INTRODUCTION
Climate change
The impact of climate change has been identified as the greatest threat to global biodiversity (Thomas et al., 2004). Climate change refers to the global rise in average surface temperatures (Razgour et al., 2013; Brown & Yoder, 2015).
Human land use
Estimating the impact of climate change and human land use on species’ ranges
Many ENM algorithms have been developed to combine event and environmental data, including: artificial neural networks (Ripley, 1996), generalized augmented models (also known as augmented regression trees; Ridgeway, 1999), and MaxEnt (Phillips et al., 2006 ). . In contrast, bionomic variables mediated by dispersal, competition and predation have a stronger influence on the Elton niche (Soberón, 2007) and are measured at fine spatial scales (Colwell & Rangel, 2009; Devictor et al., 2010; Anderson, 2013). .
Madagascar and aspects of its biodiversity
- Climate change effects on biodiversity
- Human land use effects on biodiversity
- Bats
Decreasing precipitation levels and rising temperatures could have a negative impact on Madagascar's tropical forests (Kitula et al., 2015). Madagascar is often ranked among the top 10 poorest countries in the world (Thomas et al., 2008).
Research questions, objectives and predictions
MATERIALS AND METHODS
- Study area
- Occurrence data
- Climatic data
- Land use data
- Environmental niche models
- Ensemble modelling approach
- Model evaluation
- Species richness
- Coverage of species hotspots by Madagascar protected areas
- Niche overlap
- Gap analysis
24 (Community Climate System Model), a combined climate model for simulating the Earth's climate system (Zhao et al., 2010). Two null models of niche overlap were used: a test of niche identity and a test of background similarity (Warren et al., 2008).
RESULTS
ENMs of species
Potentially suitable habitat space for 25 Malagasy bat species created by ecological niche models using only climatic variables. Suitable areas are shown in a range of colors, from red which is very suitable to blue which is not suitable. Predicted potential distributions are shown for the (a) last interglacial, (b) last glacial maximum, (c) current, and (d) future climate scenarios.
Potentially suitable habitat space for 25 Malagasy bat species generated by ecological niche models using climate/land use variables. Projected potential distributions are shown for (a) last interglacial, (b) last glacial maximum, (c) current and (d) future climate scenarios.
Species richness
- Climate only
- Climate/land use
Map of Madagascar illustrating spatial prediction of total Madagascan bat richness built on synopsis of each ENMs for climatic variables across different climatic scenarios (a) last interglacial, (b) last glacial maximum, (c) present and (d ) future. The colors indicate the number of species per cell, darker colors contain a higher number of species as indicated on the key. Map of Madagascar illustrating spatial prediction of total Madagascan bat richness built on synopsis of each ENMs for climatic/land use variables across different scenarios (a) last interglacial, (b) last glacial maximum, (c) current and (d) ) future.
Species richness statistics of Malagasy bats for all climate periods in Madagascar with climate only and climate/land use variables. Mean richness per cell is shown including standard deviation, minimum and maximum species values and number of cells that presented them. Statistics for selected hotspots for last interglacial maximum, last glacial maximum, current and future species in Madagascar for both climate-only and climate/land-use variables.
Percent within PAs – refers to hotspot area relative to total island area.
Maximum and upper quartiles hotspots
- Climate only
- Climate/land use
Hotspots identified for the four climatic periods using climate variables (colored areas): (a) last interglacial, (b) last glacial maximum, (c) current, and (d) future. The current protected areas are indicated by black lines (the black boxed areas are restricted areas to indicate where the hotspots are located) and columns indicate the two selected cell categories. Below each map is indicated the percentage of hotspots that fall within the protected areas.
Hotspots identified for the four climatic periods using climatic/land-use variables (colored areas): (a) last interglacial, (b) last glacial maximum, (c) present, and (d) future. Below each map indicates the percentage of hotspots that fall within the protected areas.
Overlap in ENMs
- Climate only
- Climate/land use
Results of identity test and background similarity test of predicted ecological niches of three functional groups (FG).
Gap analysis and additional targeted areas to conserve Malagasy bats
- Climate only
- Climate/land use
Potentially suitable habitat space for the rest of the species (12 spp., 48%) decreased on average by 37%. Under the future climate change scenario, gains in suitable habitat space were predicted in one root, five open-range (5 spp., 78%), six root-knotted and one fruit bat species (Fig. 3.8). In contrast, suitable habitat space was predicted to decrease reminiscent of FG's species;.
The conservation objectives of suitable habitat areas predicted by all 25 species were not sufficiently covered (2 – 17%) in the PA for both current and future climate projections (Fig. 3.12 & 3.13). One species with a small distribution (<50,000 km2: Miniopterus mahafaliensis) had <5% of the total area of their suitable habitat covered by the PA. The potential suitable habitat space for the remaining species (10 spp., 40%) is projected to expand by an average of 30% in the future (Fig. 3.13).
Under future climate change, gains in potentially suitable habitat space were predicted in four bats, five outdoors and one fruit bat (Fig. 3.13).
DISCUSSION
Bat hotspots under different climate change and human land use scenarios
For example, shifts in vegetation caused by climate change may result in a decrease in suitable habitat space for some species but an increase in others (Scheel et al., 1996). Rising temperatures in Madagascar may reduce plant diversity in eastern humid forests (Brown et al., 2015). Species with narrow climatic ranges may be particularly vulnerable to such climate changes (Bellard et al., 2012).
In addition, increases in the frequency of cyclones are predicted in the southwest Indian Ocean (McBride et al., 2015). The predicted decline in hotspots of bat richness is consistent with observed and predicted changes in climate and human land use across the eastern, northern and western parts of Madagascar (Hannah et al., 2008). Here, the movement of forest-adapted species into more suitable habitat space in response to climate change may be further constrained by forest encroachment, deforestation, degradation and fragmentation (Harper et al., 2007).
The putative areas to which bats are likely to migrate following climate change may already be converted to human land use and consequently do not support bats (Smith et al., 2016).
Niche overlap of functional groups
When human land use was also considered in combination with future climate change, suitable habitat space for bat species and bat-rich hotspots decreased to an even greater extent. This decline in wealth hotspots from current to future climate and land use scenarios was mainly concentrated in the northwestern and eastern parts of the island and shifted from lowland to inland areas. Brown & Yoder (2015) modeled the effects of projected climate change and human land use on suitable habitat space for a variety of lemur species, and found similar shifts in the locations of species' suitable habitats, typically associated with declines in range and geographical positions of hotspots. were changed considerably. 2012) showed that the combined effects of climate change and land use would affect Southeast Asian bats (including forest-dependent species) with changes in their predicted ranges in the future. 2016) found marked effects on the distribution of African bats with the combined effects of future land use and climate change.
As predicted, changes in climate and land use had a greater effect on the ENM of bats in roosts than bats at the roost and in the open. This may be due to bats being highly adapted to foraging in vegetation, and vegetation is likely to be affected by changes in climate and land use. Although bats have high mobility, which allows them to exploit natural habitats in land-use areas (urban settlements and agricultural landscapes), bat species show high variability in their mesoscale dispersal capabilities, even within FGs (Smith et al., 2016).
Therefore, these models could be improved using the dispersal ability of individual species (Smith et al., 2016).
Coverage of bats in Madagascar’s protected areas
This species has roots in the travel tree (Ravenala madagascariensis), which is a pioneer plant of degraded forest habitats (Ralisata et al., 2015). 63 Similarly, estimated ranges of various Malagasy taxa (ants, butterflies, frogs, geckos, lemurs and plants) under current climate change were not represented in PAs (Kremen et al., 2008), with values from a low of 16.2 % for plants, a moderate value of 38.5%. The size and location of the additional areas are similar to those Kremen et al.
Protect e.g. a den site with high species richness or increase knowledge of the hunted roosting colonies such as Pteropus rufus (MacKinnon et al., 2003). In addition, management variables such as costs, opportunities or threats must also be considered (Gardner et al., 2013). Many of the areas prioritized for protection include areas where human populations depend on local natural resources (Gardner et al., 2013).
These species must be sensitive to changes, easy to sample and provide objective results (Moreno et al., 2007).
Model limitations and future work
One way to resolve conflict between the conservation of natural forests and the fulfillment of human requirements is by substituting alternative sources of wood, for example plant plantations on degraded land (Hannah et al., 2008). However, this can be particularly expensive; growing plantations equivalent to a quarter of the natural forest outside PAs would cost approximately US$ 400 million (Hannah et al., 2008). However, few studies have tested this (Smith et al., 2016), for example the correlation between bat indicators of habitat changes with those of other taxa (e.g. mammals and birds; Brooks, 2007).
An advantage of mechanistic models is that they contain clearly defined parameters and can therefore provide a better understanding of the underlying factor(s) that determine responses to environmental change (Dormann et al., 2012). Mechanistic models can better address management questions given their ability to go beyond known conditions and identify properties that shape biogeography (Evans et al., 2015). Studies that include morphological, demographic, and genetic data can further refine predictions about bat species distributions (Razgour et al., 2016).
For example, the use of genetic data provides insight into the evolutionary history of bat populations (Flanders et al., 2011) and helps identify locations of high genetic diversity (Razgour et al., 2016).
Conclusions
Bellard C., Bertelsmeier C., Leadley P., Thuiller W., & Courchamp F. 2012) Impacts of climate change on the future of biodiversity. 2013). 2015) Pathogenic enterobacteria in lemurs associated with anthropogenic disturbance. 2012) Functional and phylogenetic approaches to predicting species responses to climate change. 2004). Distler T., Schuetz J.G., Velásquez-Tibatá J., & Langham G.M. 2015) Stacked species distribution models and macroecological models provide congruent projections of bird species richness under climate change.
Plant extinction risk under climate change: Are predicted range shifts alone a good indicator of species' vulnerability to global warming. The physical scientific basis, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Bats limit arthropods and herbivores in a tropical forest. 2013) Ecological and economic importance of bats (order Chiroptera). 2010) Correlative and mechanistic models of. species distribution provides congruent predictions under climate change.
Patterning of genetic variation in range-edge populations under past and future climate change. Wing load negatively correlates with genetic structure of eight Afro-Malagasy bat species (Molossidae). 2004) Extinction risk from climate change. Do diversification models of the Madagascar biota explain the population structure of the endemic bats Myotis goudoti (Chiroptera: Vespertilionidae).
Supplementary materials for Chapter 2
Supplementary materials for Chapter 3
Ensemble modeling performance measures resulting from fitting environmental separate models (ENMs; 10 models used) of 25 Malagasy bat species to current climate and climate/land use variables.
