Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2021 Jan 20;16(1):e0238669. doi: 10.1371/journal.pone.0238669

Protected areas network is not adequate to protect a critically endangered East Africa Chelonian: Modelling distribution of pancake tortoise, Malacochersus tornieri under current and future climates

Abraham Eustace 1,*, Luíz Fernando Esser 2, Rudolf Mremi 3, Patrick K Malonza 4, Reginald T Mwaya 3
Editor: Stephanie S Romanach5
PMCID: PMC7816999  PMID: 33471868

Abstract

While the international pet trade and habitat destruction have been extensively discussed as major threats to the survival of the pancake tortoise (Malacochersus tornieri), the impact of climate change on the species remains unknown. In this study, we used species distribution modelling to predict the current and future distribution of pancake tortoises in Zambezian and Somalian biogeographical regions. We used 224 pancake tortoise occurrences obtained from Tanzania, Kenya and Zambia to estimate suitable and stable areas for the pancake tortoise in all countries present in these regions. We also used a protected area network to assess how many of the suitable and stable areas are protected for the conservation of this critically endangered species. Our model predicted the expansion of climatically suitable habitats for pancake tortoises from four countries and a total area of 90,668.75 km2 to ten countries in the future and an area of 343,459.60–401,179.70 km2. The model also showed that a more significant area of climatically suitable habitat for the species lies outside of the wildlife protected areas. Based on our results, we can predict that pancake tortoises may not suffer from habitat constriction. However, the species will continue to be at risk from the international pet trade, as most of the identified suitable habitats remain outside of protected areas. We suggest that efforts to conserve the pancake tortoise should not only focus on protected areas but also areas that are unprotected, as these comprise a large proportion of the suitable and stable habitats available following predicted future climate change.

Introduction

Over the past few decades, there has been growing interest in species distribution models (SDMs) as fundamental tools for the studies of ecology, biogeography, and biodiversity conservation [1–4]. These models are used to enhance understanding of the factors that alter species distribution, which is critical for adjusting and designing appropriate conservation strategies under current and future climatic scenarios [3, 5, 6]. Such adjustments are necessary because climate change poses a severe threat to the conservation of natural landscapes and species across the globe and is reported to be among the primary drivers of the current loss of global biodiversity [7–9]. Climate change has also been reported to accelerate shifts in range extension [7] and the shrinkage of some species [6, 10–12].

Tropical environments are widely recognized as biodiversity regions [13] with ideal climatic conditions for the survival of different species including reptiles [14]. However, reptiles are currently facing severe threats because of climatic changes [6, 15]. Because reptiles are sensitive to environmental change, it is undeniably that climate change affects reptile biodiversity directly by altering their distribution patterns [6, 7] and indirectly by threatening conservation areas, making them less habitable for reptiles [16]. For instance, Meng et al. [15] have reported that out of the 274 Tanzania reptile species they studied, 71% (194 reptile species) are vulnerable to climate change, suggesting that climate change affects reptilian diversity both directly or indirectly. In a different study, predictions about the environmental responses of reptiles to future climatic conditions made using SDMs showed that four endemic Moroccan reptilian species are highly vulnerable to extinction in Morocco if climatic disturbance prevails as predicted [6]. The same study concluded that reductions in species-rich areas is also likely in future climatic scenarios [6].

Like other reptiles, Malacochersus tornieri hereafter referred to as the pancake tortoise, is not immune to the effects of climate change. The pancake tortoise is a small, soft-shelled, dorsoventrally flattened chelonian with discontinuous distribution in the scattered rocky hills and kopjes of the savannas of south-eastern and northern Kenya and northern, eastern, and central Tanzania [17–20]. The presence of pancake tortoises has also been reported in northern Zambia [21]. The areas in which pancake tortoises can be found are typically semi-arid; these areas are classified as having a dry climate, corresponding to both Zambezian and Somalian biogeographic regions, according to Linder et al. [22]. The Zambezian biogeographical region is a wider biogeographical region, spreading across Africa from Namibia to Tanzania, while the Somalian biogeographic region is considered a refugium for arid-adapted plants and a centre of endemism for wide-range of animal taxa including reptiles [23, 24].

Although the international animal trade and habitat destruction have been cited as the major threats to the survival of the pancake tortoise [18, 19, 25, 26], the impact of climate change on the species remains largely unknown. Despite the fact that the IUCN has identified climate change as one of the threats to pancake tortoise populations [18, 19], to our understanding, there is no study that has assessed the impact of climate change on the future distribution pattern of pancake tortoises. Understanding these climatic patterns is one of the important steps in setting appropriate plans for re-introductions and translocations of the species which are important activities for conservation of species with threatened populations or restricted range. Furthermore, the IUCN’s Guidelines for Re-Introductions and Other Conservation Translocations [27] has indicated climate-matching of recipient sites is important for understanding suitability of these areas for introduced/translocated species. Considering that the pancake tortoise is critically endangered [17–20] and listed in the CITES Appendix II [21], understanding current and future climatic habitats suitable for this species could be an essential step in charting out a realistic conservation plan for the species. Therefore, in this study, we used species distribution modeling (SDM) to determine current and future climatic habitats suitable for the pancake tortoise. Identifying these climatic suitable habitats, might help to avoid uncertainties in selecting areas for translocation or introduction while providing a higher chance of success [27–29].

With time, the ongoing impacts of climate change are expected to inflict changes to suitable habitats for pancake tortoises both within and outside of protected areas [16, 30, 31]. While protected areas remain an essential approach for conserving and protecting biodiversity against human-mediated threats [15, 32], it could be challenging to protect the endangered species that inhabit areas outside of protected lands [33] such as Kenyan pancake tortoise [18]. For the development of specific and appropriate management and conservation plans for pancake tortoises, it is crucial to understand whether these protected areas will continue to be viable for protecting suitable habitats for the species in the event of climate change. Considering species’ range varies under different climatic scenarios [3, 5, 34, 35] while the size of most protected areas tends to remain the same [36], more species may eventually be placed at risk of extinction, especially threatened species [37]. Therefore, in order to align protected areas with suitable habitat ranges [38] and enhance the conservation of threatened species in different climatic scenarios [36], SDMs can be used.

SDMs have been used to assess the impact of climate change on the distribution of different species (e.g. [30, 31, 36, 39]). These models use location data and environmental variables to predict the suitable distributional range of a species under climate change conditions [30, 36], which is essential when designing adequate species management programmes, as well as for endangered species conservation planning [40]. Although Bombi et al. [41] have used SDMs to model the distribution of all African tortoise species, including the pancake tortoise, their study did not predict the future distribution of the species. In this study, we used SDM to assess the distribution of pancake tortoises under current and future climatic conditions and to investigate how much of the climatically suitable habitat occurs within the Protected Areas Network in the Somali-Maasai and Zambezian biogeographical regions. Specifically, we assessed (i) the current and future climatically suitable areas for pancake tortoises, (ii) the occurrence of climatically persistent areas over time (henceforward, stable areas) and (iii) whether the protected areas will be viable for the conservation of the species. This study may inform species management approaches [42], including identifying suitable areas for translocation [29, 43, 44] and the establishment of nature reserves where species can be protected with minimal human intervention [10].

Methodology

Study area

We predicted current and future climatically suitable habitats for pancake tortoises within the two major biogeographical regions of Africa in which the animal occurs naturally (Fig 1B). These regions are the Somali-Masai Regional Centre of Endemism (Somali-Masai RCE) and the Zambezian Regional Centre of Endemism (Zambezian RCE), both of which fall within the semi-arid climatic belt of eastern-south Africa [45, 46]. The Somali-Masai RCE covers approximately 1.87 million km2 of arid savannah, extending from north-eastern Somalia to the north-eastern province of Kenya and reaching south through Tanzania into the valley of the Great Ruaha; it ends north of Lake Malawi [22, 46, 47]. The Somali-Masai RCE harbours approximately 4,500 plant species, of which 31.00% are endemic in the region [45, 46]. The dominant vegetation in this region is Acacia spp. The Zambezian RCE (3.77 million km2) extends in the northeast from the Somali-Masai RCE, and its distribution coincides with the Guinea savannas and woodlands and the Karoo-Namib RCE in the southwest [37, 38]. It covers the whole of south-central Africa, from the Atlantic seaboard of Angola to the entirety of Mozambique, Tanzania and the uplands of Kenya and Ethiopia [22, 47]. In terms of plant richness, it is more diverse than the Somali-Masai RCE, hosting about 8,500 plant species, out of which 54.00% are endemic in the region [45, 46].

Fig 1.

Fig 1

Open in a new tab

Current natural occurrence of pancake tortoise (Malacochersus tornieri), based on the data obtained in this study, with a 1-degree wide buffer around each presence record, at (A) continental scale with the Zambezian and Somalia biogeographical region (overlaid red polygon) and (B) regional scale. Background image was accessed from Natural Earth (public domain): http://www.naturalearthdata.com/.

Study species

The critically endangered pancake tortoise, Malacochersus [20], is a monotypic genus endemic to East Africa [18, 48]. In East Africa, the pancake tortoise is restricted to Somali-Maasai and Zambezian vegetations [20, 26, 49–51]. In Tanzania, the species distribution is discontinuously scattered from the south-eastern shores of Lake Victoria to the Maasai Steppe and southward to Ruaha National Park [18, 20, 26]. In Kenya, the distribution of pancake tortoises is disconnected from the northern to southern areas, lying between central to south-eastern regions of the country [18, 20]. In Zambia, the species has been recorded only in the northern Nakonde District that borders Tanzania [20, 21]. The preferred micro-habitats for pancake tortoises are kopjes, rock outcrops and rocky hillsides [17, 18, 26] with an annual rainfall of 250–500 mm [50] and an elevational range of 400–1,600 m above sea level [52]. From the past two generations to the next generation, the observed and expected population of pancake tortoises is expected to decline by 80.00%, with overexploitation and habitat destruction being the primary drivers [20]. Currently, the IUCN identifies biological resource use (intentional use) and agriculture (small-holder farming), as well as climate change (severe drought), as among the major threats to the habitat and population of pancake tortoises [20].

Pancake tortoise occurrence data

We obtained occurrence data from the field, online databases and previous studies. We collected pancake tortoise location data from eight sites in Tarangire National Park, three sites in the Babati district of the Manyara region, five sites in the Kondoa districts and two sites in Chemba district, both from the Dodoma region in central Tanzania. The permit for conducting field work was granted by Commission for Science and Technology (COSTECH) and Tanzania Wildlife Research Institute (TAWIRI) while free access to protected areas was granted by Tanzania National Parks (TANAPA).

We also downloaded pancake tortoise locations from the GBIF (https://www.gbif.org/) and VertNet (http://vertnet.org/) by using rgbif [53] and rvertnet [54] R packages respectively. Both databases were accessed on 5 January 2020, and we downloaded all Malacochersus tornieri locations identified in Tanzania and Kenya. We did not find any pancake tortoise occurrences in Zambia in the two databases. From the online databases, we excluded data with absent or incomplete coordinates and duplicate locations as well as non-natural locations, such as tortoise collection points, captive breeding sites and pet-animal release sites. Additionally, we searched for pancake tortoise locality records in the EMYSystem Global Turtle Database [55] and then used Elevation Map (https://elevationmap.net/) to obtain location coordinates.

From previous studies, we extracted the names of the places where pancake tortoises were recorded/observed. For Tanzania, we used sites mentioned by Klemens and Moll [26] as well as point locations collected by Zacarias [56], while for Zambia we used point locations mentioned by Chansa and Wagner [21]. In Kenya, we obtained pancake tortoise sites from Malonza [18] and Kyalo [52]. After obtaining the site names, we used Google Maps (https://www.google.co.tz/maps/), Elevation Map (https://elevationmap.net/) and Mindat (https://www.mindat.org/) to obtain coordinates for each site. If the site was not available online, we contacted individuals currently or previously working in the area in order to obtain coordinates. From all sources, we obtained data for a total of 224 occurrences, with most occurrence points falling within the current IUCN pancake tortoise distribution range (Fig 1).

Bioclimatic variables

Nineteen bioclimatic variables (BIO 1–19) were obtained from the CHELSA database [57] with 30 arc-seconds resolution. The modelling domain comprised Zambezian and Somalian biogeographical regions [22]. These regions were selected because they represent the areas where pancake tortoises exist naturally [20, 21, 26, 49–51]. We obtained variables for the two intermediate Representative Concentration Pathways (RCPs), RCP 4.5 [58] and RCP 6.0 [59], for the years 2050 (mean climate between 2041 and 2060) and 2070 (mean climate between 2061 and 2080). These mid-impact RCPs are the most desirable for future conservation planning, since they present a more realistic path compared to the extreme RCPs (2.6 and 8.5) which may incorporate too many uncertainties [60], causing projections to be unreliable. Future scenarios’ uncertainty were also accessed through ten Global Circulation Models (GCMs) available in CHELSA; we avoided those with high co-dependency [61], resulting in the selection of MIROC5, CESM1-CAM5, IPSL-CM5A-MR, FIO-ESM, GISS-E2-H, CSIRO-Mk3-6-0, GISS-E2-R, GFDL-ESM2G, MIROC-ESM-CHEM and MRI-CGCM3 (S1 Table). Each GCM is a model trying to explain how the atmosphere works. We used multiple GCMs to dissolve the effect of one unique GCM and improve predictions [62].

Variables were first submitted to a visual analysis, in which we deleted both the precipitation of the warmest quarter (BIO 18) variable and the precipitation of the coldest quarter (BIO 19) variable due to statistical artifacts, that may not represent the continuous gradient of reality, in the study region. Those artifacts are generated due to a difference in which quarter is the warmest (e.g. BIO18), causing the precipitation of one cell to be the sum from January-February-March, while the very next cell is the summed precipitation from February-March-April. We then masked variables with one degree-wide buffer from each presence record (Fig 1) and excluded variables with a high variance inflation factor (VIF > 3) and highly correlated variables (r > 0.7). This left us with six variables: the mean diurnal range (BIO 2), the isothermality (BIO 3), the mean temperature of the wettest quarter (BIO 8), the precipitation of the wettest month (BIO 13), the precipitation of the driest month (BIO 14) and the precipitation seasonality (BIO 15). These six variables were used to calculate the climatic niche of the species. The selection routine was performed using the usdm package [63] in R 3.6.2 [64]. Models were generated with variables at 30 arc-seconds resolution, while the rasters used to project models were upscaled at a factor of 10, resulting in rasters with a resolution of 2.5 arc-minutes.

Species distribution modelling

For the SDMs, we applied an ensemble method using the sdm package [65] in R 3.6.2 [64]. We implemented five algorithms using different approaches, with proper pseudo-absence selection, following Barbet-Massin et al. [66], as follows: MaxEnt, a machine-learning approach, with 1,000 randomly selected pseudo-absences; Multivariate Adaptive Regression Splines, a regression-based approach, with 100 randomly selected pseudo-absences; Multiple Discriminant Analysis, a classification approach, with 100 pseudo-absences randomly selected outside a surface-range envelope; Random Forest, a bagging approach, with 224 pseudo-absences randomly selected outside a surface-range envelope; and BIOCLIM, an envelope approach, with 100 randomly selected pseudo-absences. Algorithms were implemented using standard parameterization from the sdm package [65] hence the algorithms were not tuned. Model evaluation was performed with ten runs of a four-fold cross-validation technique (75.00% training and 25.00% test). In each run, we calculated true skill statistics (TSS) and the area under the receiver operating characteristic (AUC). To build ensemble models for each scenario, and after some pre-analysis, we selected models with TSSs and AUCs higher than the mean plus half the standard deviation. The mean AUC value was 0.958, with a standard deviation of 0.059 and a threshold equal to 0.988. The mean TSS value was 0.861, with a standard deviation of 0.112 and a threshold equal to 0.917. Selected models were projected into current and future scenarios and then binarized using the AUC threshold to avoid the use of subjective thresholds. Ensembles from future scenarios were built as a committee average of binarized rasters. Afterwards, we normalized the resulting rasters. This returned an ensemble in which 1 represents sites where all models agree with presences, 0 represents sites where all models agree with absences and the values in between are subject to uncertainty, where 0.5 represents cells with the highest uncertainty (i.e. half of the models agree with an absence, while the other half agree with a presence). We also built three potential refugees for the species by summing the normalized rasters from the five scenarios (current, RCP 4.5/2050, RCP 4.5/2070, RCP 6.0/2050 and RCP 6.0/2070). Then, we applied three thresholds, which were calculated by extracting all values greater than zero from the raster and obtaining the 90th, 95th and 99th quantiles (2.179, 2.850 and 3.930, respectively).

We calculated climatically suitable areas using a weighted method, multiplying the cell’s committee average by the cell area and summing all values within the rasters. This conservative method was intended to consider the uncertainty underlying each cell, as well as the different occupation proportions. Therefore, if a cell had 0.5 value (i.e. 50% chance of the species to occur in the cell), we calculated 50% of the cell area and add it to the total area occupied by the species. We applied this method to all scenarios, as well as, to every country present in the Zambezian and Somalian biogeographical regions [22]. We also masked results from area calculations with the World Database on Protected Areas v. 3.1 polygons [44] to estimate the climatically suitable areas under protection in the regions, countries and scenarios. Area calculations were performed in R 3.6.2 [64].

To get response curves for each variable, we extracted response data from each algorithm and made a regression analysis thorough locally estimated scatterplot smoothing (LOESS), with a span window of 0.5- and one-degree polynomial. This method fits multiple lines using half of the whole data. Each time a line is fitted, we exclude the first record from response data and include the next, until we have included all records.

Results

Our model demonstrated high performance, with an average AUC of 0.958 (SD = 0.059) and an average TSS of 0.861 (SD = 0.112). Our results show that the probability of M. tornieri occurrence increases with an increase of BIO 2, BIO 3, BIO 8, BIO 13 and BIO 15 and started to drop when the optimum condition has reached (S1 Fig). Conversely, the probability of occurrence for the species decreases with an increase in BIO 14 and started to increase after the optimum condition has reached (S1 Fig). However, BIO 3 showed the highest contribution relatively to others (S2 Fig) in predicting the distributional range of pancake tortoise in the Zambezian and Somalian biogeographical regions.

Currently, in the Zambezian and Somalian biogeographical regions, the pancake tortoise has a more extensive range in Tanzania and Kenya than in other countries present in the region (Fig 2). Although there is currently no evidence of records of pancake tortoises in Angola and Ethiopia, surprisingly, the model predicted patches of climatically suitable habitats in those countries under the current climatic scenario (Fig 2). Additionally, the model predicted that the current suitable distribution range of pancake tortoises is 90,668.75 km2, with Kenya contributing 61.10% of the current total range, followed by Tanzania (30.32%), Ethiopia (5.03%) and Angola (3.55%) (Table 1). Considering future climatic scenarios, we predicted that the pancake tortoise’s suitable habitat would not decrease. This was observed through the expansion of suitable habitats as predicted by RCP 4.5 and RCP 6.0 (Fig 2 and Table 1). The model predicted that the current distribution range would expand by 303.95% in the year 2050 and 342.47% in the year 2070 for RCP 4.5 and by 278.81% in the year 2050 and 311.99% in the year 2070 under RCP 6.0 (Table 1). Similar to the current scenario, we predict that Kenya and Tanzania will continue to have a larger suitable area (Fig 2 and Table 1) than other countries in the future. However, the distributional range is predicted to expand from the current four countries to ten countries in the future (Table 1).

Fig 2. Distribution of pancake tortoise (Malacochersus tornieri) in the Somalia and Zambezian biogeographical regions.

Fig 2

Open in a new tab

Current and future (2050 and 2070) climatic suitable habitat for pancake tortoise in the Zambezian and Somalia biogeographical regions considering six bioclimatic variables and two future climate scenarios. Warmer colours show more suitable areas, ranging from red to green. Background image was accessed from Natural Earth (public domain): http://www.naturalearthdata.com/.

Table 1. Malacochersus tornieri suitable area (km2) in the countries present in the Zambezian and Somalia biogeographical regions for the current and future climate scenarios.

Country Current Future
RCP 4.5 RCP 6.0
2050 2070 2050 2070
Total % of suitable protected habitat Total % of suitable protected habitat Total % of suitable protected habitat Total % of suitable protected habitat Total % of suitable protected habitat
Kenya 55,401.76 35.68 122,938.04 22.19 129,259.60 21.82 121,659.50 22.56 124,362.90 22.48
Tanzania 27,489.20 33.72 136,652.13 48.95 152,800.60 48.03 126,815 46.46 149,008.60 47.79
Zambia 0 0 0 0 0 0 0 0 0 0
Mozambique 0 0 1,906.56 12.75 2,424.43 16.39 996.42 6.42 1,906.07 8.95
Malawi 0 0 25.65 0 34.47 0 0 0 38.30 0
Zimbabwe 0 0 95.31 33.34 450.11 19.67 282.20 36.24 44.29 49.99
Angola 3,219.98 0 18,528.14 1.20 19,549.08 0.73 18,689.62 1.50 20,321.14 1.64
Namibia 0 0 0 0 0 0 0 0 0 0
Botswana 0 0 0 0 0 0 0 0 0 0
Somalia 0 0 10,233.67 0 12,181.94 0 11,891.61 0 7,692.55 0
Ethiopia 4,557.81 6.84 71,044.64 28.52 76,881.24 26.89 60,338.92 29.05 65,144.88 27.52
Democratic Republic of Congo 0 0 4,283.13 0.69 6,532.77 3.30 2,327.59 4.43 4,388.17 1.40
Burundi 0 0 0 0 0 0 0 0 0 0
Rwanda 0 0 0 0 0 0 0 0 0 0
Uganda 0 0 11.15 0 31.08 0 44.59 0 38.85 0
South Africa 0 0 0 0 0 0 0 0 0 0
TOTAL 90,668.75 32.37 366,257.22 31.39 401,179.70 30.69 343,459.60 30.41 373,547.30 31.50
% of change relative to the current scenario 303.95% 291.73% 342.47% 319.48% 278.81% 255.85% 311.99% 300.94%
Open in a new tab

‘Total’ is the total suitable area and ‘% of protected suitable habitat’ is the percentage of protected area that overlaps with the species suitable habitat.

The highly suitable areas (indicated by higher committee averages) are currently present in Kenya and Tanzania (Fig 2), where the species occurs naturally. In the future, highly suitable habitats will expand into Ethiopia as well (Fig 2); however, the species has not yet been recorded in that country. Although there were observations of pancake tortoises in Zambia (Fig 1), our model predicted that the area is not climatically suitable for pancake tortoises in the current and future scenarios (Fig 2).

Considering protected lands, we found that a larger suitable habitat for pancake tortoises lies outside of the current Protected Areas Network in both current and future climatic scenarios (Table 1). Currently, 32.37% of the suitable pancake tortoise habitat lies inside of protected areas (Table 1). In the future, we predicted that the protected suitable area for pancake tortoises will expand from 114,969.40 km2 (in 2050) to 123,112.71 km2 (2070) in RCP 4.5 and 104,437.30 km2 (in 2050) to 117,672.83 km2 (in 2070) in RCP 6.0 (Table 1), given the current Protected Area Network. However, we predicted that the protected suitable habitat of pancake tortoises will continue to be smaller in the future (RCP 4.5: 31.39% in 2050 and 30.69% in 2070; RCP 6.0: 30.41% in 2050 and 31.50% in 2070; Table 1).

We identified Kenya, Tanzania, Ethiopia and Angola as the countries that maintain the most stable habitat for pancake tortoises over time (Table 2). However, the highest stability occurs within Kenya, Tanzania and Ethiopia (Fig 3), with only Kenya having a highly stable habitat inside the protected areas (Table 2). We predicted that the stable habitats for pancake tortoises within the current Protected Areas Network will continue to be smaller than those of habitats in unprotected areas (percentage of stable habitat present in protected areas: less stable [33.08%], average stability [27.97%] and highly stable [14.87%, present in Kenya only]; Table 2).

Table 2. Potential climatic stable areas/habitats (in km2) for the pancake tortoise per each country present in the Zambezian and Somalia biogeographical regions.

Country Current suitable habitat (km2) Less stable Mid stable Highly stable
Total (km2) % of protected stable habitat Total (km2) % of protected stable habitat Total (km2) % of protected stable habitat
Kenya 55,401.76 92,198.02 35.31 57,252.66 30.89 16,066.27 15.42
Tanzania 27,489.20 50,153.64 35.54 16,299.88 26.19 255.98 0
Zambia 0 0 0 0 0 0 0
Mozambique 0 0 0 0 0 0 0
Malawi 0 0 0 0 0 0 0
Zimbabwe 0 0 0 0 0 0 0
Angola 3,219.98 2,174.18 0 83.67 0 0 0
Namibia 0 0 0 0 0 0 0
Botswana 0 0 0 0 0 0 0
Somalia 0 0 0 0 0 0 0
Ethiopia 4,557.81 19,904.86 20.12 8,513.63 12.01 340.53 0
Democratic Republic of Congo 0 0 0 0 0 0 0
Burundi 0 0 0 0 0 0 0
Rwanda 0 0 0 0 0 0 0
Uganda 0 0 0 0 0 0 0
South Africa 0 0 0 0 0 0 0
TOTAL 90,668.75 164,430.70 30.08 82,149.83 27.97 16,662.78 14.87
Open in a new tab

Climatic stability was inferred by extracting all values greater than zero from the raster and obtaining the 90th, 95th and 99th quantiles, returning low, mid and high stability, respectively. Area was calculated with a suitability weighted approach. The variation between current suitable habitat and the stable areas is influenced by the weight set in cells where for those of stable areas weight is one, while cells in current suitable area weight is less than one. ‘Total’ is the total stable area and ‘% of protected stable habitat’ is the percentage of protected area that overlaps with the species stable habitat.

Fig 3. Potential climatic stable areas for the pancake tortoise in the Zambezian and Somalia biogeographical regions.

Fig 3

Open in a new tab

(A) Location of stable areas in Africa. (B) Stable areas in Angola. (C) Stable areas in Tanzania, Kenya and Ethiopia. The stable areas were obtained by considering three thresholds from the sum of the five normalized climatic scenarios (current, RCP 4.5/2050, RCP 4.5/2070, RCP 6.0/2050 and RCP 6.0/2070). The brighter red colour indicates the more stable site through time. Background image was accessed from Natural Earth (public domain): http://www.naturalearthdata.com/.

Discussion

We predicted the climatic suitable habitat for pancake tortoises in the Zambezian and Somalian biogeographical regions in the current and future scenarios. The six bioclimatic variables indicated that pancake tortoise occurrence can either increase or decrease until the optimum condition has been reached (S1 Fig). This might be due to the fact that reptiles need the optimum condition for laying and hatching their eggs [67]. Therefore, climatic changes can significantly affect reptile’s reproductive success [68]. Although isothemality (BIO 3) did have the highest contribution in predicting climatic suitable habitat for pancake tortoise in the Zambezian and Somalian biogeographical regions (S2 Fig), this variable was not selected by Bombi et al. [41] when modelling the distribution of 16 species of Testudinidae in Africa. However, two of our variables (mean temperature of the wettest quarter (BIO 8) and precipitation seasonality (BIO 15)) did match with the one selected by Bombi et al. [41]. These variations could be due to differences in geographical range considered, species involved, number of occurrences and modelling approach.

We predicted that the suitable climatic habitat for pancake tortoises would be less discontinuously scattered in the Zambezian and Somalian biogeographical regions in the future than in current climatic scenarios (Fig 2). The disjointed distribution of pancake tortoises was also observed in the countries in which they currently exist naturally, which are Tanzania [17, 20] and Kenya [18, 20]. We further predicted that the distributional range of pancake tortoises would expand in the future (Fig 2 and Table 1). The expansion of the future distributional ranges of reptiles has also been recorded by Houniet et al. [69] for Bradypodion occidentale, González-Fernández et al. [70] for Thamnophis melanogaster, Fathinia et al. [71] for Pseudocerastes urarachnoides and Sousa-Guedes et al. [72] for 13 different reptile species.

Apart from area expansion, our model also predicted an increase in the number of climatically suitable habitats in countries in which pancake tortoises do not exist naturally from the current two to eight future countries, with Angola being isolated in the far west of the region (Fig 2 and Table 1). This scenario of isolation of pancake tortoise populations has been also recorded in Tanzania [50] and Kenya [18], where the species exists naturally. The absence of pancake tortoise in the countries which we predicted to have the climatically suitable habitats could be due to their behaviour of being non-migrant [18–20]. On the other hand, Malonza [18] has suggested that the absence of pancake tortoises in potential habitats is mainly due to elevation, with species occurring from 500–1,800 m above sea level [20]. Pancake tortoises do not occur in some climatically suitable habitats (Fig 2). The prevalence of non-Precambrian rock types between areas where we predict suitable habitat may preclude occupancy by pancake tortoises, as they prefer areas dominated by Precambrian rocks [13].

We did not predict the existence of climatic suitable habitats for pancake tortoise under current and future scenarios in Zambia (Fig 2 and Table 1) although the species have been reported to occur in the country [13]. This could mean that pancake tortoise recorded in Zambia were the result of the international animal trade [19], as a result of which animals from East Africa were exported illegally from the country [20]. However, this argument would require a genetic analysis for confirmation. Conversely, Zambia could be located at the limit of the climatically suitable niche and thus has low climatic suitability, which, when applied at a threshold, turns into an absence.

We predicted that Tanzania, Kenya, Ethiopia and Angola (Fig 3 and Table 1) will have climatically stable habitats over time. As pancake tortoises have not yet been recorded in Ethiopia and Angola, these areas could hold potential for the translocation and introduction of the species. We recommend robust habitat suitability studies of these countries and further quantification of occupancy status given the species apparently occupies suitable habitats in the nearby Zambezian and Somalian biogeographical regions [13, 37].

Protected areas are critical tools for biodiversity conservation [33], yet the African Protected Areas Network offers inconsistent protection to tortoise species [41]. In the Zambezian and Somalian biogeographical regions, only 32.37% of the current climatically suitable area for pancake tortoises fall within protected areas, and this percentage is predicted to decline in the future to 30.41% - 31.50% (Table 1). Additionally, from 66.92% - 85.13% of the stable climatic habitat is predicted to be outside of protected areas. Our results are inconsistent with Bombi et al. [41] argument, who mentioned that the established protected areas in East Africa for wildlife conservation offer sufficient presentation for tortoises. On the other hand, we agree with Bombi et al. [41] findings on pancake tortoise where they found that across the entire range of the species only 22.60% of its range are protected [41]. In Kenya, only 5.00% of the pancake tortoise population is protected, while in Zambia, the species does not occur within the Protected Area Network [20, 21]. Based on occurrence points we collected, in Tanzania, only four out of 22 national parks are occupied by pancake tortoises. The pancake tortoise’s suitable habitat is largely unprotected in both the current and the future scenarios, likely increasing the risk of overexploitation and exacerbating negative effects of habitat destruction as in Tanzania [26], Kenya [18, 20] and Zambia [20, 21]. Likewise, ectoparasite prevalence is higher outside of the protected area [73], potentially increasing risk to the species.

Management implications

Because the current natural range for pancake tortoise does not include some of our current and future predictions for climatically suitable habitats, we recommend future studies be conducted in areas where pancake tortoises do not exist to confirm the absence of the species. As White [45] and Chansa and Wagner [21] have pointed out, pancake tortoises could exist in the entire Zambezian and Somalian biogeographical regions, provided that suitable habitat is present; therefore, confirmatory studies on the existence of the species in the climatically suitable habitats are essential for conservation planning for the species. However, we caution that the existence of pancake tortoises is not solely dependent on the presence of climatically suitable habitats, as Malonza [18] has confirmed the non-presence of pancake tortoises in typical habitats for the species in Kenya. Furthermore, the available suitable and stable habitats outside of the current range could be used as baseline areas for the translocation and introduction of the species where necessary. Therefore, we support the IUCN [27] and Bellis et al. [29], who have suggested the importance of conducting SDMs before translocation and species introduction/re-introduction. Our model did not predict the existence of climatically suitable habitats for pancake tortoises in Zambia (Fig 2). Therefore, we recommend the maximization of conservation efforts in Zambia in order to maintain the recorded pancake tortoise populations, since they seem to be highly threatened as they are all located outside of protected lands [21].

Furthermore, the presence of a large proportion of the climatically suitable habitat for pancake tortoises outside of protected areas could imply the need for more conservation efforts outside the protected range. These efforts might include the establishment of new protected areas aimed at biodiversity conservation to include suitable habitats for pancake tortoises and therefore minimize anthropogenic impacts on the species [18, 43]. Since the current increase in Protected Area Network have rarely strategically considered global biodiversity maximization [33], establishing protected areas within species suitable habitats could be one of the strategies for protecting global biodiversity. This will also help to reduce the extinction risk for different species under climate change [74].

Area protection, management of international animal trade, species recovery plans and conservation awareness are some conservation actions prioritized by IUCN to save the species under current situation [20]. In addition, majority of the of our predicted suitable habitats and current and future climatic scenarios do fall under highest spatial prioritization for land conservation to minimize extinction risk under climate change in the Afrotropics [74]. Nonetheless, environmental and social context will decide which option is better [75], therefore all conservation stakeholders including local, regional and international organizations, scientists, practitioners and the general public should join forces to save the global biodiversity.

Conclusion and study limitations

We predict expansion of suitable habitats for pancake tortoises in the future, which may conserve populations of this critically endangered reptile. Importantly, the largest proportion of suitable habitats is outside of the current Protected Area Network, therefore, we suggest the pancake tortoise be upgraded in its listing status from CITES Appendix II to Appendix I.

Because our results were largely based on use of climatic variables, our findings should not be treated as ready-made for on-the-ground application but could be used as one of many tools to help in conservation planning of pancake tortoises. Our decision to use primarily climate variables was because they drive most of the species’ distribution [76]. Although Giannini et al. [77], de Araújo [9] and Palacio and Girini [77] have pointed out that the inclusion of biotic factors significantly improves SDMs, we were unable to obtain these data for our study. We recommend that future studies to consider the inclusion of pre-Cambrian rock (as it provides a preferred habitat for pancake tortoises), the international pet trade, land-use changes and ecological interactions as predictor variables. However, in the current situation, it is difficult to obtain these data ready-made for SDM, especially for future scenarios. Furthermore, the application of SDM can be limited by; first, assuming there is a balance between environmental changes and spatial distribution of the species [2, 78, 79], second, the representability of distributional spectrum of the species in relation to the occurrence data used in modelling [78–80], third, data adequacy and resolution applied during modelling, fourth, model performance and fifth, the reliability of climatic future predictions [79, 81].

All in all, our study provides a solid foundation for future development of conservation measures aimed at protecting populations of the critically endangered pancake tortoise.

Supporting information

S1 Fig. Response curves from the ensemble models to the six selected bioclimatic variables.

Response curves were fitted through locally estimated scatterplot smoothing (LOESS). Grey background shows a scale from y-axis which is replicated to every graph.

(DOCX)

Click here for additional data file. (236.9KB, docx)
S2 Fig. Variable importance for six less correlated climatic variables of the ensemble species distribution model.

BIO 8 = mean temperature of the wettest quarter, BIO 3 = the isothermality, BIO 2 = mean diurnal range, BIO 15 = precipitation seasonality, BIO 14 = precipitation of the driest month and BIO 13 = precipitation of the wettest month.

(DOCX)

Click here for additional data file. (103.3KB, docx)
S1 Table. Ten Global Circulation Models (GCMs) used in our study.

(DOCX)

Click here for additional data file. (20.1KB, docx)

Acknowledgments

We thank Tanzania Wildlife Research Institute (TAWIRI) and Commission for Science and Technology (COSTECH) for granting permit and Tanzania National Park (TANAPA) for giving free access into Tarangire National Park. We also thank Deo Tarimo for his help in obtaining pancake GPS locations from Mkomazi National Park and some from Tarangire National Park. Furthermore, we appreciate all the logistical support which were provided by the College of African Wildlife Management, Mweka during fieldwork. We also thank Nicola van Wilgen and the other anonymous reviewer for their comments which significantly improved this manuscript.

Data Availability

According to IUCN the pancake tortoise is critically endangered and threatened by international pet trade therefore the location data collected by authors is not publicly available. However, these data are available from the College of African Wildlife Management upon reasonable request. Please, contact the Head of Research and Consultancy Department through mweka@mwekawildlife.ac.tz if you need these data. Other secondary data used in this study are available from the original source as cited in the method section.

Funding Statement

The author(s) received no specific funding for this work.

References

  • 1.Franklin J. Species distribution models in conservation biogeography: Developments and challenges. Divers Distrib. 2013;19: 1217–1223. 10.1111/ddi.12125 [DOI] [Google Scholar]
  • 2.Guisan A, Thuiller W. Predicting species distribution: Offering more than simple habitat models. Ecol Lett. 2005;8: 993–1009. 10.1111/j.1461-0248.2005.00792.x [DOI] [PubMed] [Google Scholar]
  • 3.Báez JC, Estrada A, Torreblanca D, Real R. Predicting the distribution of cryptic species: The case of the spur-thighed tortoise in Andalusia (southern Iberian Peninsula). Biodivers Conserv. 2012;21: 65–78. 10.1007/s10531-011-0164-3 [DOI] [Google Scholar]
  • 4.Mammola S, Leroy B. Applying species distribution models to caves and other subterranean habitats. Ecography (Cop). 2018;41: 01–14. 10.1111/ecog.03464 [DOI] [Google Scholar]
  • 5.Real R, Romero D, Olivero J, Estrada A, Márquez AL. Estimating How Inflated or Obscured Effects of Climate Affect Forecasted Species Distribution. PLoS One. 2013;8 10.1371/journal.pone.0053646 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Martínez-Freiría F, Argaz H, Fahd S, Brito JC. Climate change is predicted to negatively influence Moroccan endemic reptile richness. Implications for conservation in protected areas. Naturwissenschaften. 2013;100: 877–889. 10.1007/s00114-013-1088-4 [DOI] [PubMed] [Google Scholar]
  • 7.Burrows MT, Schoeman DS, Richardson AJ, Molinos JG, Hoffmann A, Buckley LB, et al. Geographical limits to species-range shifts are suggested by climate velocity. Nature. 2014;507: 492–495. 10.1038/nature12976 [DOI] [PubMed] [Google Scholar]
  • 8.Ceballos G, Ehrlich PR, Barnosky AD, García A, Pringle RM, Palmer TM. Accelerated modern human-induced species losses: Entering the sixth mass extinction. Sci Adv. 2015;1: 9–13. 10.1126/sciadv.1400253 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Garcia RA, Cabeza M, Rahbek C, Araújo MB. Multiple dimensions of climate change and their implications for biodiversity. Science (80-). 2014;344: 1–10. 10.1126/science.1247579 [DOI] [PubMed] [Google Scholar]
  • 10.McHenry J, Welch H, Lester SE, Saba V. Projecting marine species range shifts from only temperature can mask climate vulnerability. Glob Chang Biol. 2019;00: 1–14. 10.1111/gcb.14828 [DOI] [PubMed] [Google Scholar]
  • 11.Srinivasulu A, Srinivasulu C. All that glitters is not gold: A projected distribution of the endemic Indian Golden Gecko Calodactyflodes aureus (Reptilia: Squamata: Gekkonfidae) indicates a major range shrinkage due to future climate change. J Threat Taxa. 2016;8: 8883–8892. 10.11609/jot.2723.8.6.8883-889 [DOI] [Google Scholar]
  • 12.Ahmadi M, Hemami MR, Kaboli M, Malekian M, Zimmermann NE. Extinction risks of a Mediterranean neo-endemism complex of mountain vipers triggered by climate change. Sci Rep. 2019;9: 1–12. 10.1038/s41598-018-37186-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Deharveng L, Bedos A. Biodiversity in the tropics. 3rd ed. Encyclopedia of Caves. 3rd ed. Elsevier Inc.; 2019. pp. 146–162. 10.1016/b978-0-12-814124-3.00040-6 [DOI]
  • 14.Rajpar MN. Tropical Forests Are An Ideal Habitat for Wide Array of Wildlife Species. Tropical Forests—New Edition. 2018. p. 37 10.5772/intechopen.73315 [DOI] [Google Scholar]
  • 15.Meng H, Carr J, Beraducci J, Bowles P, Branch WR, Capitani C, et al. Tanzania’s reptile biodiversity: Distribution, threats and climate change vulnerability. Biological Conservation. 2016. 10.1016/j.biocon.2016.04.008 [DOI] [Google Scholar]
  • 16.Araújo MB, Alagador D, Cabeza M, Nogués-Bravo D, Thuiller W. Climate change threatens European conservation areas. Ecol Lett. 2011;14: 484–492. 10.1111/j.1461-0248.2011.01610.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Raphael B, Klemens M, Moehlman P, Dierenfeld E, Karesh W. Blood values in free-ranging pancake tortoises (Malacochersus tornieri). J Zoo Wildl Med. 1995;25: 63–67. [Google Scholar]
  • 18.Malonza PK. Ecology and Distribution of the Pancake Tortoise, Malacochersus tornieri in Kenya. J East African Nat Hist. 2003;92: 81–96. 10.2982/0012-8317(2003)92[81:eadotp]2.0.co;2 [DOI] [Google Scholar]
  • 19.Mwaya RT, Moll D, Malonza PK, Ngwaya JM. Malacochersus tornieri (Siebenrock 1903)–Pancake Tortoise, Tornier’s Tortoise, Soft-shelled Tortoise, Crevice Tortoise, Kobe Ya Mawe, Kobe Kama Chapati. Rhodin AGJ, Iverson JB, van Dijk PP, Stanford CB, Goode EV, Buhlmann KA, et al., editors. Chelonian Research Monographs. 2018. 10.3854/crm.5.107.tornieri.v1.2018 [DOI]
  • 20.Mwaya RT, Malonza PK, Ngwava JM, Moll D, Schmidt FA, Rhodin AGJN. Malacochersus tornieri. The IUCN Red List of Threatened Species 2019. IUCN Red List Threat. 2019. https://doi.org/Malacochersus tornieri
  • 21.Chansa W, Wagner P. On the status of Malacochersus tornieri (Siebenrock, 1903) in Zambia. Salamandra Bonn. 2006;42: 187. [Google Scholar]
  • 22.Linder HP, de Klerk HM, Born J, Burgess ND, Fjeldså J, Rahbek C. The partitioning of Africa: Statistically defined biogeographical regions in sub-Saharan Africa. J Biogeogr. 2012;39: 1189–1205. 10.1111/j.1365-2699.2012.02728.x [DOI] [Google Scholar]
  • 23.Thulin M. Aspects of disjunct distributions and endemism in the arid parts of the Horn of Africa, particularly Somalia. Proceedings of the 13th Plenary Meeting, AETFAT, Zomba, Malawi. 1994. pp. 1105–1119.
  • 24.Burgess N, Hales JD, Underwood E, Dinerstein E, Olson D, Itoua I, et al. Terrestrial Ecoregions of Africa and Madagascar A Conservation Assessment. Island Press; 2004. [Google Scholar]
  • 25.Moll D, Klemens MW. Ecological characteristics of the pancake tortoise, Malacochersus tornieri, in Tanzania. Chelonian Conserv Biol. 1996;2: 26–35. Available: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=Moll+and+Klemens+%281996%29&btnG=
  • 26.Klemens MW, Moll D. An assessment of the effects of commercial exploitation on the pancake tortoise, Malacochersus tornieri. Tanzania. Chelonian Conserv Biol. 1995;1: 197–206. [Google Scholar]
  • 27.IUCN. Guidelines for Reintroductions and Other Conservation Translocations. Version 1.0. Gland, Switzerland: IUCN; 2013.
  • 28.White TH, de Melo Barros Y, Develey PF, Llerandi-Román IC, Monsegur-Rivera OA, Trujillo-Pinto AM. Improving reintroduction planning and implementation through quantitative SWOT analysis. J Nat Conserv. 2015;28: 149–159. 10.1016/j.jnc.2015.10.002 [DOI] [Google Scholar]
  • 29.Bellis J, Bourke D, Maschinski J, Heineman K, Dalrymple S. Climate suitability as a predictor of conservation translocation failure. Conserv Biol. 2020. 10.1111/cobi.13518 [DOI] [PubMed] [Google Scholar]
  • 30.Coetzee BWT, Robertson MP, Erasmus BFN, van Rensburg BJ, Thuiller W. Ensemble models predict important bird areas in southern Africa will become less effective for conserving endemic birds under climate change. Glob Ecol Biogeogr. 2009;18: 701–710. 10.1111/j.1466-8238.2009.00485.x [DOI] [Google Scholar]
  • 31.Esser HJ, Mögling R, Cleton NB, Van Der Jeugd H, Sprong H, Stroo A, et al. Risk factors associated with sustained circulation of six zoonotic arboviruses: A systematic review for selection of surveillance sites in non-endemic areas. Parasites and Vectors. 2019;12: 1–17. 10.1186/s13071-018-3256-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Craigie ID, Baillie JEM, Balmford A, Carbone C, Collen B, Green RE, et al. Large mammal population declines in Africa’s protected areas. Biol Conserv. 2010;143: 2221–2228. 10.1016/j.biocon.2010.06.007 [DOI] [Google Scholar]
  • 33.Rodrigues ASL, Akçakaya HR, Andelman SJ, Bakarr MI, Boitani L, Brooks TM, et al. Global Gap Analysis: Priority Regions for Expanding the Global Protected-Area Network. Bioscience. 2004;54: 1092–1100. 10.1641/0006-3568(2004)054[1092:ggaprf]2.0.co;2 [DOI] [Google Scholar]
  • 34.Warren R, VanDerWal J, Price J, Welbergen JA, Atkinson I, Ramirez-Villegas J, et al. Quantifying the benefit of early climate change mitigation in avoiding biodiversity loss. Nat Clim Chang. 2013;3: 678–682. 10.1038/nclim ate1887 [DOI] [Google Scholar]
  • 35.Beaumont LJ, Graham E, Duursma DE, Wilson PD, Cabrelli A, Baumgartner JB, et al. Which species distribution models are more (or less) likely to project broad-scale, climate-induced shifts in species ranges? Ecol Modell. 2016;342: 135–146. 10.1016/j.ecolmodel.2016.10.004 [DOI] [Google Scholar]
  • 36.Krechemer F da S, Marchioro CA. Past, present, and future distributions of bumble bees in South America: Identifying priority species and areas for conservation. Journal of Applied Ecology. 2020. 10.1111/1365-2664.13650 [DOI] [Google Scholar]
  • 37.Araújo MB, Williams PH, Fuller RJ. Dynamics of extinction and the selection of nature reserves. Hungarian Q. 2008;49: 1971–1980. 10.1098/rspb.2002.2121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Terribile LC, Lima-Ribeiro MS, Araújo MB, Bizão N, Collevatti RG, Dobrovolski R, et al. Areas of climate stability of species ranges in the Brazilian cerrado: Disentangling uncertainties through time. Nat a Conserv. 2012;10: 152–159. 10.4322/natcon.2012.025 [DOI] [Google Scholar]
  • 39.Khan S, Nathand A, Das A. The distribution of the elongated tortoise (Indotestudo elongata)on the indian subcontinent: Implications for conservation and management. Herpetol Conserv Biol. 2020;15: 212–227. [Google Scholar]
  • 40.Muñoz AR, Márquez AL, Real R. Updating Known Distribution Models for Forecasting Climate Change Impact on Endangered Species. PLoS One. 2013;8: 1–9. 10.1371/journal.pone.0065462 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Bombi P, D’Amen M, Luiselli L. From Continental Priorities to Local Conservation: A Multi-Level Analysis for African Tortoises. PLoS One. 2013;8: 1–9. 10.1371/journal.pone.0077093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Villero D, Pla M, Camps D, Ruiz-Olmo J, Brotons L. Integrating species distribution modelling into decision-making to inform conservation actions. Biodivers Conserv. 2017;26: 251–271. 10.1007/s10531-016-1243-2 [DOI] [Google Scholar]
  • 43.Church RL, Stoms DM, Davis FW. Reserve selection as a maximal covering location problem. Biol Conserv. 1996;76: 105–112. [Google Scholar]
  • 44.UNEP-WCMC, IUCN. Protected Planet: The World Database on Protected Areas (WDPA), Cambridge, UK: UNEP-WCMC and IUCN. In: 2020 [Internet]. 2020. Available: www.protectedplanet.net
  • 45.White F. The vegetation of Africa. A descriptive memoir to accompany the UNESCO-AETFAT-UNSO vegetation map of Africa, UNESCO, Paris. 1983; 356.
  • 46.Coe MJ, Skinner JD. Connections, disjunctions and endemism in the eastern and southern african mammal faunas. Trans R Soc South Africa. 1993;48: 233–255. 10.1080/00359199309520273 [DOI] [Google Scholar]
  • 47.Linder HP, Lovett J, Mutke JM, Barthlott W, Jürgens N, Rebelo T, et al. A numerical re-evaluation of the sub-Saharan phytochoria of mainland Africa. Biol Skr. 2005;55: 229–252. [Google Scholar]
  • 48.Kabigumila J. Morphometrics of the Pancake tortoise (Malacochersus tornieri) in Tanzania. Tanzania J Sci. 2002;28: 34–46. [Google Scholar]
  • 49.Wood RC, MacKay A. The distribution and status of the pancake tortoises, Malacochersus tornieri, Kenya. Conservation, Restorafion and Management of Tortoises and Turtles-An Internafional Conference. 1997.
  • 50.Spawls S, Howell K, Drewes R, Ashe J. A Field Guide to the Reptiles of East Africa—Kenya, Tanzania, Uganda, Rwanda and Burundi. San Diego, San Francisco, New York, Boston, London: AP Natural World; 2002.
  • 51.Mwaya RT. The floristic composition of the habitat of Malacochersus tornieri at a hill in Tarangire National Park, Tanzania. Salamandra. 2009;45: 115–118. [Google Scholar]
  • 52.Kyalo NS. Conservation, Management and Control of Trade in Pancake Tortoise Malacochersus tornieri (Siebenrock, 1903) in Kenya: The Non-Detriment Finding Studies Case Study. International Expert Workshop on CITES Non-Detriment Findings. Mexico,; 2008.
  • 53.Chamberlain S, Barve V, Mcglinn D, Oldoni D, Desmet P, Geffert L, et al. rgbif: Interface to the Global Biodiversity Information Facility API. 2020. Available: https://cran.r-project.org/package=rgbif
  • 54.Chamberlain S. rvertnet: Search “Vertnet”, a “Database” of Vertebrate Specimen Records. 2018. Available: https://cran.r-project.org/package=rvertnet
  • 55.Iverson JB, Kiester AR, Hughes LE, Kimerling AJ. The EMYSystem world turtle database. In: http://emys.geo.orst.edu [Internet]. 2003 [cited 27 Apr 2020]. Available: http://emys.geo.orst.edu
  • 56.Zacarias DA. Habitat analysis and population structure of the pancake tortoise (Malacochersus tornieri) in the Tarangire Ecosystem: an approach using GIS (Unpublished). College of African Wildlife Management, Mweka. 2007. Available: https://www.academia.edu/2563124/Aplicação_dos_SIG_e_análise_estatística_no_maneio_da_fauna_bravia_o_caso_das_tartarugas_de_carapaça_mole_Malacochersus_tornieri_
  • 57.Karger DN, Conrad O, Böhner J, Kawohl T, Kreft H, Soria-Auza RW, et al. Climatologies at high resolution for the earth’s land surface areas. Sci Data. 2017;4: 1–20. 10.1038/sdata.2017.122 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Thomson AM, Calvin K V, Smith SJ, Kyle GP, Volke A, Patel P, et al. RCP4.5: A pathway for stabilization of radiative forcing by 2100. Clim Change. 2011;109: 77–94. 10.1007/s10584-011-0151-4 [DOI] [Google Scholar]
  • 59.Masui T, Matsumoto K, Hijioka Y, Kinoshita T, Nozawa T, Ishiwatari S, et al. An emission pathway for stabilization at 6 Wm-2 radiative forcing. Clim Change. 2011;109: 59–76. 10.1007/s10584-011-0150-5 [DOI] [Google Scholar]
  • 60.IPCC. Climate Change 2014: Synthesis Report. Contribution of Working Group I, II and III to the Fifth Assessment Report of the Intergovenmental Panel on Climate Change [Core Writing Team, P.K. Pachauri and L.A. Meyer (eds)] IPcc, geneva, Switzerland, 151 pp. 2014.
  • 61.Sanderson BM, Knutti R, Caldwell P. A representative democracy to reduce interdependency in a multimodel ensemble. J Clim. 2015;28: 5171–5194. 10.1175/JCLI-D-14-00362.1 [DOI] [Google Scholar]
  • 62.Oh SG, Suh MS. Comparison of projection skills of deterministic ensemble methods using pseudo-simulation data generated from multivariate Gaussian distribution. Theor Appl Climatol. 2017;129: 243–262. 10.1007/s00704-016-1782-1 [DOI] [Google Scholar]
  • 63.Naimi B, Hamm NAS, Groen TA, Skidmore AK, Toxopeus AG. Where is positional uncertainty a problem for species distribution modelling? Ecography (Cop). 2014;37: 191–203. 10.1111/j.1600-0587.2013.00205.x [DOI] [Google Scholar]
  • 64.R Core Team. R: A language and environment for statistical computing. Vienna, Austria. 2019. Available: https://www.r-project.org/
  • 65.Naimi B, Araújo MB. Sdm: A reproducible and extensible R platform for species distribution modelling. Ecography (Cop). 2016;39: 368–375. 10.1111/ecog.01881 [DOI] [Google Scholar]
  • 66.Barbet-Massin M, Jiguet F, Albert CH, Thuiller W. Selecting pseudo-absences for species distribution models: How, where and how many? Methods Ecol Evol. 2012;3: 327–338. 10.1111/j.2041-210X.2011.00172.x [DOI] [Google Scholar]
  • 67.Ackerman RA, Lott DB. Thermal, hydric and respiratory climate of nests. Deeming DC, ed Reptilian Incubation: Environment, Evolution, and Behaviour. Nottingham: Nottingham University Press; 2004. pp. 15–43.
  • 68.Tomillo PS, Saba VS, Blanco GS, Stock CA, Paladino F V., Spotila JR. Climate driven egg and hatchling mortality threatens survival of Eastern Pacific leatherback turtles. PLoS One. 2012;7: 1–7. 10.1371/journal.pone.0037602 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Houniet DT, Thuiller W, Tolley KA. Potential effects of predicted climate change on the endemic south african dwarf chameleons, bradypodion. African J Herpetol. 2009;58: 28–35. 10.1080/21564574.2009.9635577 [DOI] [Google Scholar]
  • 70.González-Fernández A, Manjarrez J, García-Vázquez U, D’Addario M, Sunny A. Present and future ecological niche modeling of garter snake species from the Trans-Mexican Volcanic Belt. PeerJ. 2018; 1–20. 10.7717/peerj.4618 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.Fathinia B, Rödder D, Rastegar-pouyani N, Hosseinzadeh MS, Kazemi SM. Zoology in the Middle East The past, current and future habitat range of the Spider-tailed Viper, Pseudocerastes urarachnoides (Serpentes: Viperidae) in western Iran and eastern Iraq as revealed by habitat modelling. Zool Middle East. 2020; 1–9. 10.1080/09397140.2020.1757910 [DOI] [Google Scholar]
  • 72.Sousa-Guedes D, Arenas-Castro S, Sillero N. Ecological niche models reveal climate change effect on biogeographical regions: The Iberian Peninsula as a case study. Climate. 2020;8: 1–18. 10.3390/cli8030042 [DOI] [Google Scholar]
  • 73.Mwaya RT, Mremi R, Eustace A, Ndibalema V. Prevalence of ticks (Acari: Ixodidae) parasitism on Pancake Tortoises, Malacochersus tornieri (Testudinidae), is lower inside than outside Tarangire National Park, Tanzania (In press). Chelonia Conserv Biol. Forthcoming.
  • 74.Hannah L, Roehrdanz PR, Marquet PA, Enquist BJ, Midgley G, Foden W, et al. 30% Land Conservation and Climate Action Reduces Tropical Extinction Risk By More Than 50%. Ecography (Cop). 2020;43: 1–11. 10.1111/ecog.05166 [DOI] [Google Scholar]
  • 75.Brown JH, Curtin CJ, Braithwaite RW. Management of the Semi-Natural Matrix In Bradshaw G.A., Marquet P.A. (eds) How Landscapes Change. Ecological Studies (Analysis and Synthesis). Springer Berlin Heidelberg; 2003. [Google Scholar]
  • 76.Salas EAL, Seamster VA, Harings NM, Boykin KG, Alvarez G, Dixon KW. Projected future bioclimate-envelope suitability for reptile and amphibian species of concern in South Central USA. Herpetol Conserv Biol. 2017;12: 522–547. [Google Scholar]
  • 77.Palacio FX, Girini JM. Biotic interactions in species distribution models enhance model performance and shed light on natural history of rare birds: a case study using the straight-billed reedhaunter Limnoctites rectirostris. J Avian Biol. 2018;49: 1–11. 10.1111/jav.01743 [DOI] [Google Scholar]
  • 78.Barry S, Elith J. Error and uncertainty in habitat models. J Appl Ecol. 2006;43: 413–423. 10.1111/j.1365-2664.2006.01136.x [DOI] [Google Scholar]
  • 79.Martínez-Freiría F, Tarroso P, Rebelo H, Brito JC. Contemporary niche contraction affects climate change predictions for elephants and giraffes. Divers Distrib. 2016;22: 432–444. 10.1111/ddi.12406 [DOI] [Google Scholar]
  • 80.Araújo MB, Guisan A. Five (or so) challenges for species distribution modelling. J Biogeogr. 2006;33: 1677–1688. 10.1111/j.1365-2699.2006.01584.x [DOI] [Google Scholar]
  • 81.Franklin J. Moving beyond static species distribution models in support of conservation biogeography. Divers Distrib. 2010;16: 321–330. 10.1111/j.1472-4642.2010.00641.x [DOI] [Google Scholar]

Decision Letter 0

Stephanie S Romanach

6 Oct 2020

PONE-D-20-25666

Protected areas network is not adequate to protect a critically endangered East Africa Chelonian: Modelling distribution of pancake tortoise, Malacochersus tornieri under current and future climates

PLOS ONE

Dear Dr. Eustace,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Your paper has the potential to make a useful contribution to the literature, but as presented, there are too many gaps for the reader follow your methods through to your conclusions. It is important that you provide enough information about your modeling approach so that the reader can understand how you worked with the species and climate data and that you explicitly describe the outcomes of the critical modeling steps along the way. I have a number of concerns with the modeling that could be addressed during revision, but this will likely require a major overhaul including re-doing the modeling and presenting the information with greater detail and clarity. Both reviewers provide excellent, detailed suggestions to help if you choose to revise. Please note that revision does not guarantee acceptance.

Please submit your revised manuscript by Nov 20 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Stephanie S. Romanach, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Methods section, please provide additional information regarding the permits you obtained for the work. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. We note that Figures 1-3 in your submission contain map images which may be copyrighted.

All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (a) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (b) remove the figures from your submission:

a. You may seek permission from the original copyright holder of Figures 1-3 to publish the content specifically under the CC BY 4.0 license. 

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

5. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper investigates the suitability of future climate space for the critically endangered pancake tortoise, with a view to informing translocations and/or population management that will maximize sustainability of future populations in the face of climate change and other stressors (this species is harvested for the international pet trade). The study finds that similar to the current situation, most future suitable habitat is outside of protected areas. Further they identify additional sites where the species is not currently known to occur that could serve to bolster the population.

While this topic is a very important and the investigation a needed contribution in terms of critically endangered species management, there were some fundamental aspects that relate to species distribution modelling and species conservation that have not been covered/discussed. I have made a number of specific comments, but in general, think that the following need to be addressed to strengthen the paper:

1. More rigorous assessment of the climatic variables selected. How do these relate to the species’ life history and what mechanism of impact are they expected to have on population viability. Also, in terms of the model, what was the relationship with these variables and tortoise occurrence, and what was the relative effect size of each? Are there any potentially informative variables that were not included?

Related to the above, are there any other non-climatic variables for which information exist that might be added? The authors suggest that rock substrate is important. The authors also mention a previously published distribution model. How do the results of these models compare?

2. The tables and figures are useful, but could their utility could definitely be enhanced further with the addition of details. For example, protected areas are not shown and in some instances country borders are also omitted. For the tables, it would be useful to compare current to predicted future situations.

3. The extent to which the current distribution has been impacted by overharvesting is not clear. The current models do not highlight any suitable habitat in Zambia, but only one occurrence was included from the Zambian border. Is it likely that the species used to be more widespread in Zambia? If so, the discussion should rather conclude that lack of current records from the area preclude investigation of habitat suitability in Zambia

4. Some more general discussion/framing of species management under climate change would be useful. There are so many species and large change is expected. Is moving species to far off locations really viable? How are these interventions prioritized? And by whom? The paper by Hannah (2020) details some of the most important corridors for conservation in Africa. There is definite overlap with the areas identified in this study and would be worth mentioning.

Specific comments:

Line 118 – there is no mention (later) of whether your current distribution compares well to this previous model. What the major findings and reasons for any potential differences?

Line 170, in what way would aquaculture threaten the tortoise? I assume this is a grouped categorization by the IUCN, but perhaps leave off the aquaculture part or include in inverted commas?

Lines 179-182 combine sentences to reduce repetition

Fig. 1 Caption: I suggest reordering and rewriting along the following lines to make the buffer area clear from the outset: Current natural occurrence of pancake tortoise (Malacochersus tornieri), based on data obtained in this study, with a 1-degree wide buffer around each presence record, at (A) continental scale and (C) regional scale. (B) shows the Zambezian and Somalia biogeographical region.

Line 233: is some measure of maximum temperature not important from a species perspective under climate change? How close to thermal maxima is the current study area likely to get? How do the variables selected compare to those used in previous studies on tortoises?

242-251 I am not quite sure that I understand the different algorithms – are you saying that you had 5 methods of accounting for pseudo-absences?

Section starting line 241 “Species distribution modelling”– it is not clear at which point the future projections are brought into the modelling. These are referred to as part of a general analysis in line 269, but there is no explicit separation of how current versus future estimates of range were produced.

Line 290 – with regard to the potential habitat in Angola. Was this also found to be the case when pseudo-absences were only generated from within the boundary of the current range? See the description of how different selection of pseudo-absences influences model outcomes and utility in Merow et al. 2013. The selection of pseudoabsences should also be informed by the type of conservation intervention that is most feasible. For example, I would imagine that conservation interventions within the species current countries of occupancy would be more likely to succeed than attempts in countries far beyond where the species has ever occurred. Should the area in Angola really be considered as part of the conservation strategy? The species has no means to natural dispersal to reach the area in Angola, so any such intervention would need to be human mediated, and require significant international collaboration. Is this feasible? If not, I would concentrate statistics on countries in the proximity of the species current range and leave additional areas as a passing mention (unless the politics of such species/climate change management is dealt with in more depth).

Having said this, what happens at the border of the Zambezian Somalia bioregion? Is the habitat entirely unsuitable? What do future models beyond this border suggest?

Table 1 – please indicate whether the area protected is the protected area that overlaps with the species distribution/potential distribution or whether it indicates the total area under protection in a particular country. I assume the former. It may also be useful to provide the protected area as a percentage. Combining the information in Tables 1 and 2 to indicate the proportion of PA that remains stable might also be useful?

Lines 319-333 and Tables. It is hard how to determine how the much larger numbers and percentages presented in lines 319-325 compare to the smaller percentages of stable protected area in lines 326-333. It might be better to concentrate the results on the area within protected areas currently suitable for these tortoises and how stable this remains over time, as well as loss and addition of potential new protected area that might become suitable?

Table 2: It would also be useful to add information on the current situation to Table 2 for comparison purposes (unless these numbers are intended to sum to the current total)? Is there any protected area that was suitable that becomes entirely unsuitable?

Line 312, Fig. 2 – it is hard to tell how suitable Zambia is without the inclusion of country borders on the map. Is the model working correctly if it does not predict currently suitable habitat in areas where observations have been made? What is so different about Zambian habitat that causes it not to be selected? Plotting part A and B of Figure 1 together might assist in identifying where the edge of the bioregion is in relation to the country borders.

Results: There are some key factors missing from the current results/discussion section:

• Why are the core drivers of the model not discussed? It would be useful to see response curves for individual variables, as well as effect sizes

• Related to the above, the discussion should include some mechanistic drivers of the potential patterns – why are the key variables the ones picked up by the analysis

• Are there any potential climate variables thought to influence the species breeding or foraging success that have not been included in the current models?

• Are soil data not available? It would be useful to include these data in the models if they are available.

• How does the representation of the current distribution compare to other models done for this species in the past

• How stable does the area currently under protection remain? To what extent is climate change really expected to add to the threat of the species? It appears that conservation of the species outside of the protected area is currently a priority regardless of climate change, but the extent to which climate change may change this in future is what needs to be discussed. Is there anywhere in particular, considering land-use and potential for other threats, where you would recommend the establishment of new protected areas to protect this species? [This would be in areas least impacted by climate change and most suitable for the species].

For the reported suitable ranges (lines 290-300), how was ‘suitable area’ quantified? I understand from the methods, there was a committee averaging technique applied, but this resulted in cases where some of the colours represent almost no agreement of presence. Which degree of consensus of suitable area (or colour on the map) was classified as ‘suitable’ for the numbers reported in the Tables?

Fig. 2 There does not appear to be a large difference between the RCPs in terms of high match areas (>0.5; yellow to red). Why might this be?

Fig. 3 please indicate the location of protected areas on this map

Merow, C., Smith, M.J. & Silander, J.A. (2013) A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography, 36, 1058-1069.

Hannah, L., Roehrdanz, P.R., Marquet, P.A., Enquist, B.J., Midgley, G., Foden, W., . . . Svenning, J.C. (2020) 30% land conservation and climate action reduces tropical extinction risk by more than 50%. Ecography.

Reviewer #2: General comments:

This paper examines the distribution of critically endangered pancake tortoises. The authors used occurrences obtained from Tanzania, Kenya and Zambia to estimate suitable and stable areas in other countries (projecting/predicting). They also used a protected area network to assess how many of the suitable and stable areas are protected for the conservation. The authors conclude most of the identified suitable habitats remain outside of protected areas. I think this is a sound contribution to the literature. I have provided specific comments below, aimed at clarification of methods and results. In addition, particularly in the discussion, I have suggested ways the text can be streamlined to flow better and be more concise and direct. Finally, I think some added text addressing some of the problematic areas of SDM at the end of the discussion might make this paper stronger. Specific comments are below, but in general I am referring to how SDMs are indeed frequently used, but they are also very controversial. So I just suggest you consider briefly addressing this and indicate how your study deals with these issues (e.g., by careful selection of data points etc.).

Specific comments:

Line 43: It seems like a few more citations might be helpful to demonstrate the growing interest in species distribution modeling.

Line 53: Additional citations seem warranted here.

Line 54: Provide citation, if possible.

Line 81: Consider starting the sentence differently because the previous sentence also beings with “although”.

Line 86: It would be more concise to say “…has indicated climate-matching of recipient sites is important for understanding suitability of these areas for introduced/translocated species”, or something similar. In general, this sentence is a bit cumbersome and seems to be discussing translocation, so you might want to set that up before this sentence with a better transition. Maybe just lead in with something simple transitioning from “no study has assessed climate change in these tortoises” to why re-introductions might be necessary.

Line 93: I think this phrasing is a little problematic “because SDM is the most widely accepted method of predicting climatically suitable habitats”. I might eliminate this part. SDM is frequently used, true, but it is also very controversial, and refers to many different methods, including occupancy modeling. So SDM ends up being an umbrella term for several of different ways of mapping species distribution.

Line 97: What is the implication for “there are endangered species that inhabit areas outside of protected lands”? I think this information will help make a stronger point. In essence—what is the point you are trying to make? I would suggest moving Lines 102-104 to the beginning of the paragraph, so the main point is clearer from the beginning. Only slight sentence re-structuring would be needed for the rest of the paragraph if you re-arranged this way.

Line 112: See earlier comment and consider removing “SDMs are essential”. This part is unnecessary. What is essential is the examination of how well protected areas perform, not the SDM. Plenty of other methods can be used to achieve the same goal.

Lines 113-114: Correct, and there is a lot of controversy over SDM’s, for example, in the data used to generate them (see below). I am not saying you need to reference these papers, I’m just illustrating a point that drastically different results occur with even small differences in data inputs. I wonder if controversy should be addressed here, or perhaps in the discussion in a way that shows you have thought about this and carefully selected your data and approach.

• Rodda, G. H., C. S. Jarnevich, and R. N. Reed. 2009. What parts of the US mainland are climatically suitable for invasive alien pythons spreading from Everglades National Park? Biological Invasions 11:241–252.

• Pyron, R. A., F. T. Burbrink, and T. J. Guiher. 2008. Claims of Potential Expansion throughout the U.S. by Invasive Python Species Are Contradicted by Ecological Niche Models. PLoS ONE 3:e2931

• Rodda, G. H., C. S. Jarnevich, and R. N. Reed. 2011. Challenges in Identifying Sites Climatically Matched to the Native Ranges of Animal Invaders. PLoS ONE 6:e14670.

Line 125: “the occurrence of more stable areas over time” seems like it needs to be more explicitly defined, but hopefully this will be more clear in the methods.

Line 127: Consider re-phrasing as “This study may inform specie management approaches, including identifying suitable areas for translocation and the establishment of nature reserves where species can be protected with minimal human intervention [10]. This way is more concise (the sentence is long). But more importantly, the way you have this written reads as a promotion for SDM’s and I don’t think that is warranted without more information on how they need to be used carefully. The discussion probably needs this information.

Line 159: Remove “up”

Lines 160-162: “…tortoises are disjointly distributed from the northern to southern areas, passing though the central to south-eastern regions of the country”. This reads awkwardly and is a bit confusing.

Line 175: Remove “on the ground”

Lines 175-178: Can you reference one of your maps for where your field sites were, generally? Or is this best left unsaid because of poaching concerns? I mention this because many will not know where these areas are without a visual.

Lines 185-187: I like that you specified how you cleaned the records and would say that as much information here as possible is warranted, because even one incorrect data point can throw off model results.

Lines 188-190: It is not clear why you would need elevation data to get coordinates. Please clarify. Further, is this a source of error? These are extrapolated data and could inflate errors in your predictions. Again, I think some careful explanation of how these models are sensitive to user inputs, and how you addressed this carefully in your approach is important.

Lines 216-219: Provide a citation indicating they are more realistic.

Line 214: What is a Representative concentration pathway?

Line 219: Variations in future scenarios? Please consider adding a little more detail.

Line 220: Are you saying you ran 10 models to quantify variation? Please clarify.

Lines 222-223: What are all these acronyms? Define them as models and list out what they stand for or provide a citation that identifies them.

Line 225: What is BIO 18? Bio 19? Clarify.

Lines 224-230 are confusing. What are you doing? Perhaps say something like “To evaluate different models we first eliminated x, because of …., and then eliminated y, to assess …..” Or some similar phrasing.

Line 252: What standard procedures?

Line 261: How does this avoid subjective thresholds? Perhaps re-phrase as “The selected models were binarized using the AUC threshold to avoid use of a subjective threshold”. Include citation?

Line 272: Consider removing “furthermore”.

Line 284: You are letting R average the models built with different algorithms, which is fine. I just wonder if you can extract the information and depict a table of each algorithm’s relative performance. This is just a suggestion.

Line 300-301: You say “will continue to” and “will expand” etc. Clarify that you are predicting these things with “we predict that…” for example. These results are just predictions, not an indication of what will happen with 100% certainty. Fix this throughout the Results where applicable. E.g., Lines 320, 323, 331

Line 329: Define stability in this context; likewise define in Table caption on Line 341. How did you determine stability?

Line 335: Is ‘red sheds’ a typo?

Line 345: Begin with “We predicted that…” rather than “Our SDM predicted”

Line 359: Reads awkwardly. What do you mean by “the isolation of pancake tortoise populations is also occurring within Tanzania…”? Does this refer to future predictions? Should you re-phrase as “Based on our model predictions we expect tortoise populations to become isolated…”?

Line 361: Re-phrase, reads awkwardly.

Line 363: “Furthermore” is not necessary and reads a little awkwardly in this context.

Lines 356-370 could use some tightening up; the flow if the text is a bit choppy and could be streamlined a bit. Instead of “As pancake tortoises prefer areas featuring Precambrian rocks, the presence of other rock types between the suitable habitats could act as a distribution barrier [13]. Therefore, this could be another reason for the non-existence of pancake tortoises in some climatically suitable habitats; however, the species may occur in the Zambezian floral region, provided that suitable habitat is available” try “Pancake tortoises do not occur in some climatically suitable habitats (citation/Fig?). The prevalence of non-Precambrian rock types between areas where we predict suitable habitat may preclude occupancy by Pancake tortoises, as they prefer areas dominated by Precambrian rocks [13].”

Line 370: This is confusing “however, the species may occur in the Zambezian floral region, provided that suitable habitat is available [13,37]” given the previous statement saying that they often don’t occur in suitable habitat.

Lines 371-383: As with the previous comment this text could be re-written to be more concise and to flow better.

Line 387: Remove “Additionally”. Also, as above, this sentence could use some restructuring for conciseness and clarity. For example, I might re-phrase as follows: “We recommend robust habitat suitability studies in these countries and further quantification of occupancy status given the species apparently occupies suitable habitats in the nearby Zambezian and Somalian biogeographical regions [37],[13].”

Line 391-393: Consider re-phrasing as “Protected areas are critical tools for biodiversity conservation [23], yet the African Protected Areas Network offers inconsistent protection to tortoise species [33].

Line 402: “none of the recorded species is within” reads awkwardly. Consider re-phrasing as “while in Zambia, the species does not occur within the Protected Area Network [12,13].”

Line 403: Consider re-phrasing as “In Tanzania, only 4 out of 22 national parks are occupied by pancake tortoises.” Also, is this based on your results? If so, clarify with something like “we predict that in Tanzania only 4 of 22 parks….”

Lines 404-409: Consider re-phrasing for better flow: “The pancake tortoise’s suitable habitat is largely unprotected in both the current and the future scenarios, likely increasing the risk of over exploitation and exacerbating negative effects of habitat destruction as in Tanzania [18], Kenya [10,12] and Zambia [12,13]. Likewise, ectoparasite prevalence is higher outside of protected area [59], potentially increasing risk to the species”.

Line 412: Consider re-phrasing as “Because the range of Pancake tortoises does not include current and future predictions for climatically suitable habitats, we recommend future studies be conducted in areas where pancake tortoises do not exist to confirm absence of the species.”

Line 444: “or even their inversion into a trend of growth.” This part seems a bit speculative at this point. I would consider removing, but if you keep, perhaps consider re-phrasing this sentence with something similar to: “We predict expansion of suitable habitat for pancake tortoises in the future, which may conserve populations of this critically endangered reptile”.

Line 444-448: This is an important sentence. I would re-phrase as “Importantly, the largest proportion of suitable habitat is outside of the current Protected Area Network, therefore we suggest the Pancake tortoise be upgraded in its listing status from CITES Appendix II to Appendix I.” This is more streamlined and direct.

Line 449-452: Re-phrase as “Because our results were largely based on use of climactic variables, our findings should not be treated as ready-made for on-the-ground application but could be used as one of many tools to help in conservation planning of Pancake tortoises.” This is a bit more streamlined and concise.

Line 453: Re-phrase as “Our decision to use primarily climate variables was because climate variables drive most of the specie’s distribution”. This is a lot more concise, or less repetitive.

Line 462: I think a stronger ending is warranted. I might suggest saying “Our study provides a solid foundation for future development of conservation measures aimed at protecting populations of the critically endangered pancake tortoise.” or something similar.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Nicola van Wilgen

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Attachment

Submitted filename: Merow et al. 2013 Maxent.pdf

Click here for additional data file. (879.2KB, pdf)
Attachment

Submitted filename: Hannah et al 2020 Extinction risk mitigation.pdf

Click here for additional data file. (2MB, pdf)
PLoS One. 2021 Jan 20;16(1):e0238669. doi: 10.1371/journal.pone.0238669.r002

Author response to Decision Letter 0

24 Dec 2020

Please, see the attached document named 'PLOS ONE Reviewer comments and responses' for response to editor and reviewers comments.

Attachment

Submitted filename: PLOS ONE Reviewer comments and responses.docx

Click here for additional data file. (64KB, docx)

Decision Letter 1

Stephanie S Romanach

7 Jan 2021

Protected areas network is not adequate to protect a critically endangered East Africa Chelonian: Modelling distribution of pancake tortoise, Malacochersus tornieri under current and future climates

PONE-D-20-25666R1

Dear Dr. Eustace,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Stephanie S. Romanach, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

- e.g., L 379, L 455: "Pancake" is capitalized but elsewhere is lowercase. Please check manuscript text for consistency.

- Fig S1 middle left panel x-axis label is incomplete. Perhaps remove "of" to fit all text or expand figure size or reduce font size.

- Table S1: should read "GCMs" not GCM's"

Reviewers' comments:

Acceptance letter

Stephanie S Romanach

11 Jan 2021

PONE-D-20-25666R1

Protected areas network is not adequate to protect a critically endangered East Africa Chelonian: Modelling distribution of pancake tortoise, Malacochersus tornieri under current and future climates

Dear Dr. Eustace:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Stephanie S. Romanach

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Fig. Response curves from the ensemble models to the six selected bioclimatic variables.

    Response curves were fitted through locally estimated scatterplot smoothing (LOESS). Grey background shows a scale from y-axis which is replicated to every graph.

    (DOCX)

    Click here for additional data file. (236.9KB, docx)
    S2 Fig. Variable importance for six less correlated climatic variables of the ensemble species distribution model.

    BIO 8 = mean temperature of the wettest quarter, BIO 3 = the isothermality, BIO 2 = mean diurnal range, BIO 15 = precipitation seasonality, BIO 14 = precipitation of the driest month and BIO 13 = precipitation of the wettest month.

    (DOCX)

    Click here for additional data file. (103.3KB, docx)
    S1 Table. Ten Global Circulation Models (GCMs) used in our study.

    (DOCX)

    Click here for additional data file. (20.1KB, docx)
    Attachment

    Submitted filename: Merow et al. 2013 Maxent.pdf

    Click here for additional data file. (879.2KB, pdf)
    Attachment

    Submitted filename: Hannah et al 2020 Extinction risk mitigation.pdf

    Click here for additional data file. (2MB, pdf)
    Attachment

    Submitted filename: PLOS ONE Reviewer comments and responses.docx

    Click here for additional data file. (64KB, docx)

    Data Availability Statement

    According to IUCN the pancake tortoise is critically endangered and threatened by international pet trade therefore the location data collected by authors is not publicly available. However, these data are available from the College of African Wildlife Management upon reasonable request. Please, contact the Head of Research and Consultancy Department through mweka@mwekawildlife.ac.tz if you need these data. Other secondary data used in this study are available from the original source as cited in the method section.

    Articles from PLoS ONE are provided here courtesy of PLOS

    RESOURCES