M., Roberge, J.M., Arnberg, W., Olsson, J., 2003. Two-dimensional gap analysis: a tool for efficient conservation planning and biodiversity policy implementation. Ambio 32, 527–534. Arponen, A., Moilanen, A., Ferrier, S., 2008. A successful community-level strategy for conservation prioritization. Journal of Applied Ecology 45, 1436–1445. Austin, M., 1992. Modelling the environmental niche of plants: implications for plant community response to elevated CO2 levels. Australian Journal of Botany 40, 615–630. Austin, M., 2007. Species distribution models and ecological theory: a critical assessment and some possible new approaches. Ecological Modelling 200, 1–19. Austin, M., Nicholls, A., Doherty, M., Meyers, J., 1994. Determining species response functions to an environmental gradient by means of a function. Journal of Vegetation Science, 215–228. Ball, I., Possingham, H., 2000. Marxan (v 1.8.6): Marine Reserve Design Using Spatially Explicit Annealing. User Manual. The University of Queensland, Australia. Barbosa, A.M., Real, R., Mario Vargas, J., 2009. Transferability of environmental favourability models in geographic space: the case of the Iberian desman (Galemys pyrenaicus) in Portugal and Spain. Ecological Modelling 220, 747–754. Belbin, L., 1993. Environmental representativeness: regional partitioning and reserve selection. Biological Conservation 66, 223–230. Beven, K.J., Kirkby, M.J., 1979. A physically based variable contributing area model of basin hydrology. Hydrological Sciences Bulletin 24, 43–69. Burley, F.W., 1988. Monitoring biological diversity for setting priorities in conservation. In: Wilson, E.O. (Ed.), Biodiversity. National Academy Press, Washington, DC, USA, pp. 227–230. Cattle, S.R., Meakin, S.N., Ruszkowski, P., Cameron, R.G., 2003. Using radiometric data to identify aeoilian dust additions to topsoil of the Hillston district, western NSW. Australian Journal Soil Research 41, 1439–1456.

Commonwealth of Australia, 1999. Australian Guidelines for Establishing the National Reserve System. Environment Australia, Canberra. Cook, S.E., Corner, R.J., Groves, P.R., Grealish, G.J., 1996. Use of airborne gamma radiometric data for soil mapping. Australian Journal Soil Research 34, 183–194. Estrada, A., Real, R., Vargas, J.M., 2008. Using crisp and fuzzy modelling to identify favourability hotspots useful to perform gap analysis. Biodiversity and Conservation 17, 857–871. ET GeoWizards, 2005. ArcGIS 9.3 Extension. Available at http://www.ian-ko.com. Fairbanks, D.H.K., Benn, G.A., 2000. Identifying Regional Landscapes for Conservation Planning: A Case Study from KwaZulu-Natal, South Africa. Faith, D., Ferrier, S., Walker, P., 2004. The ED strategy: how species-level surrogates indicate general biodiversity patterns through an ‘environmental diversity’perspective. Journal of Biogeography 31, 1207–1217. Faith, D., Walker, P., 1996. Environmental diversity: on the best-possible use of surrogate data for assessing the relative biodiversity of sets of areas. Biodiversity and Conservation 5, 399–415. Faith, D.P., 2003. Environmental diversity (ED) as surrogate information for specieslevel biodiversity. Ecography 26, 374–379. Faith, D.P., 2010. Environmental Diversity (ED) As Surrogate Information for Specieslevel Biodiversity. The Australian Museum, Sydney, Australia. Faith, D.P., Margules, C., Walker, P., Stein, J., Natera, G., 2001. Practical application of biodiversity surrogates and percentage targets for conservation in Papua New Guinea. Conservation Biology 6, 289–303. Ferrier, S., 2002. Mapping spatial pattern in biodiversity for regional conservation planning: where to from here? Systematic Biology 51, 331. Ferrier, S., Powell, G.V.N., Richardson, K.S., Manion, G., Overton, J.M., Allnutt, T.F., Cameron, S.E., Mantle, K., Burgess, N.D., Faith, D.P., 2004. Mapping more of terrestrial biodiversity for global conservation assessment. BioScience 54, 1101–1109. Fitzsimons, J.A., Robertson, H.A., 2005. Freshwater reserves in Australia: directions and challenges for the development of a comprehensive, adequate and representative system of protected areas. Hydrobiologia 552, 87–97. Franklin, J., 1998. Predicting the distribution of shrub species in southern California from climate and terrain-derived variables. Journal of Vegetation Science 9, 733–748. Fu, P., Rich, P.M., 2000 The Solar Analyst. Helios Environmental Modeling Institute. User Manual. USA. Hobson, R.D., 1972. Surface roughness in topography: quantitative approach. In: Chorley, R.J. (Ed.), Spatial Analysis in Geomorphology. Harper and Row, New York, USA, pp. 221–245. Houlder, D., Hutchinson, M., Nix, H., MacMahon, J., 2000. ANUCLIM 5.1 Users Guide. The Australian National University. Center for Resource and Environmental Studies, Canberra. Jennings, M.D., 1995. Gap analysis today: a confluence of biology, ecology, and geography for management of biological resources. Wildlife Society Bulletin 23, 658–662.

352

S.M. Sharafi et al. / Ecological Indicators 18 (2012) 342–352

Jennings, M.D., 2000. Gap analysis: concepts, methods, and recent results. Landscape Ecology 15, 5–20. Keith, D.A., Bedward, M., 1999. Native vegetation of the South East Forests region, Eden New South Wales. Cunninghamia 6, 1–218. Kirkpatrick, J., 1983. An iterative method for establishing priorities for the selection of nature reserves: an example from Tasmania. Biological Conservation 25, 127–134. Kirkpatrick, J., Brown, M., 1994. A comparison of direct and environmental domain approaches to planning reservation of forest higher plant communities and species in Tasmania. Conservation Biology 8, 217–224. Lacher, T.E., 1998. In: Savitsky, B.G., Lacher, T.E. (Eds.), GIS Methodologies for Developing Conservation Strategies: Tropical Forest Recovery and Wildlife Management in Costa Rica. Columbia University Press, New York, pp. 199–209. Lehner, B., Verdin, K., Jarvis, A., 2006. HydroSHEDS Technical Documentation. In: World Wildlife Fund US, Washington, DC. Available at http://hydrosheds.cr.usgs.gov. Lindenmayer, D., Burgman, M.A., 2005. Practical Conservation Biology. Csiro. Mackey, B.G., Berry, S.L., Brown, T., 2008. Reconciling approaches to biogeographical regionalization: a systematic and generic framework examined with a case study of the Australian continent. Journal of Biogeography 35, 213–229. Mackey, B., Lindenmayer, D., 2001. Towards a hierarchical framework for modelling the spatial distribution of animals. Journal of Biogeography 28, 1147–1166. Mackey, B.G., Nix, H.A., Hutchinson, M.F., Macmahon, J.P., Fleming, P.M., 1988. Assessing representativeness of places for conservation reservation and heritage listing. Environmental Management 12, 501–514. Margules, C., Nicholls, A., Pressey, R., 1988. Selecting networks of reserves to maximise biological diversity. Biological Conservation 43, 63–76. Margules, C.R., Pressey, R.L., 2000. Systematic conservation planning. Nature 405, 243–253. Mark, A., 1985. The botanical component of conservation in New Zealand. New Zealand Journal of Botany 23, 789–810. Martinez, I., Carreno, F., Escudero, A., Rubio, A., 2006. Are threatened lichen species well-protected in Spain? Effectiveness of a protected areas network. Biological Conservation 133, 500–511. Maxted, N., Dulloo, E., 2008. Gap analysis: a tool for complementary genetic conservation assessment. Diversity and Distributions 14, 1018–1030. Moilanen, A., 2007. Landscape zonation, benefit functions and target-based planning: unifying reserve selection strategies. Biological Conservation 134, 571–579. Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Natural Resource Management Ministerial Council, 2004. Directions for the National Reserve System. A Partnership Approach. Australian Government, Department of the Environment and Heritage, Canberra, ACT. Newell, G., White, M., Griffioen, P., 2009. Potential Impact of Climate on Selected Terrestrial Ecosystems of Northern Victoria. Department of Sustainability and Environment Report. NSW NPWS, 1999. C-Plan Conservation Planning Software. User Manual for C-Plan Version 3.2. Oldfield, T.E.E., Smith, R.J., Harrop, S.R., Leader-Williams, N., 2004. A gap analysis of terrestrial protected areas in England and its implications for conservation policy. Biological Conservation 120, 303–309. Pearlstine, L., Smith, S., Brandt, L., Allen, C., Kitchens, W., Stenberg, J., 2002. Assessing state-wide biodiversity in the Florida Gap analysis project. Journal of Environmental Management 66, 127–144. Peterson, A.T., Nakazawa, Y., 2008. Environmental data sets matter in ecological niche modelling: an example with Solenopsis invicta and Solenopsis richteri. Global Ecology and Biogeography 17, 135–144. Pimm, S.L., Russell, G.J., Gittleman, J.L., Brooks, T.M., 1995. The future of biodiversity. Science 269, 347.

Pressey, R., 1994. Ad hoc reservations: forward or backward steps in developing representative reserve systems? Conservation Biology 8, 662–668. Pressey, R., Whish, G., Barrett, T., Watts, M., 2002. Effectiveness of protected areas in north-eastern New South Wales: recent trends in six measures. Biological Conservation 106, 57–69. Pyke, C.R., Fischer, D.T., 2005. Selection of bioclimatically representative biological reserve systems under climate change. Biological Conservation 121, 429–441. Rodrigues, A.S.L., Andelman, S.J., Bakarr, M.I., Boitani, L., Brooks, T.M., Cowling, R.M., Fishpool, L.D.C., da Fonseca, G.A.B., Gaston, K.J., Hoffmann, M., 2004. Effectiveness of the global protected area network in representing species diversity. Nature 428, 640–643. Rodríguez, J.P., Brotons, L., Bustamante, J., Seoane, J., 2007. The application of predictive modelling of species distribution to biodiversity conservation. Diversity and Distributions 13, 243–251. Sappington, J.M., Longshore, K.M., Thomson, D.B., 2007. Quantifying landscape ruggedness for animal habitat analysis: a case study using bighorn sheep in the Mojave Desert. Journal of Wildlife Management 71 (5), 1419– 1426. Sarkar, S., Margules, C., 2002. Operationalizing biodiversity for conservation planning. Journal of Biosciences 27, 299–308. Sarkar, S., Pressey, R.L., Faith, D.P., Margules, C.R., Fuller, T., Stoms, D.M., Moffett, A., Wilson, K.A., Williams, K.J., Williams, P.H., 2006. Biodiversity conservation planning tools: present status and challenges for the future. Annual Review of Environment and Resources 31, 123–159. Schmidt, F., 2002. TOPOCROP Version 1.2. An extension for Arcview 3.x that implements various Terrain Indices, Institute for Geodesy and Geo-Informatics, University of Rostock. Scott, J.M., 1993. Gap analysis: a geographic approach to protection of biological diversity. Wildlife Monographs. Scott, J.M., Davis, F.W., McGhie, R.G., Wright, R.G., Groves, C., Estes, J., 2001. Nature reserves: do they capture the full range of America’s biological diversity? Ecological Applications 11, 999–1007. Smith, R.J., Goodman, P.S., Matthews, W.S., 2006. Systematic conservation planning: a review of perceived limitations and an illustration of the benefits, using a case study from Maputaland, South Africa. Oryx 40, 400–410. Snelder, T.O.N.H., Dey, K.L., Leathwick, J.R., 2007. A procedure for making optimal selection of input variables for multivariate environmental classifications. Conservation Biology 21, 365–375. Strittholt, J.R., Boerner, R.E.J., 1995. Applying biodiversity gap analysis in a regional nature reserve design for the Edge of Appalachia, Ohio (USA). Conservation Biology 9, 1492–1505. Tear, T.H., Kareiva, P., Angermeier, P.L., Comer, P., Czech, B., Kautz, R., Landon, L., Mehlman, D., Murphy, K., Ruckelshaus, M., 2005. How much is enough? The recurrent problem of setting measurable objectives in conservation. BioScience 55, 835–849. Tognelli, M.F., de Arellano, P.I.R., Marquet, P.A., 2008. How well do the existing and proposed reserve networks represent vertebrate species in Chile? Diversity and Distributions 14, 148–158. Trakhtenbrot, A., Kadmon, R., 2005. Environmental cluster analysis as a tool for selecting complementary networks of conservation sites. Ecological Applications 15, 335–345. Wessels, K., Freitag, S., Van Jaarsveld, A., 1999. The use of land facets as biodiversity surrogates during reserve selection at a local scale. Biological Conservation 89, 21–38. Wilford, J.R., Bierwirth, P.N., Craig, M.A., 1997. Application of air-borne gammaray spectrometry in soil/regolith mapping and applied geomorphology. ASGO Journal Australian Geology and Geophysics 17, 201–216. Woinarski, J., Price, O., Faith, D., 1996. Application of a taxon priority system for conservation planning by selecting areas which are most distinct from environments already reserved. Biological Conservation 76, 147–159.