Is Severe Drought Increase Or Decrease In Carrying Capacity

• Loading metrics

Open Admission

Peer-reviewed

Inquiry Article

Predicting the Future Touch on of Droughts on Ungulate Populations in Arid and Semi-Barren Environments

• Clare Duncan,
• Aliénor 50. M. Chauvenet,
• Louise G. McRae,
• Nathalie Pettorelli

ten

• Published: December 17, 2022
• https://doi.org/10.1371/journal.pone.0051490

Figures

Abstruse

Droughts can have a severe touch on the dynamics of animal populations, peculiarly in semi-arid and barren environments where herbivore populations are strongly express by resource availability. Increased drought intensity under projected climatic change scenarios can be expected to reduce the viability of such populations, yet this impact has seldom been quantified. In this study, we aim to make full this gap and assess how the predicted worsening of droughts over the 21^st century is likely to touch on the population dynamics of twelve ungulate species occurring in arid and semi-barren habitats. Our results provide support to the hypotheses that more than sedentary, grazing and mixed feeding species volition be put at high risk from hereafter increases in drought intensity, suggesting that management intervention nether these conditions should be targeted towards species possessing these traits. Predictive population models for all sedentary, grazing or mixed feeding species in our written report testify that their probability of extinction dramatically increases under future emissions scenarios, and that this extinction adventure is greater for smaller populations than larger ones. Our study highlights the importance of quantifying the current and time to come impacts of increasing extreme natural events on populations and species in club to ameliorate our ability to mitigate predicted biodiversity loss nether climatic change.

Citation: Duncan C, Chauvenet ALM, McRae LM, Pettorelli N (2012) Predicting the Hereafter Impact of Droughts on Ungulate Populations in Arid and Semi-Arid Environments. PLoS ONE 7(12): e51490. https://doi.org/10.1371/periodical.pone.0051490

Editor: Frank Seebacher, University of Sydney, Commonwealth of australia

Received: September iv, 2022; Accepted: November 5, 2022; Published: Dec 17, 2022

Copyright: © 2022 Duncan et al. This is an open-admission article distributed under the terms of the Creative Eatables Attribution License, which permits unrestricted apply, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: The authors have no back up or funding to report.

Competing interests: The authors have declared that no competing interests exist.

Introduction

In light of the current global extinction crisis, understanding how and where drivers of population refuse will have effect has never been more of import [ane]. Climate change is expected to be a major driver of species extinctions in the 21^st century [2], [3]. Average changes in greenhouse gas concentrations are expected to produce directional changes in climatic conditions, and increase the level of inter-almanac variability in these conditions [4]. Droughts are a significant component of such climatic variability, and can take a devastating impact on animal populations [5]–[eight]. Through processes such as recurrent reductions in population numbers and the consequent genetic effects caused past demographic bottlenecks [5], droughts have the potential to lead populations, and unabridged species, to extinction.

In a recent publication, the Intergovernmental Panel on Climate Modify (IPCC) reported a likely increase in droughts over the 21^st century in diverse regions of the world, including southern Europe and the Mediterranean, central Europe, central North America, Cardinal America and United mexican states, northeast Brazil, and southern Africa [nine]. This predicted increase is a potential cause for conservation business concern: although the impact of droughts on population size fluctuations has been assessed past many [10]–[xiii], the potential bear on of expected changes in drought conditions on wildlife populations has near never been quantified (simply run into [fourteen]). Moreover, to appointment no comparative study has been conducted to assess the touch on of future changes in drought conditions for species exhibiting contrasting life histories. The chance of extinction of species in response to various threats is partially shaped by their intrinsic biological characteristics, east.thousand., body mass, feeding strategy, reproductive strategy, territoriality and domicile range size [15], [sixteen]. Certain life history traits can exist expected to make species more than susceptible to increased droughts, such as strong dependence on permanent water-sources [17], [eighteen]; obligate grazing or mixed feeding (due to whole or partial dependence on drought-intolerant food species [19], [20]); sedentary behaviour (due to being unable to escape the effects of drought conditions on resource availability [20]–[21]). Notwithstanding, how possessing these traits will shape the future susceptibility of populations to changes in climatic conditions is currently unknown. Such information is still deemed necessary to ameliorate our ability to mitigate predicted biodiversity loss under climatic change.

As highlighted by the IPCC 2022 report [9], the need to quantitatively assess how predicted changes in drought conditions are likely to impact wildlife is particularly neat for populations inhabiting semi-arid and arid regions. Main productivity in these environments is already heavily limited by precipitation [22]. Every bit a result, even a slight increase in drought elapsing, intensity or frequency in these regions has the potential to severely touch resources availability and thus plant eater abundance [12], [23]–[25]. Increased drought conditions may likewise indirectly impact herbivore populations in these habitats as reduced forage and water availability tin lead to increased vulnerability to predation, due to e.1000., increased densities at thin water points [26]–[27]. This study thus proposes to quantify the touch of future changes in drought weather condition on time to come growth rate of terrestrial ungulate species with contrasted life histories inhabiting arid and semi-arid environments. Based on the knowledge available on ungulate susceptibility to drought [18]–[23], nosotros hypothesised that population growth rates of species that are grazers or mixed feeders (H1), and that are relatively sedentary (H2) should show a pregnant negative human relationship with drought intensity, and be more than negatively impacted by drought intensity than population growth rates of other species groups.

Materials and Methods

Drought Data

An farthermost natural issue, such every bit a drought, can exist defined equally an event that is rare within its statistical reference distribution. It is normally equally rare or rarer than the 10^th or 90^thursday percentile [4], [28], and can exist quantified in relation to a specific (possibly impact-related) threshold [9]. However, it is difficult to quantitatively ascertain a drought every bit in that location are several unlike forms (e.thousand., meteorological drought, agricultural drought and hydrological drought; see [29]). Equally a consequence, several unlike drought indices accept been developed [thirty], [31]. I index adult to quantify meteorological drought is the Palmer Drought Severity Index (PDSI; [32]). The PDSI is a common meteorological drought index [33], which has been used to quantify both historical and projected long-term trends in global dehydration (run into [34]). The PDSI is a standardized index, incorporating both previous and current moisture supply (atmospheric precipitation) and demand (potential evapotranspiration, PE), which ranges from −10 (dry out) to +x (wet).

Palmer'due south original PDSI was calibrated using fixed coefficients from limited information from the central U.s., and to improve spatial comparability several attempts have been fabricated to recalibrate the PDSI index since (meet [35]). Self-calibrating PDSI (sc_PDSI) [36] has been found to be more spatially comparable than Palmer's original PDSI, while calculation of PE using the more sophisticated Penman-Monteith equation (PDSI_pm) as opposed to the original Thornwaite equation, has likewise improved the efficiency of the PDSI [35]. We used global PDSI information (time to come sc_PDSI_pm) for the years 1970 to 2005. It is calculated using observed or modelled monthly air surface temperature and atmospheric precipitation, calibrated with Penman-Monteith PE based on historical information and gridded to a 2.5°x2.5° grid ([35], [36];http://world wide web.cgd.ucar.edu/cas/catalog/climind/pdsi.html).

Study Species and Population Data

Only ungulate species occurring in arid and semi-arid areas where droughts are predicted to become more common and more than intense over the 21^st century were considered for this analysis. To identify relevant species, we offset compiled ungulate species distribution data (genera Artiodactyla, Perissodactyla, Proboscidea) from the International Wedlock for the Conservation of Nature (IUCN) mammal species dataset (Geographic Information Systems data available at http://www.iucnredlist.org/technical-documents/spatial-data), and then overlaid these species distribution maps with a map of global arid areas of the Köppen-Geiger climate classification ([37]; Geographic Information Systems information available at http://koeppen-geiger.vu-wien.ac.at/shifts.htm).

Time-series affluence data for 1970–2005 for the ungulate populations found in arid and semi-arid areas were collated from the WWF/ZSL Living Planet Index (LPI) database [38]. The database is comprised of yearly affluence data (either population size estimates, population density, relative affluence, biomass, index information, proxy data, samples or measures per unit endeavour) for vertebrate populations, collated from information in published scientific literature, unpublished reports and online databases. Data is only included in the database if the method of drove or estimation, the geographic location of the population, and the units of measurement are known, and if the information source is referenced and traceable.

Every bit our report focused on the bear on of drought on population growth rate, abundance records were just included in the analysis when population estimates were for 2 consecutive years; i.eastward. if two affluence records had a ane or more than years gap between them they were not included. Only species for which the sample size of growth rates over all populations was greater than xx observations were included in the analysis. Moreover, populations for which the initial starting abundance record was smaller than the average herd size range for that given species were removed in order to ensure that we did non include unstable populations in our analysis. In addition, individual populations for which heavy management practices (e.g. African elephant (Loxodonta africana), African buffalo (Syncerus caffer), blue wildebeest (Connochaetes taurinus) and impala (Aepyceros melampus) in Kruger National Park, Southward Africa), or poaching activities (e.g. African elephant in Ruaha National Park, Tanzania and white rhinoceros (Ceratotherium simum) in Garamba National Park, Democratic Republic of Congo) were known to occur over the years of the abundance record (i.e. where references in the literature confirm the presence of such processes) were also excluded in lodge to eliminate potentially biased affluence information. Then, populations for which estimates were collated from a number of unlike sources and using a number of different estimation methods were as well omitted due to the resulting high level of sampling bias. Finally, populations that contained growth rates higher than that which is physiologically possible for the given species (e.thousand., higher than possible if all members of the population in a given year were female, each gave nativity to the maximum number of offspring possible for that species in one year, and there was no bloodshed) were not used (n = iv). Such loftier growth rates are very probable attributable to population increases caused by alternative processes such as immigration, the intentional introduction of individuals into national parks, or wrong estimates of population size.

The resulting dataset comprised time-serial affluence data from 71 populations of 12 ungulate species (Tabular array S1). Most ungulate species currently covered past the LPI database occur in eastern and southern Africa, due to increased sampling effort in these areas. As a result, nearly of our study species and populations also occur in these regions. The geographic coordinates for each population were taken from the LPI database [38], and monthly sc_PDSI_pm values for each location were collated from the ii.5°x2.5° filigree pixel in which the population savage and the years in which they were surveyed.

Calculating Drought Indices

A drought tin can be described by 3 axes: duration, frequency, and intensity [39]. Drought duration refers to the timescale of the drought occurrence, e.thousand., the length of the drought episode. Frequency refers to the average interval (or distance) between drought events at a given location, which can vary betwixt two years in extreme arid regions and 100 years in extremely wet regions [39]. Intensity refers to the extent of the precipitation arrears, and is usually calculated in relation to the duration as the cumulative wet deficiency across the drought duration [39], [forty].

For this study, a year is classified as a "drought twelvemonth" if the sc_PDSI_pm value of at least one calendar month within that year is below the value of the 10th percentile of its statistical reference distribution at a detail location (e.grand., threshold θ). Drought intensity then refers to the number of consecutive months within a given "drought year" in which q<θ (where q is the sc_PDSI_pm value for a given month). In addition, drought frequency (or recurrence interval) is defined every bit the average altitude (in years) between "drought years" beyond all report locations for each species between 1970 and 2005.

In guild to test the impact of changes in drought conditions on the population growth rate of ungulate species in arid and semi-arid areas, we developed four potential predictor variables of drought: the total number of months of the preceding yr (T ), and the preceding two years (T_{t 2}), in which q<θ; the maximum number of consecutive months of the preceding year (C ), and the preceding 2 years (C_{t 2}), in which q<θ. The variables C and C_{t 2} were developed based on the definition of a drought effect provided by Sheffield and Forest [40]. The variables T and T_{t 2} were developed in order to create indices of drought that incorporated potential small breaks in drought occurrence, which could not be incorporated under the variables C and C_{t 2}. The correlation between our chosen drought index and both annual average PDSI and annual modal PDSI beyond all study populations was tested using the Spearman's rank correlation. All analyses were carried out in R v. 2.14.2 [41].

Modelling the Time to come Impact of Drought on Ungulate Populations

Establishing a link betwixt drought indices and growth rates.

For all ungulate species considered, we calculated observed population growth rates (r_t ) for a given twelvemonth t between 1970 and 2005 as the change in abundance over time, normalised by the natural logarithm: where Northward_t is the population size at time t, and N_{t +1} is the population size at time t+1.

We so carried out single predictor regressions of observed population growth rate (r_t ) against C, T, C_{t 2} and T_{t 2} using linear mixed effect models with population location and species as random effects in order to place the best predictor of r_t for all the populations and species considered. This helped us place the best drought indicator variable of the four.

In order to test our 2 hypotheses, we then created four subset species groups based on descriptions of individual species nutrition and motion behaviour in the literature [42]: (one) non-sedentary species (due east.g., species that are described as being nomadic, migratory, displaying seasonal movements, or extremely wide-ranging) that wholly or partially depend on drought-intolerant nutrient species (e.chiliad., species that are described every bit pure grazers or 'by and large' grazers, and mixed feeders) (hereafter MG), (ii) non-sedentary species that practise non depend on drought-intolerant food species (eastward.1000., species that are described as pure browsers or 'mostly' browsers, and omnivores) (futurity MB), (3) sedentary species that wholly or partially depend on drought-intolerant food species (time to come SG), and (4) sedentary species that do not depend on drought-intolerant food species (hereafter SB). Nosotros then carried out single predictor regressions of population growth charge per unit (r_t ) for each of these groups against the best drought indicator variable; we used linear mixed models with population location and species as random furnishings.

Model description.

For each grouping of species (MG, MB, SG, and SB) for which our best performing variable of drought intensity was found to be a significant predictor of r_t we developed species-specific stochastic population models to predict the impact of future drought conditions on their viability. Because each species was divided into several populations, for which records of abundance were dissimilar, we built the model to projection each population separately, e.chiliad., the initial population size N ₀ was different per population. Our model took the course: where

and

R_t is the modelled population growth rate at time t. b_t is a coefficient describing the impact of drought conditions (D_t ) on the modelled growth charge per unit for each group (MG, MB, SG, and SB): information technology was estimated using the outputs of the linear mixed effects models for each group of ungulate species. a_t is the average growth charge per unit in the absenteeism of drought: this coefficient was estimated using the observed average growth rate (r_av ) for each private species. D_t describes the drought conditions of a given year t and reflects the construction of the best drought indicator variable found above.

While modelling D_t we aimed to reproduce observed drought patterns in our data. To do so, we showtime determined the average lengths of drought and non-drought episodes per years (i.e., the average number of months in droughts or not in drought per years) across all populations of the species for which nosotros congenital a stochastic population model. This was to be able to simulate the length (in months) of a drought episode if a given modelled year was in drought. To determine if a modelled yr experienced drought or not, we offset generated an initial value D_t at time t ₀ by comparing a random number sampled from a uniform distribution between 0 and 1, to an 'initial threshold' (Table S2). This initial threshold number was the probability that a given twelvemonth t would be in drought based on observed drought patterns for each species between 1970 and 2005. If the random number was greater than the threshold number, year t was considered to be not in drought (D_t  = 0). If the random number was lower than the threshold number, year t was considered to be in drought, and D_t was assigned a value between i and 12 to represent a number of months in drought. The value of D_t when in drought was randomly sampled from the observed distribution of our best performing drought indicator variable for the given grouping of species.

So, at each time-step (i.east., each simulated year) we assigned a drought or non-drought condition to the year based on new 'distance thresholds', representative of observed drought consequence length (in years) and observed non-drought event length (in years). D_{t +1} was computed by comparison a random number (once more from a uniform distribution between 0 and i) to these 'distance thresholds'; if in the preceding year, at time t, D_t >0, then D_{t +1} would be computed by comparing the random number generated to the 'drought distance threshold' (Figure 1). The drought distance threshold for each group of species is the observed probability for each group that if year t was in drought, then year t+1 would too be in drought. However, if at time t, D_t  = 0, and then D_{t +1} would be determined past comparison the random number generated to the 'non-drought distance threshold' (Figure 1). Similarly, the non-drought distance threshold is the observed probability for each group of species that if year t was non in drought (D_t  = 0), and so year t+ane would also non be in drought. If D_t  = 0 and the random number generated was greater than the non-drought altitude threshold, then D_{t +ane} = 0, or if D_t was >0 and the random number generated was greater than the drought altitude threshold then the value of D_{t +1} was once more generated from a random sample based on the probability distribution of our best performing drought variable, described above.

Figure ane. Model generation of D_{t +ane} at each time step, where R is a random number sampled from a compatible distribution, and 'ddt' and 'nddt' represent to the 'drought distance threshold' and the 'non-drought distance threshold' respectively.

https://doi.org/10.1371/journal.pone.0051490.g001

To examination how well how the stochastic population models performed, we calculated the r-squared values between the observed abundances between 1970 and 2005 and the model-predicted abundances for each population of each species. Moreover, we assessed the ability of the models to reproduce observed patterns in the average length of drought and non-drought events (in years) by performing Wilcoxon signed-rank tests for the 95% conviction intervals of the median for both the observed and model predicted drought and non-drought event lengths (in years).

Simulations.

For all groups of species for which a significant human relationship was found betwixt growth rates and the best performing drought indicator variable, we used emissions scenario predictions for the region in which each species occurs to train our species-specific model simulations. Iii scenarios of future drought occurrence were considered: the first scenario assumed the continuation of current conditions (hereafter 20C). The two other scenarios were based on Sheffield and Wood's projections [twoscore] for the southern African region. Under the second scenario, a doubling of curt-term (4–half dozen months) droughts detectable by 2025 was considered (time to come B1); under the third scenario, a tripling of short-term droughts detectable by 2040 was considered (hereafter A2) [40]. Changes in occurrence of medium-term (seven–11 months) droughts were not investigated past Sheffield and Wood [40]. As we were modelling drought occurrence on a yearly basis (e.g., ane–12 months length), nosotros made the assumption that medium-term droughts would showroom similar increases as short-term droughts under scenarios B1 and A2, which make our results a cautious underestimate of futurity drought occurrence.

For each population of each species, nosotros ran a model for which the initial abundance corresponded to that population's commencement abundance record. Considering the model was stochastic, it was run for 5000 simulations in social club to generate a large number of possible population trajectories from which to draw a mean observation. The total number of time-steps for each population was (1) 2005-t ₀ (as each population had a different starting year t ₀), in order to reproduce observed abundance, and (2) 2099-t ₀, in guild to model futurity abundance up to 2099 nether different climatic change scenarios. A given population was considered extinct when its size was ≤5 individuals.

Results

Iii (C, T and T_{t two}) of the 4 derived indices of drought conditions were constitute to bear witness a pregnant negative relationship with observed growth rates (r_t ) across all study species (C: slope = −0.01, p<0.01; T: gradient = −0.01, p<0.01; T_{t 2}: slope = −0.01, p = 0.03), while the relationship betwixt the maximum number of consecutive months of drought (q<θ) over the preceding year (C_{t 2}) and observed growth rates (r_t ) was not significant (slope = −0.01, p = 0.11). All iv drought indices were highly correlated with each other, with C_{t 2} showing the lowest correlations with all other variables (Tabular array S3), potentially explaining the not-significance of its relationship with r_t . The two measures of drought intensity over the preceding yr (C and T) displayed the most meaning relationships with observed growth rates (r_t ). Thus, based on previous definitions of drought occurrence in the literature [xl], we elected to use C only in further analyses. There was a loftier degree of correlation between C and the annual average PDSI (rho = −0.60, p<0.001), and C and the annual modal PDSI (rho = −0.59, p<0.001) across all study populations.

Every bit expected (H1 & H2), species with different life histories did not exhibit the same level of susceptibility to drought atmospheric condition: when modelling observed growth rates as a function of C for each species group (SB, SG, MB, MG), the only group of species for which C showed a significant negative relationship with growth rates was the grouping of sedentary species that either wholly or partially depend on drought-intolerant food species (SG group; slope = −0.04, p = 0.001) (Figure 2; Figure S1). Species that fell within this group were buffalo (Syncerus caffer), impala (Aepyceros melampus), hartebeest (Alcelaphus buselaphus) and waterbuck (Kobus ellipsiprymnus), and the population dynamics impacts of future increases in drought conditions were therefore only investigated for these four species.

The stochastic population models showed a certain level of heterogeneity in their ability to mimic observed abundances across individual species and across private populations. For example, our model explained 70% of the variance in abundance of buffalo in the Serengeti-Mara ecosystem and in Addo Elephant Park, but but 20% in the Narok District; information technology explained 36% of variance in affluence of waterbuck in Malilangwe Biological reserve but only 2% in Kruger National Park (Tabular array one). Withal, observed average drought (observed: pseudomedian = 2.0, lower 95% Confidence Interval = 1.five, upper 95% CI = 2.0; predicted: pseudomedian = 2.0, lower 95% CI = 1.5, upper 95% CI = 2.0) and not-drought episode lengths (observed: pseudomedian = 3.0, lower interval = 2.5, upper 95% CI = four.0; predicted: pseudomedian = iii.0, lower 95% CI = 2.five, upper 95% CI = 3.five) were well replicated.

Population projections for the four species in group SG showed that boilerplate population growth rates (λ) decreased, while extinction probabilities (E) and the variation in average growth rates steadily increased, when successively considering scenario 20C, B1 and A2 (Table 2; Table S4). There was all the same piddling variation in projected boilerplate growth rates and extinction risks between scenarios B1 and A2, probable due to the late onset of increases in drought intensity. Waterbuck populations had extremely low negative boilerplate growth rates (λ) even nether connected electric current conditions (scenario 20C), with the average take a chance of going extinct at the end of this century being 100%. Contrastingly, buffalo and impala displayed positive projected average growth rates, with both species showing a negligible drought-related average risk of extinction nether all scenarios (Table 2). Hartebeest populations showed negative projected average growth rates under all scenarios, with a relatively high take a chance of extinction (51.one%) nether continued electric current conditions which increased to 66.4% and 69.1% under scenarios B1 and A2 respectively (Tabular array 2). In improver, our results besides showed that smaller populations of all species will exist put at higher run a risk of extinction from increasing futurity drought occurrence than larger populations (Tabular array S4 and Table S5).

Table 2. Modelled average growth rate (λ) and mean extinction probability (E; in %) to 2099 across all populations of each sedentary grazer species under model scenarios 20C, B1 and A2. SD stands for standard difference.

https://doi.org/x.1371/journal.pone.0051490.t002

Discussion

Investigations into the potential time to come impacts of climate modify on biodiversity more often than not consider directional changes in environmental weather condition (meet eastward.g. [43], [44]), and studies into the furnishings of climate extremes are few at present. Here nosotros provide research to help close this gap, and nowadays a model framework to quantitatively assess the impact of potential future increases in such highly variable and devastating events. Our study shows that future climate alter will negatively bear on certain ungulate species in arid and semi-arid environments, dramatically increasing extinction risk from drought occurrence for some of them. Our results besides provide back up to the hypotheses that the species most at risk from increasing future drought intensity are those that are relatively sedentary, and that are wholly or partially dependent on drought-intolerant food species (e.g., grazers and mixed feeders).

Our findings that sedentary ungulate species (as opposed to nomadic or migratory species), which are dependent on drought-intolerant food species (as opposed to browsers or omnivorous species) are more at hazard from current and future drought atmospheric condition are in line with frequent reports in the literature of species exhibiting these life history traits suffering high mortality during individual drought events [xiii], [20]–[22]. Our results suggest that now the frequency and intensity of drought occurrence is sufficiently low that it is not inflicting a significantly negative bear on on populations that are currently able to escape the effects of resources depletion in dry conditions. Even so, as drought intensity increases under hereafter climate change its impact on such species may become pregnant.

The results of the predictive stochastic population models show that hartebeest and waterbuck will exist put at extremely high take a chance from future increases in drought intensity nether climate change over this century. Conversely, the risk of buffalo and impala populations going extinct volition remain depression. Those patterns are likely to reflect reported general population trends for these species throughout their global ranges, as well as trends occurring at the level of the study populations. In fact, population trends for hartebeest, buffalo and waterbuck across their global ranges are reported by the IUCN to be by and large decreasing at present [45]–[47]. In particular, according to our simulations, waterbuck have an extremely high probability of extinction and low average population growth rates fifty-fifty under continued present weather condition. While reported decline in the species is often largely due to poaching [46], the species is too one of the most h2o-dependent of African ungulate species and has a high-protein dietary requirement [48]. As such it is extremely susceptible to drought atmospheric condition, explaining our simulations' results. The positive average growth rates and depression gamble of extinction displayed past buffalo under all scenarios, notwithstanding, does not necessarily evidence an immunity of this species to drought atmospheric condition. Instead, the impact of drought on buffalo population dynamics could be dampened, or fifty-fifty counteracted, by alternative processes. For case, all study populations occur in large national parks where populations are doing well, with an increasing population trend in Addo National Park, South Africa, Mountain Zebra National Park, South Africa and Lewa Nature Conservancy, Kenya since 1991, 2002 and 1990 respectively [49], [50]. In addition, the only years in which affluence data are bachelor for two consecutive years for the population in the Serengeti-Mara Ecosystem, Tanzania occur over 1970–1977 when the population was undergoing a rapid improvement following the eradication of the rinderpest disease in the region [51]. As a effect, our population projections for buffalo nether the 3 scenarios over the 21^st century may be considerably modest.

The results presented here exercise have some limitations. First, sample size was minor for some of our study species due to populations being omitted for reasons listed above. Second, in the absence of data regarding potential alternative processes that may exist interim on the study populations, the chief supposition of this work is that drought intensity is the sole procedure acting on growth rates. Even so, management practices could besides be acting on the population dynamics of studied populations, equally many of them occur within protected areas and national parks and management deportment are often listed as 'unknown' within the LPI database. For case, water- or food-provisioning [52]–[54], culling [21] or fencing [55]–[57] may serve to either increase or subtract the resilience of ungulate populations to drought occurrence. Similarly, processes such as poaching (particularly in the example of the black rhinoceros; [58]), predation [54], [59], dispersal [xiii], [60], disease outbreaks [61], [62], or the impacts of other farthermost natural events (e.g., wild fire or flooding; [5]), which too take an effect on some ungulate populations, cannot be accounted for in our data. The influence of such processes could be buffering the impact of the effect of drought on populations of our study species. Third, interpretation methods of yearly abundance within our dataset also differed between individual populations of our report species, with some methods resulting in coarse resolution data (e.g., scaling-upwards from walking transects), which may similarly have buffered the effect of drought occurrence on these populations. Individual population size estimates may also suffer some degree of inaccuracy due to the difficult nature of obtaining counts for game species [63]–[66]. Finally, the population models congenital to project species' abundance under future climate change scenarios remain relatively simple. For example, they ignore processes such as density-dependence, predation or the departure in survival and reproduction rates of individuals of different sexes and ages, which are known to affect population dynamics [67]–[69]. While our results provide a practiced kickoff model of variation in future population abundance in response to drought, nosotros admit that additional more complex studies will be required in guild to gain a consummate understanding of the potential futurity impact of increases in drought intensity under climate change on the persistence of ungulate populations.

In the face of our results, should highly sedentary, grazing and mixed feeding ungulate species exist targeted by park government in the time to come for management in drought years? Some such species are clearly highly susceptible, but potential management strategies for their benefit may in reality come at a cost to them and to other ungulate species in arid and semi-arid environments. Studies have institute that provisioning with bogus water-holes tin enable less sedentary species to aggrandize their ranges within national parks and can promote increases in these populations, in plow resulting in heightened ungulate densities at such water points, and negatively impacting rarer sedentary species such every bit the waterbuck, roan antelope (Hippotragus equinus) and tsessebe (Damaliscus lunatus) through resources exhaustion and increased predation [53], [54], [lxx]. This upshot has too been found to extend to the predators of ungulate species, with lions (Panthera leo) in the Kruger National Park, Southward Africa, benefitting from college casualty densities around artificial water-points and causing competitive exclusion of the less common chocolate-brown hyena (Hyaena brunnea) [71]. Indeed, the crusade for caution in determining effective spacing of artificial h2o-points for the maintenance of ungulate affluence and diverseness, and ecosystem heterogeneity both within and exterior protected areas has been highlighted extensively in the literature [53], [72], [73]. Such provisioning can also disrupt movement patterns of migratory ungulate species, and result in heightened inter-specific competition and die-off populations of these in dry out years [twenty]. This raises the effect of whether the run a risk of increasing drought severity and frequency over the 21^st century [9] may potentially further exacerbate the issue of artificial water-holes on both the competitive exclusion of rarer, more than sedentary ungulate species and the dry flavour survival of migratory species. Climate change is predicted to alter the timing of migrations and the migration routes of terrestrial mammals, through altering the distribution of forage and surface-water availability [74]–[76]. Hence under hereafter increases in drought intensity and frequency, both wet and dry season ranges of migratory ungulate species may become less able to support these populations and may exist altered, which could have serious implications for the future conservation of such species. Fencing effectually national parks can have a severe touch on on the survival of the migratory populations within, disrupting migratory pathways and heightening the impact of drought through disabling such populations to access their dry flavor ranges [55]–[57]. Under increasing future drought intensity, national parks should focus on the effective spacing of artificial water-points and on enabling greater connectivity for migratory ungulates. Such management strategies would assist in limiting the negative impacts of water-provisioning on both sedentary and migratory populations.

Birthday, our work illustrates that climatic change and increased drought conditions could lead to the extinction of certain populations over the 21^st century. Our findings provide further evidence that increasing futurity drought atmospheric condition will pose a greater take chances to ungulate species that are highly sedentary, and that are wholly or partially dependent on drought-intolerant nutrient species. Although none of our study species are threatened species and have been classified by the International Union for the Conservation of Nature (IUCN) as 'Least Business organization' [45]–[47], [77], our findings may have implications for some other highly threatened ungulate species and subspecies in areas where drought intensity is predicted to increase over the 21^st century, such as Mountain zebra (Equus zebra), European bison (Bison bonasus), Cuvier's gazelle (Gazella cuvieri) and Tora hartebeest (Alcelaphus buselaphus tora) [78]–[81]. In addition, these drought-intolerant life history traits are also likely to crusade enhanced susceptibility to increasing future drought intensity of sure species in other taxonomic groups. Our report conspicuously stresses the importance of long-term monitoring in order to provide a ground on which to explore the impacts of extreme natural events on animate being populations under future climate change. Future studies should exist conducted in guild to determine the susceptibility of species in differing environments and taxonomic groups, especially threatened species and pocket-size or isolated populations, to these increased climate extremes, in order to develop appropriate management strategies.

Supporting Information

Figure S1.

C as a predictor of growth rates r for all (a) sedentary, browsing species and (b) migratory or nomadic, grazing or mixed-feeding species. C is the maximum number of consecutive months of the preceding year in which q<θ. For sedentary browsing species, there were six species, n = 373, gradient = −0.01, p = 0.20. For migratory or nomadic, grazing or mixed feeding species, there were 2 species, northward = 165, slope = 0.001, p = 0.88.

https://doi.org/10.1371/journal.pone.0051490.s001

(DOCX)

Tabular array S2.

Probabilities under current weather of a given yr existence in drought, and of different lengths of drought occurring if that's the case, for sedentary grazing or mixed feeding species. (a) Observed probabilities nether current conditions of a given year beingness in drought (initial threshold; C >0) and non in drought (C =0) and that if one year is a drought year that the following year volition also be a drought year (the drought altitude threshold; ddt) or if one year is a non-drought yr that the following year will too be a non-drought year (the non-drought distance threshold; nddt), beyond all populations of all sedentary grazing or mixed feeding species. (b) Observed probabilities under current weather that if a given year is a drought twelvemonth, C will be equal to each value betwixt one and 12, beyond all populations of all sedentary grazing or mixed feeding species.

https://doi.org/ten.1371/periodical.pone.0051490.s003

(DOC)

Table S5.

Relative starting affluence of populations of sedentary, grazing species (SG; relative abundance hither is initial population abundance at year t in comparison to the initial abundance at year t for all populations of that species). Due to confidential information sources, raw abundance values cannot be published.

https://doi.org/10.1371/periodical.pone.0051490.s006

(DOC)

Acknowledgments

We are grateful to Stefanie Deinet for her aid collating the LPI ungulate species information. We would similar to give thanks the two anonymous reviewers for their help in improving the manuscript.

Writer Contributions

Conceived and designed the experiments: NP ALMC. Performed the experiments: CD ALMC NP. Analyzed the data: CD ALMC NP. Contributed reagents/materials/analysis tools: LMM. Wrote the paper: CD NP ALMC.

References

1. ane. Ceballos G, Ehrlich PR (2002) Mammal population losses and the extinction crisis. Science 296: 904–907.
   • View Article
   • Google Scholar
2. 2. Millennium Ecosystem Assessment (2005) Ecosystems and human well-being: biodiversity synthesis. Washington, D.C., United states of america: World Resources Institute. 86 p.
3. iii. Foden W, Mace GM, Vié J-C, Angulo A, Butchart S, et al.. (2008) Species susceptibility to climate change impacts. In: Vié J-C, Hilton-Taylor C, Stuart SN, editors. The 2008 review of the IUCN Red List of Threatened Species. Gland, Switzerland: IUCN/SSC. 77–88.
4. 4. Intergovernmental Panel on Climate change (2007) Climate alter 2007: The physical scientific discipline basis. Cambridge, United Kingdom and New York, USA: Cambridge University Press. 996 p.
5. v. Immature TP (1994) Natural die-offs of large mammals: implications for conservation. Conserv Biol 8: 410–418.
   • View Article
   • Google Scholar
6. six. Saltz D (2002) The dynamics of equid populations. In: Moehlmann P, IUCN/SSC Equid Specialist Group, editors. Zebra, asses, and horses: an action plan for the conservation of wild equids. Gland, Switzerland: IUCN/SSC. 110 p.
7. 7. Epps CW, McCullough DR, Wehausen JR, Bleich VC, Rechel JL (2004) Furnishings of climate change on population persistence of desert-habitation mountain sheep in California. Conserv Biol eighteen: 102–113.
   • View Article
   • Google Scholar
8. 8. Foley C, Pettorelli Northward, Foley 50 (2008) Astringent drought and dogie survival in elephants. Biol Lett 4: 541–iv.
   • View Article
   • Google Scholar
9. 9. Intergovernmental Panel on Climate Change (2012) Managing the risks of extreme events and disasters to advance climatic change adaptation. A special written report of Working Groups I and II of the Intergovernmental Panel on Climate change. Cambridge, United Kingdom and New York, USA: Cambridge University Printing. 582 p.
10. 10. Armbruster P, Lande R (1993) A population viability analysis of African elephant (Loxodonta africana): how big should reserves exist? Conserv Biol vii: 602–610.
    • View Commodity
    • Google Scholar
11. 11. Owen-Smith N, Mason DR, Ogutu JO (2005) Correlates of survival rates for 10 African ungulate populations: density, rainfall and predation. J Anim Ecol 74: 774–788.
    • View Article
    • Google Scholar
12. 12. Ogutu JO, Piepho HP, Dublin HT, Bhola N, Reid RS (2008) Rainfall influences on ungulate population affluence in the Mara-Serengeti ecosystem. J Anim Ecol 77: 814–829.
    • View Article
    • Google Scholar
13. 13. Augustine DJ (2010) Response of native ungulates to drought in semi-arid Kenyan rangeland. Afr J Ecol 48: 1009–1020.
    • View Article
    • Google Scholar
14. fourteen. Saltz D, Rubenstein DI, White GC (2006) The impact of increased ecology stochasticity due to climatic change on the dynamics of Asiatic wild ass. Conserv Biol 20: 1402–1409.
    • View Article
    • Google Scholar
15. 15. Isaac NJB, Cowlishaw G (2004) How species reply to multiple extinction threats. Philos Trans R Soc Lond B Biol Sci 271: 1135–1141.
    • View Article
    • Google Scholar
16. 16. Collen B, McRae LM, Deinet S, De Palma A, Carranza T, et al. (2011) Predicting how populations decline to extinction. Philos Trans R Soc Lond B Biol Sci 366: 2577–86.
    • View Article
    • Google Scholar
17. 17. Dunham KM (1994) The outcome of drought on the big mammal populations of the Zambezi riverine woodlands. J Zool 234: 489–526.
    • View Commodity
    • Google Scholar
18. eighteen. Kay RNB (1997) Responses of African livestock and wild herbivores to drought. J Barren Environ 37: 683–694.
    • View Article
    • Google Scholar
19. 19. Hillman JC, Hillman AKK (1977) Mortality of wildlife in Nairobi National Park, during the drought of 1973–1974. Afr J Ecol fifteen: 1–18.
    • View Article
    • Google Scholar
20. 20. Knight MH (1995) Drought-related mortality of wildlife in the southern Kalahari and the function of human. Afr J Ecol 33: 377–394.
    • View Commodity
    • Google Scholar
21. 21. Walker BH, Emslie RH, Owen-Smith RN, Scholes RJ (1987) To choose or not to cull: Lessons from a southern African drought. J Appl Ecol 24: 381–401.
    • View Article
    • Google Scholar
22. 22. Rutherford MC (1980) Annual found production-atmospheric precipitation relations in barren and semi-barren regions. S Afr J Sci 76: 53–56.
    • View Article
    • Google Scholar
23. 23. Ogutu JO, Owen-Smith Northward (2003) ENSO, rainfall and temperature influences on extreme population declines among African savanna ungulates. Ecol Letts 6: 412–419.
    • View Article
    • Google Scholar
24. 24. Illius AW, O'Connor TG (2000) Resources heterogeneity and ungulate population dynamics. Oikos 89: 283–294.
    • View Article
    • Google Scholar
25. 25. Georgiadis Northward, Hack M, Turpin K (2003) The influence of rainfall on zebra population dynamics: implications for management. J Appl Ecol twoscore: 125–136.
    • View Article
    • Google Scholar
26. 26. Loveridge AJ, Hunt JE, Murindagomo F, Macdonald DW (2006) Influence of drought on predation of elephant (Loxodonta africana) calves by lions (Panthera leo) in an African wooded savannah. J Zool 270: 523–530.
    • View Commodity
    • Google Scholar
27. 27. Owen-Smith N, Mills MGL (2006) Manifold interactive influences on the population dynamics of a multispecies ungulate assemblage. Ecol Monogr 76: 73–92.
    • View Article
    • Google Scholar
28. 28. Intergovernmental Console on Climate Change (2001) Climate change 2001: The scientific ground. Cambridge, U.k. and New York, U.s.: Cambridge Academy Press. 881 p.
29. 29. AMS (American Meteorological Society) (1997) Meteorological drought-policy argument. B Am Meteorol Soc 78: 847–849.
    • View Article
    • Google Scholar
30. thirty. Heim RR (2000) Drought indices: a review. In: Wilhite DA, editor. Drought: a global assessment. Oxford, United Kingdom: Taylor & Francis. 159–167.
31. 31. Keyantash J, Dracup JA (2002) The quantification of drought: an evaluation of drought indices. B Am Meteorol Soc 83: 1167–1180.
    • View Article
    • Google Scholar
32. 32. Palmer WC (1965) Meteorological drought. Washington, D.C., Us: Atmospheric condition Bureau Inquiry Paper No. 45.
33. 33. Heim RR (2002) A review of twentieth-century drought indices used in the United States. B Am Meteorol Soc 83: 1149–1165.
    • View Commodity
    • Google Scholar
34. 34. Dai A (2011) Drought under global warming: a review. Wiley Interdiscip Rev Clim Modify 2: 45–65.
    • View Commodity
    • Google Scholar
35. 35. Dai A (2011) Characteristics and trends in various forms of the Palmer Drought Severity Index during 1900–2008. J Geophys Res 116: 1–26.
    • View Article
    • Google Scholar
36. 36. Wells N, Goddard S, Hayes MJ (2004) A cocky-calibrating Palmer Drought Severity Alphabetize. J Clim 17: 2335–2351.
    • View Article
    • Google Scholar
37. 37. Rubel F, Kottek M (2010) Observed and projected climate shifts 1901–2100 depicted by world maps of the Köppen-Geiger climate classification. Meteorol Z 19: 135–141.
    • View Article
    • Google Scholar
38. 38. WWF/ZSL (2012) The Living Planet Database. Available: world wide web.livingplanetindex.org. Accessed 2022 Jul 26.
39. 39. Ponce VM, Pandey RP, Ercan S (2000) Characterization of drought beyond climatic spectrum. J Hydrol Eng 5: 222–224.
    • View Article
    • Google Scholar
40. twoscore. Sheffield J, Wood EF (2008) Projected changes in drought occurrence under time to come global warming from multi-model, multi-scenario, IPCC AR4 simulations. Clim Dynam 31: 79–105.
    • View Commodity
    • Google Scholar
41. 41. Team RDC (2012) R: a language and surround for statistical computing, Vienna, Austria. Available: www.R-project.org. Accessed 2022 Feb 29.
42. 42. Wilson DE, Mittermeier RA (2011) Handbook of the mammals of the world - volume 2: hoofed mammals. Barcelona, Spain: Lynx Edicions. 885 p.
43. 43. Erasmus BFN, van Jaarsveld A, Chown SL, Kshatriya M, Wessels KJ (2002) Vulnerability of South African creature taxa to climate change. Glob Alter Biol 8: 679–693.
    • View Article
    • Google Scholar
44. 44. Thuiller Westward, Broennimann O, Hughes G, Alkemade JRM, Midgley GF, et al. (2006) Vulnerability of African mammals to anthropogenic climate change nether conservative land transformation assumptions. Glob Change Biol 12: 424–440.
    • View Article
    • Google Scholar
45. 45. IUCN/SSC Antelope Specialist Group (2008a) Alcelaphus buselaphus. In: IUCN, editors. IUCN Red List of Threatened Species. Version 2022.2. Available: www.iucnredlist.org. Accessed 2022 May 21.
46. 46. IUCN/SSC Antelope Specialist Grouping (2008b) Kobus ellipsiprymnus. In: IUCN, editors. IUCN Carmine List of Threatened Species. Version 2022.2. Bachelor: www.iucnredlist.org. Accessed 2022 May 21.
47. 47. IUCN/SSC Antelope Specialist Group (2008c) Syncerus caffer. In: IUCN, editors. IUCN Red Listing of Threatened Species. Version 2022.ii. Bachelor: world wide web.iucnredlist.org. Accessed 2022 May 21.
48. 48. Spinage CA (1982) A territorial antelope: the Uganda waterbuck. London, UK: Academic Press. 334 p.
49. 49. Chege G, Kisio E (2008) Lewa Wild fauna Salvation Enquiry and Monitoring Annual Report. Nairobi, Kenya: Lewa Wildlife Conservancy. 27 p. Available: www.lewa.org. Accessed 2022 Oct 10.
50. 50. Southward African National Parks (2008) Due south African National Parks Data Repository, SANParks. Available: www.sanparks.org. Accessed 2022 Oct 10.
51. 51. Sinclair ARE, Mduma SAR, Hopcraft JGC, Fryxell JM, Hilborn R, et al. (2007) Long-term ecosystem dynamics in the Serengeti: lessons for conservation. Conserv Biol 21: 580–590.
    • View Article
    • Google Scholar
52. 52. van Dierendonck MC, Wallis de Vries MF (1996) Ungulate reintroductions: experiences with the Takhi or Przewalski equus caballus (Equus ferus przewalskii) in Mongolia. Conserv Biol 10: 728–740.
    • View Article
    • Google Scholar
53. 53. Owen-Smith N (1996) Ecological guidelines for waterpoints in extensive protected areas. S Afr J Wildl Res 26: 107–112.
    • View Article
    • Google Scholar
54. 54. Harrington R, Owen-Smith N, Viljoen PC, Biggs HC, Mason DR, et al. (1999) Establishing the causes of the roan antelope decline in the Kruger National Park, South Africa. Biol Conserv 90: 69–78.
    • View Commodity
    • Google Scholar
55. 55. Ben-Shahar R (1993) Does fencing reduce the carrying-chapters for populations of large herbivores? J Trop Ecol 9: 249–253.
    • View Commodity
    • Google Scholar
56. 56. Boone RB, Hobbs NT (2004) Lines effectually fragments: effects of fencing on large herbivores. Afr J Range Provender Sci 21: 147–158.
    • View Article
    • Google Scholar
57. 57. Mbaiwa BYJE, Mbaiwa OI (2006) The effects of veterinary fences on wild fauna populations in Okavango Delta, Botswana. Methods 12: 17–24.
    • View Commodity
    • Google Scholar
58. 58. Emslie R (2011) Diceros bicornis. In: IUCN, editors. IUCN Red List of Threatened Species. Version 2022.1. Available: www.iucnredlist.org. Accessed 2022 Jul 2.
59. 59. Gasaway WC, Gasaway KT, Berry HH (2006) Persistent low densities of plains ungulates in Etosha National Park, Namibia: testing the nutrient-regulating hypothesis. Can J Zool 74: 1556–1572.
    • View Article
    • Google Scholar
60. 60. Verlinden A (1998) Seasonal motility patterns of some ungulates in the Kalahari ecosystem of Botswana between 1990 and 1995. Afr J Ecol 36: 117–128.
    • View Article
    • Google Scholar
61. 61. Joly Do, Messier F (2004) Testing hypotheses of bison population refuse (1970–1999) in Forest Buffalo National Park: synergism between exotic illness and predation. Tin can J Zool 82: 1165–1176.
    • View Commodity
    • Google Scholar
62. 62. Jolles AE, Cooper DV, Levin SA (2005) Hidden furnishings of chronic tuberculosis in African buffalo. Ecology 86: 2358–2364.
    • View Article
    • Google Scholar
63. 63. Bouché P, Lejeune P, Vermeulen C (2012) How to count elephants in Westward African savannahs? Synthesis and comparison of main game count methods. Biotechnol Agron Soc xvi, 77–91.
64. 64. Brockett BH (2002) Accuracy, bias and precision of helicopter-based counts of black rhinoceros in Pilanesberg National Park, S Africa. South Afr J Wildl Res 32, 121–136.
65. 65. Jachmann H (2001) Estimating abundance of African wildlife: an help to adaptive direction. Boston, USA: Kluewer Bookish Publishers. 285 p.
66. 66. Magin CD (1989) Variability in full ground counts on a Kenyan game ranch. Afr J Ecol 27, 297–303.
67. 67. Coulson T, Catchpole EA, Albon SD, Morgan BJT, Pemberton JM, et al. (2001) Age, sex, density, winter atmospheric condition, and population crashes in Soay sheep. Science 292: 1528–1531.
    • View Commodity
    • Google Scholar
68. 68. Leirs H, Stenseth NC, Nichols JD, Hines JE, Verhagen R, et al. (1997) Stochastic seasonality and nonlinear density-dependent factors regulate population size in an African rodent. Nature 389: 178–180.
    • View Article
    • Google Scholar
69. 69. Clutton-Brock TH (1988) Reproductive success- studies of private variation in contrasting breeding systems. Chicago: University of Chicago Press. 548 p.
70. seventy. Ogutu JO, Owen-Smith N, Piepho HP, Kuloba B, Edebe J (2012) Dynamics of ungulates in relation to climatic and land use changes in an insularized African savanna ecosystem. Biodivers Conserv 21: 1033–1053.
    • View Commodity
    • Google Scholar
71. 71. Gaylard A, Owen-Smith Due north, Redfern J (2003) Surface water availability: implications for heterogeneity and ecosystem processes. In: Du Toit JT, Biggs H, Rogers KH, editors. The Kruger feel: ecology and management of savanna heterogeneity. Washington, D.C., USA: Island Printing. 171–188.
72. 72. Du Toit JT, Cumming DHM (1999) Functional significance of ungulate diversity in African savannas and the ecological implications of the spread of pastoralism. Biodivers Conserv 8: 1643–1661.
    • View Article
    • Google Scholar
73. 73. de Leeuw J, Waweru MN, Okello OO, Maloba M, Nguru P, et al. (2000) Distribution and diverseness of wildlife in northern Kenya in relation to livestock and permanent h2o points. Biol Conserv 100: 297–306.
    • View Article
    • Google Scholar
74. 74. UNEP/CMS (2006) Migratory species and climate change: impacts of a changing environs on wild animals. Bonn, Deutschland: UNEP/CMS Secretariat. 63 p.
75. 75. Berger J (2004) The last mile: how to sustain long-distance migration in mammals. Conserv Biol 18: 320–331.
    • View Article
    • Google Scholar
76. 76. Robinson R, Crick HQP, Learmonth JA, Maclead IMD, Thomas CD, et al. (2009) Travelling through a warming earth: climatic change and migratory species. Endanger Species Res seven: 87–99.
    • View Article
    • Google Scholar
77. 77. IUCN/SSC Antelope Specialist Group (2008d) Aepyceros melampus. In: IUCN, editors. IUCN Ruby List of Threatened Species. Version 2022.2. Available: www.iucnredlist.org. Accessed 2022 May 21.
78. 78. IUCN/SSC Antelope Specialist Group (2008e) Alcelaphus buselaphus ssp. tora. In: IUCN, editors. IUCN Cherry List of Threatened Species. Version 2022.i. Available: www.iucnredlist.org. Accessed 2022 Jul ii.
79. 79. Mallon DP, Cuzin F (2008) Gavella cuvieri. In: IUCN, editors. IUCN Red List of Threatened Species. Version 2022.i. Available: www.iucnredlist.org. Accessed 2022 Jul 18.
80. 80. Novellie P (2008) Equus zebra. In: IUCN, editors. IUCN Red List of Threatened Species. Version 2022.1. Available: www.iucnredlist.org. Accessed 2022 Jul 2.
81. 81. Olech W (2008) Bison bonasus. In: IUCN, editors. IUCN Ruby-red List of Threatened Species. Version 2022.1. Available: www.iucnredlist.org. Accessed 2022 Jul xviii.

Is Severe Drought Increase Or Decrease In Carrying Capacity,

Source: https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0051490

Posted by: merrittshowle.blogspot.com

Share this post