**3. Identifying species and distributions**

Although over 320 species of bat are currently described from SEA (Simmons, 2005; Kingston, 2010) research in the area has been sporadic and the rate of species discovery is now high for not only bats (Bumrungsri et al., 2006), but across many other taxa (Duckworth & Hedges, 1998; Bain et al., 2003; Giam et al., 2010). Recent research has revealed that many bats previously regarded as one species are in reality complexes, comprising a number of cryptic species (Soisook et al., 2008, 2010; Francis et al., 2010). Therefore before any conservation measures can be put in place the distribution and status of current species must first be established. SEA has some of the highest diversity of bats on the planet in addition rate of species discovery (Simmons & Wetterer, 2011). A projection of the species richness of 171 species throughout SEA (Fig.1) shows that most forested regions still retain high species richness, and therefore present priority regions for research.

However recent research has clearly demonstrated that currently known SEA bat species only represent a fraction of total species numbers (Francis et al., 2010; Giam et al., 2010; A.C.Hughes et al., in prep a). Both recent taxonomic and genetic research show that much further work is needed in order to identify all species in the region, and similar trends are liable to exist across biotic groups. Species identification is clearly a priority, because it is impossible to try to develop effective conservation strategies when there is little understanding of the true ranges of many species; and when species currently classified as showing large distributions are in actuality made up of a number of cryptic species with small ranges and much smaller populations (A.C.Hughes et al., in prep a). Both taxonomic (Soisook et al., 2008, 2010) and genetic work (Francis et al., 2010) demonstrate that there are many currently undescribed and potentially cryptic species throughout SEA.

 Methods used to determine species present obviously involve detailed taxonomic surveys (as advocated by Webb et al., 2010), in addition to genetic analyses where possible. However other protocols for species identification and monitoring may also be valuable components of species discovery in some taxa, such as the use of call analysis to identify cryptic bat species (e.g. G. Jones & Van Parijs, 1993). In such cases the identification of potentially cryptic species may begin with call analysis, as was recently found to be the case in Hipposideros bicolor, (Douangboubpha et al., 2010). Acoustic monitoring also provides a means of potentially monitoring population trends as well as identifying possible cryptic species (K. E. Jones et al., 2011). Two protocols have recently been developed which describe the potential for using localised call libraries for identifying bat species in SEA (A.C.Hughes et al., 2010, in press). Once acoustic identification libraries have been developed then acoustic surveys and inventories of surrounding regions (e.g. 1o of the areas used to develop the library) can be made to identify species present (using discriminant function analysis) and the presence of species outside their known range. The presence of novel call variants could cue and promote further research to determine if sub-species or cryptic species are present, and the spatial distributions of call variants of some species suggests spatial

effective impact mitigation strategies. Though in this chapter bats will provide the main case study (due to their potential as indicator species) most of what will be discussed here is broadly applicable for the conservation of biodiversity throughout SEA, and in developing strategies for mitigating species loss in other regions of the world which faces similar issues

Although over 320 species of bat are currently described from SEA (Simmons, 2005; Kingston, 2010) research in the area has been sporadic and the rate of species discovery is now high for not only bats (Bumrungsri et al., 2006), but across many other taxa (Duckworth & Hedges, 1998; Bain et al., 2003; Giam et al., 2010). Recent research has revealed that many bats previously regarded as one species are in reality complexes, comprising a number of cryptic species (Soisook et al., 2008, 2010; Francis et al., 2010). Therefore before any conservation measures can be put in place the distribution and status of current species must first be established. SEA has some of the highest diversity of bats on the planet in addition rate of species discovery (Simmons & Wetterer, 2011). A projection of the species richness of 171 species throughout SEA (Fig.1) shows that most forested regions still retain

However recent research has clearly demonstrated that currently known SEA bat species only represent a fraction of total species numbers (Francis et al., 2010; Giam et al., 2010; A.C.Hughes et al., in prep a). Both recent taxonomic and genetic research show that much further work is needed in order to identify all species in the region, and similar trends are liable to exist across biotic groups. Species identification is clearly a priority, because it is impossible to try to develop effective conservation strategies when there is little understanding of the true ranges of many species; and when species currently classified as showing large distributions are in actuality made up of a number of cryptic species with small ranges and much smaller populations (A.C.Hughes et al., in prep a). Both taxonomic (Soisook et al., 2008, 2010) and genetic work (Francis et al., 2010) demonstrate that there are

 Methods used to determine species present obviously involve detailed taxonomic surveys (as advocated by Webb et al., 2010), in addition to genetic analyses where possible. However other protocols for species identification and monitoring may also be valuable components of species discovery in some taxa, such as the use of call analysis to identify cryptic bat species (e.g. G. Jones & Van Parijs, 1993). In such cases the identification of potentially cryptic species may begin with call analysis, as was recently found to be the case in Hipposideros bicolor, (Douangboubpha et al., 2010). Acoustic monitoring also provides a means of potentially monitoring population trends as well as identifying possible cryptic species (K. E. Jones et al., 2011). Two protocols have recently been developed which describe the potential for using localised call libraries for identifying bat species in SEA (A.C.Hughes et al., 2010, in press). Once acoustic identification libraries have been developed then acoustic surveys and inventories of surrounding regions (e.g. 1o of the areas used to develop the library) can be made to identify species present (using discriminant function analysis) and the presence of species outside their known range. The presence of novel call variants could cue and promote further research to determine if sub-species or cryptic species are present, and the spatial distributions of call variants of some species suggests spatial

high species richness, and therefore present priority regions for research.

many currently undescribed and potentially cryptic species throughout SEA.

to those discussed here.

**3. Identifying species and distributions** 

segregation which could denote cryptic species (A.C.Hughes et al., in prep a). Monitoring surveys are also essential to determine distribution and population trends, however funds and specialists are not always available to carry out this valuable work when it requires repeated taxonomic surveys and specialist knowledge. Acoustic analysis and monitoring only requires specialists initially, during the creation of acoustic libraries, and surveys can then be carried out by non-specialists or automated software programs (K.E. Jones et al., 2011). Thus protocols such as these provide a viable means of both identifying species present and subsequently monitoring trends, and may be able to detect variation over shorter periods than in trapping-based monitoring which has been previously been advocated (Meyer et al., 2010). Acoustic surveys are currently limited in species coverage, and are biased towards bat taxa that use high-intensity echolocation calls. Acoustic surveys are therefore best used side-by-side with conventional survey techniques such as using mistnets and harp traps in a standardised manner (MacSwiney G et al., 2008). However invasive trapping techniques are expensive and require highly trained experts, whereas acoustic surveys can be carried out with little training and recordings can then be forwarded to highly trained researchers for analysis, or analysed by software to provide standardised and comparable data for any region. If initially surveys combine both trapping and acoustic techniques to establish acoustic libraries within a given area then those libraries can subsequently be employed to monitor trends in many species across wide areas. The use of common species as indicators for abundance and distribution of rarer species has been found to be accurate in previous studies, as correlations have been found in the trends of common species with other species present (Pearman et al., 2010). Therefore even if acoustic surveys cannot cover all species, the trends in the distributions and populations of common species may still be more widely applicable.

Logistical constraints also mean that it is not always possible to survey all areas in a region, and thus methods which determine range based on limited spatial knowledge of an organism's total distribution provides a valuable tool when applied properly (i.e. predictive modelling approaches, Box 1, Fig. 1). Former distributions of species and zoogeographic constraints must also be considered and included in analyses of species distributions. Within SEA the geophysical history is to a large extent responsible for the current patterns of diversity and species' distributions, and thus analyses of present species distributions cannot be conducted without by making reference to the past (Woodruff, 2003). The connections and separations of the various parts of SEA during past time periods not only influence current distributions but further constrain possible responses to future change. A zoogeographic transition in the distributions of some animal groups centred around the Isthmus of Kra has persisted for over a million years (De Bruyn et al., 2004). Recent analyses (A.C. Hughes et al., 2011) show that although breaks in the distribution patterns of bats are apparent along the Thai peninsula, they occur not only at the Isthmus of Kra and are influenced by climatic discontinuities in conjunction with biogeographic consequences associated with the narrow breadth of the peninsula; and it is probable that these circumstances have also caused divisions known to occur in the distributions of other taxa in the region (J.B. Hughes, et al., 2003). Zoogeographic transitions have persisted over long time periods along the peninsula because the position of climatic boundaries appears remarkably constant. Climatic discontinuities continue to affect the distributions of species, and will also affect how effectively species can respond to climatic change in the future.

Mapping a Future for Southeast Asian Biodiversity 7

and therefore is a conservative projection, and though it is possible to prevent the loss of species due to deforestation in protected areas it is not possible to prevent species loss due to climatic change. Forest is becoming increasingly fragmented even within "protected areas", and mining rates in SEA are the highest in the tropics (Day & Ulrich, 2000). Mining not only destroys important roost sites (Clements et al., 2006), but also degrades areas and increases accessibility to previously remote areas (which in turn facilitates deforestation, McMahon, et al., 2000; Laurance, 2008a). Therefore not only are current suitable habitat and roosts being destroyed, but the distance between suitable areas may actually be increasing for the same reasons. Other factors such as fires are also prevalent through SEA, and fires have increasingly been found to move north in response to climatic change, therefore posing an increasing threat to the biodiversity of SEA (Taylor et al., 1999). Projections of total biodiversity loss currently estimate the extinction of up to 85% of current biodiversity in SEA within this century (Sodhi et al., 2010). However the estimates of undiscovered species show that we may potentially have only discovered around half of the species in many orders (Giam et al., 2010) and only around 40% of bat species (A.C. Hughes et al., in prep a). Groups containing cryptic species are likely to have particularly high numbers of undiscovered species, and this is highlighted in bats by recent genetic work (Francis et al., 2010). Species with smaller distributions are more likely to have specialist requirements (limiting overall distribution), and will be more susceptible to loss of range and therefore have a higher probability of extinction (Kotiaho et al., 2005). Hence many species currently regarded as widespread, and thus of "Least Concern" by the IUCN may comprise a complex of cryptic species each of which will show higher categories of threat. As almost all species analysed here (fig. 2) showed a loss in original habitat in all scenarios, and many of those species may be species complexes it is likely that impacts for many of the species will be worse than estimated during this study (fig. 2). Projections here (Fig. 2) only account for climate change, but cannot consider hunting, fires, mining and the plethora of other threats. Fungal diseases have recently devastated populations of North American bats (Blehert et al., 2009), in addition to South American frogs (Berger et al., 1999). Moreover the spread of pathogens has been associated with temperature change, for example the spread of chytrid fungus is believed to be related to global warming (Pounds et al., 2006; Boyles & Willis, 2010). Therefore the effect of climate change on species is dynamic and complex, as it has both direct and indirect implications for distributions and populations of all species. Furthermore climatic changes have already been shown to cause changes to the distribution of different biomes (Salazar et al., 2007), and hence has profound

SEA is currently in the midst of a biodiversity crisis which has been described as a 6th mass extinction (Myers, 1988). There are some undeniable implications of the current threats, and others such as the possibility of 'no-analogue' communities (Stralberg et al., 2009) and the effect of invasive species, which are less certain. However native species are likely to attempt to either migrate north spatially, or move to higher altitudes (Malcolm et al., 2006). Continued decreases in the patch sizes of rainforest will decrease species richness, and increasing accessibility for humans will increase the probability of hunting within areas. Increases in human population will negatively affect biodiversity, if current unsustainable practices continue. Not only is the modification of human activities necessary to decrease further species loss, but human intervention is necessary to allow species any opportunity to respond effectively to climatic changes. The methods to mitigate possible threats require

implications for species within those biomes.

detailed evaluation to try to curb species extinctions.

Identification of species present, their ranges and trends in distribution and population form an important first step in the development of effective conservation plans. Once these steps have been fulfilled then threats to current distributions and diversity can be analysed (Fig. 1) and necessary conservation actions planned.
