conservation outside protected areas and the effect of human-dominated landscapes on stress hormones...

Contributed Paper

Conservation outside Protected Areas and the Effectof Human-Dominated Landscapes on StressHormones in Savannah ElephantsM. A. AHLERING,∗†§†† J. E. MALDONADO,†‡ L. S. EGGERT,∗ R. C. FLEISCHER,† D. WESTERN,§AND J. L. BROWN∗∗∗Division of Biological Sciences, University of Missouri, 226 Tucker Hall, Columbia, MO 65211, U.S.A.†Center for Conservation and Evolutionary Genetics, Smithsonian Conservation Biology Institute, National Zoological Park, 3001Connecticut Avenue NW, Washington, D.C. 20008, U.S.A.‡Department of Vertebrate Zoology National Museum of Natural History, Smithsonian Institution, MRC 108, Washington, D.C. 20013,U.S.A.§African Conservation Centre, Nairobi, Kenya∗∗Center for Species Survival, Smithsonian Conservation Biology Institute, National Zoological Park, 1500 Remount Road, Front Royal,VA 22630, U.S.A.

Abstract: Biodiversity conservation strategies are increasingly focused on regions outside national protectedareas, where animals face numerous anthropogenic threats and must coexist with human settlements, live-stock, and agriculture. The effects of these potential threats are not always clear, but they could have profoundimplications for population viability. We used savannah elephants (Loxodonta africana) as a case study toassess the physiological stress associated with living in a human-livestock-dominated landscape. We collectedsamples over two 3-month periods in 2007 and 2008. We used fecal DNA to identify 96 individual elephants ina community conservation area (CCA) and measured fecal glucocorticoid metabolite (FGM) concentrationsas a proxy for stress. The CCA is community Maasai land managed for livestock and wildlife. We comparedthe FGM concentrations from the CCA to FGM concentrations of 40 elephants in Amboseli National Park and32 elephants in the Maasai Mara National Reserve, where human settlements and intense livestock grazingwere absent. In the CCA, we found no significant individual differences in FGM concentrations among theelephants in 2007 (p = 0.312) or 2008 (p = 0.412) and no difference between years (p = 0.616). Theelephants in the CCA had similar FGM concentrations to the Maasai Mara population, but Amboseli elephantshad significantly lower FGM concentrations than those in either Maasai Mara or the CCA (Tukey pairwise test,p < 0.001), due primarily to females excreting significantly lower FGM relative to males (p = 0.025). In theCCA, there was no relation among female group size, average pairwise group relatedness, and average groupFGM concentration. We found no clear evidence of chronic stress in elephants living on CCA communal land,which is encouraging for conservation strategies promoting the protection of animals living outside protectedareas.

Keywords: community-based conservation, fecal DNA, fecal glucocorticoids, Loxodonta africana, noninvasivehormone analysis, protected areas

Conservacion Fuera de Areas Protegidas y el Efecto de Paisajes Dominados por Humanos sobre Hormonas delEstres en Elefantes Africanos

Resumen: Las estrategias de conservacion se enfocan cada vez mas hacia regiones fuera de areas nacionalesprotegidas, donde los animales enfrentan numerosas amenazas antropogenicas y deben coexistir con asen-tamientos humanos, ganado y agricultura. Los efectos de esas amenazas potenciales no siempre son claros,pero podrıan tener implicaciones profundas para la viabilidad poblacional. Utilizamos elefantes africanos

††Current address: The Nature Conservancy, 10 Cornell Street Stop 9019, Grand Forks, ND 58202, U.S.A., email [email protected] submitted February 24, 2012; revised manuscript accepted January 31, 2013.

569Conservation Biology, Volume 27, No. 3, 569–575Conservation Biology C© 2013 Society for Conservation BiologyNo claim to original US government worksDOI: 10.1111/cobi.12061

570 Conservation outside Parks

(Loxodonta africana) como un estudio de caso para evaluar el estres fisiologico asociado con vivir en unpaisaje dominado por humanos y ganado. Recolectamos muestras en 2 perıodos de 3 meses en 2007 y 2008.Utilizamos ADN fecal para identificar a 96 elefantes individuales en un area de conservacion comunitaria(ACC) y medimos las concentraciones del metabolito glucocorticoide fecal (MGF) como un indicador deestres. El ACC es tierra Maasai comunitaria manejada para ganado y vida silvestres. Comparamos lasconcentraciones de MGF en el ACC con las concentraciones de MGF de 40 elefantes del Parque NacionalAmboseli y 32 elefantes de la Reserva Nacional Maasai Mara, donde no hay asentamientos humanos nipastoreo intensivo. En el ACC, no encontramos diferencias individuales significativas en las concentracionesde MGF entre los elefantes en 2007 (p = 0.312) ni 2008 (p = 0.412) ni entre anos (p = 0.616). Los elefantesen el ACC tuvieron concentraciones similares de MGF a la poblacion de Maasai Mara, pero los elefantes deAmboseli tuvieron concentraciones de MGF significativamente menores que las de Maasai Mara o el ACC.(prueba pareada de Tukey, p < 0.001), debido principalmente a que las hembras excretan significativamentemenos MGF que los machos (p = 0.025). En el ACC, no hubo relacion entre el tamano del grupo de hembras,la similitud promedio de grupos pareados, ni la concentracion de MGF promedio del grupo. No encontramosevidencias claras de estres cronico en elefantes viviendo en el ACC, lo cual es alentador para las estrategiasde conservacion que promueven la proteccion de animales que viven fuera de areas protegidas.

Palabras Clave: ADN fecal, analisis de hormonas no invasivo, areas protegidas, conservacion basada en comu-nidades, glucocorticoides fecales, Loxodonta africana

Introduction

Biodiversity conservation is increasingly focused on theprotection of species in the human-dominated matrix out-side traditional protected areas. Although protected areasremain the cornerstone of conservation efforts (Stokes etal. 2010), there is increasing recognition that protectedareas are often not large enough to protect all species orto sustain populations (Tabarelli et al. 2010). Conserva-tion of species outside parks requires understanding andaddressing anthropogenic threats and issues not foundinside parks. A major problem is human-wildlife conflict(Newmark et al. 1994), which threatens people and an-imals and erodes support for conservation efforts (Gadd2005). However, not all anthropogenic effects are imme-diately apparent; some may cause physiological changesthat have a delayed effect on the fitness of an animal.Human activities, such as recreation and tourism can havea delayed effect on the survival and physiology of bothmammal and bird species inside protected areas (e.g.,Creel et al. 2002; Ellenberg et al. 2007). Outside parks,animals encounter different types of effects from humanactivities. For example, savannah elephants (Loxodontaafricana) in east Africa are increasingly expanding out-side protected areas into areas primarily used for livestockgrazing. To date, little attention has been given to thephysiological effect of human activities outside parks onelephants or other wildlife.

Measuring glucocorticoid (GC) hormone production,which increases in response to stress, is one way to quan-tify the physiological effect of human activities on wildlife(Palme et al. 2005). In the short term, a stress response isadaptive, but chronically elevated GC levels can lead todecreased fitness (Keay et al. 2006; Tarlow & Blumstein2007). Measuring GC responses to potential stressors forlarge, elusive wildlife species has recently become more

feasible with the development of fecal metabolite anal-ysis techniques (Tarlow & Blumstein 2007). Fecal GCmetabolite (FGM) monitoring has a number of advantagesfor measuring stress responses, including the ease of col-lection, minimal effect of sample collection on the studyanimals, and a smoothing of short-term GC differences(Palme et al. 2005). Fecal GC metabolite concentrationsare related to gut-passage time, which for a large animalcan be many hours, and elevations are an indication ofchronic stress (Wasser et al. 2000).

For effective wildlife conservation, both the direct andindirect effects of anthropogenic activities on animal pop-ulations need to be identified and quantified (Tarlow& Blumstein 2007). We used the savannah elephant toevaluate the indirect physiological effects of living out-side a traditional protected area in a human-livestock-dominated landscape. Savannah elephants are ideal fora case study because like many species not effectivelycovered by protected areas they are wide ranging, mi-gratory animals with large home ranges that require vastareas to maintain viable populations (Armbruster & Lande1993), and knowledge of the direct threats to elephantsoutside parks is fairly well established (Blanc et al. 2003).The greatest direct threats to elephants are illegal killing(Douglas-Hamilton 1987; Blanc et al. 2003; Gobush et al.2008) and human-elephant conflict (Blanc et al. 2003;Gadd 2005; Graham et al. 2010). In the 1970s and 1980s,elephants retreated to protected areas after dramatic de-clines due to poaching (Douglas-Hamilton 1987), butbecause elephant populations are on the rise in someareas and over 80% of elephant habitat is outside nationalprotected areas (Blanc et al. 2003; Gadd 2005), their pop-ulations are expanding beyond park boundaries.

Our goal was to examine the FGM concentrations ofelephants that recently recolonized communal lands out-side the boundaries of protected areas in Kenya. We

Conservation BiologyVolume 27, No. 3, 2013

Ahlering et al. 571

evaluated the effect of human pastoral settlements andlivestock grazing on FGM concentrations by comparingelephants on communal land with elephants living innearby protected areas without dense human settlementsor intense livestock grazing. We hypothesized that ele-phants living in the pastoral landscape would exhibithigher FGM concentrations as a result of increased stressthan those in nearby protected areas. The elephants in theCCA are not targeted for harassment, and poaching is notcommon. However, they do experience frequent encoun-ters with everyday human and livestock activities. Givenrecent findings that females in disrupted social groupsexhibited elevated FGM levels (Gobush et al. 2008), wealso evaluated gender differences and the relation be-tween female group size, average group relatedness, andaverage group FGM concentration.

Methods

Study Area

We focused on a group of elephants recently estab-lished on Maasai community land in the southern RiftValley province of Kenya (2◦S, 35◦E). In 2000, 2 Maa-sai group ranches, Olkiramatian and Shompole, estab-lished a 20,000-ha community conservation area (CCA)for wildlife and ecotourism and as a grass bank for domes-tic livestock during droughts. After decades of absence,elephants began to recolonize the area shortly afterwardand have now been consistently using the area for over 5years. The area is dominated by open savannas with denseacacia woodlands along the riparian areas and higher el-evations on the rift valley escarpment. The Maasai arepastoralists who use the group ranches largely for graz-ing cattle, sheep, and goats, but they also grow cropsin a small area along the Ewaso Nyiro River. Elephantsin the CCA are exposed to dense human settlements,agricultural areas, and intense livestock grazing on a dailybasis.

Collection of Samples

From May through July 2007 and February through April2008, we searched the study area and followed recentelephant tracks to locate fresh dung piles. Dung pilesfrom which we took samples were <12 hours old, andmost were only a few hours old. Effort was made tosample all individuals in the CCA in both years. We alsocollected dung samples from a number of individuals inthe CCA on multiple occasions, which allowed us to eval-uate individual variation.

From each dung pile, we collected 2 samples fromwhich we extracted DNA and determined the concen-tration of FGMs for the elephants. First, we collected asample for DNA analysis from the external region of 1dung bolus and measured the circumference of up to3 boli to estimate the age of the elephant; dung bolus

circumference is correlated with the age of the elephant.We assigned elephants to an age class on the basis ofthe dung bolus criteria of Jachmann and Bell (1984) anddemographic data from Moss (2001). Elephants estimatedto be 1–4 years old were classified as dependent juveniles,5–12 years old as subadults, and individuals over 12 yearsas adults. Unlike the samples from the national parks, sam-ples from the CCA were collected in 2 different years. Tocollect the second sample for hormone analysis, we thenmixed all the dung boli thoroughly to evenly distributethe hormones from the entire pile and took a 10-g sam-ple. Hormone samples were frozen immediately in liquidnitrogen and transferred to a −20 ◦C freezer in Nairobifor storage. Samples for hormonal analyses were treatedwith 2% acetic acid to kill pathogens and refrozen at−20 ◦C prior to exportation to the United States. Sampleswere shipped to the United States overnight on dry iceand stored in a freezer at −20 ◦C. The DNA samples weretreated following Ahlering et al. (2012). Dung sampleswere exported to the United States for analysis under aU.S. Department of Agriculture permit (48529).

To compare elephants in the CCA with elephants inthe 2 adjacent parks, Amboseli and Maasai Mara, we col-lected samples from both parks in June 2007. Sampleswere collected across the parks to capture the range ofhormone variability. We used the same field protocols asin the CCA.

Genetic Analyses

We used the DNA extraction and amplification proto-cols in Ahlering et al. (2012). We used 12 microsatel-lite loci to identify individuals (FH60R, FH94R, FH48R,FH67, FH126, LA4, LA5, LA6R, LafMS02, LaT05, LaT08,LaT13R) (Nyakaana & Arctander 1998; Comstock et al.2000; Eggert et al. 2000; Comstock et al. 2002; Archie etal. 2003; Eggert et al. 2007). We used DNA from a savannaelephant in the North American captive population as apositive control to standardize allele scores. We includeda negative control (no DNA added) in all reactions todetect PCR contamination.

Because fecal DNA is more prone to error than DNAfrom tissue, we used a multiple- tubes approach (Taberletet al. 1996) and the program RelioType (Miller et al. 2002)to reduce the possibility of assigning incorrect genotypes.We confirmed heterozygous genotypes at least 2 timesand all homozygous genotypes at least 5 times. Samplesthat did not amplify the required 2–5 times at each locuswere discarded. We used the Microsatellite Toolkit add-in in Microsoft Excel to identify unique individuals andgenotype recaptures of the same individual (Park 2001).To determine the sex of the individuals, we used themolecular sexing method of Munshi-South et al. (2008).To eliminate the possibility of identifying the sexes incor-rectly, samples sexed as males were confirmed twice, andsamples sexed as females were confirmed 3 times. We



calculated expected (HE) and observed heterozygosity(HO) values and tested for deviations from expectationsunder Hardy-Weinberg equilibrium and linkage disequi-librium in Genepop 4.0 (Raymond & Rousset 1995; Rous-set 2008). We used the program Relatedness (version5.0.8) (Queller & Goodnight 1989) to estimate averagepairwise relatedness.

Hormone Analyses

We measured FGM concentrations using methods val-idated for African elephants (Wasser et al. 2000). Wefroze, lyophilized, crushed, and sifted dung samples toremove fibrous material. About 0.2 g of fecal powderwere extracted with 5 mL of 90% methanol by shakingfor 30 min and centrifuging for 20 min at 3000 rpm toform a pellet of fecal material. After we poured off the su-pernatant, we dried samples under air and reconstitutedthem with 1 mL of 100% methanol. We used a sonicatorfor 15 min to remove dried material from the sides of thetube. We dried samples under air again and reconstitutedthem in 1 mL of dilution buffer for analyses. Extractionefficiencies were evaluated by adding 3H-corticosteroneto the samples before extraction; extraction efficienciesfor all the samples averaged 68%.

We measured FGM concentrations with a commercialcorticosterone 125I radioimmunoassay (ICN Biomedicals,Costa Mesa, California) validated for elephants (Wasser etal. 2000). We followed manufacturer instructions, exceptthat we halved the volumes for all reagents as per stan-dard practice (Millspaugh et al. 2007). Sensitivity of the as-say at maximum binding was 12.5 ng/mL. The intra-assaydifference was 1.90%, and the interassay difference was2.98%. We reported all FGM concentrations in nanogramsper gram (dry weight).

Statistical Analyses

We used data from the multiple detections to examinedifferences in FGM concentrations within individuals inthe CCA. We evaluated individual differences in FGMconcentrations within each year following the methodsof Munshi-South et al. (2008). We performed a nonpara-metric Wilcoxon signed rank test in the program R (RDevelopment Core Team 2008) to determine whetherthe difference between the first and last sample collectedwas significantly different from zero. We performed aSpearman’s nonparametric correlation in R (R Develop-ment Core Team 2008) to test the relation between thedifference in the first and last sample collected per indi-vidual and the number of days between collection dates.

We used analysis of variance (ANOVA) to test theeffects of age, sex, year, and location on FGM concen-tration, and we used a Tukey test to evaluate multiplecomparisons for the main effects. To ensure the year ofcollection did not affect results of the ANOVA, we com-

pared results from 3 separate ANOVAs. First, we usedonly the CCA data from 2007, coinciding with collectiondate of the national park samples. Second, we includedboth years in the full model and included year as a fac-tor level. We used ANOVA to test the CCA populationseparately for the effect of year on FGM concentrations.Because of nonnormality in residuals, for all parametrictests we log transformed the FGM concentration values,which normalized the residuals.

We used a generalized linear model in the program R (RDevelopment Core Team 2008) to evaluate the relationbetween female group size, average group relatedness,and average FGM concentrations for the group. Groupassignments and size were determined by daily monitor-ing of elephant movements (Ahlering et al. 2012). Withinthe CCA, we identified 7 groups of females for whichwe calculated the average pairwise relatedness (Ahleringet al. 2012). For this analysis, we used only the 2008 dataand averaged the FGM data by group for comparison withthe average relatedness. Average group relatedness wascalculated with the program Relatedness (version 5.0.8)(Queller & Goodnight 1989) following Ahlering et al.(2012).

Results

We collected and genotyped 272 samples for which weobtained FGM concentrations. We sampled 40 individualsfrom Amboseli and 32 individuals from Maasai Mara. Weonly detected all individuals within the national parksonce. From the 200 samples collected in the CCA, wedetected 96 individuals. We detected 36 of those 96 indi-viduals (23 males and 13 females) more than once. Mostof the multiple detections were within a year, but 12males and 4 females were detected in both years. Noneof the 12 loci showed evidence of linkage disequilibriumor departure from Hardy-Weinberg equilibrium. We hadsufficient power to distinguish individuals and identifymultiple detections (P[ID]sib = 0.0000624). RelioTypeshowed the probability of the reliability of our genotypesas 0.99 (Waits et al. 2001).

Within years, there was no significant difference be-tween the first and last FGM concentration measured foran individual (2007, V = 26, p = 0.3125; 2008, V = 140,p = 0.412) and no correlation between the FGM con-centration difference and the number of days betweenmeasurements (2007, S = 68.533, p = 0.662; 2008, S =1091.563, p = 0.200). These results indicate there was notrend over time. Additionally, year had no effect on FGMconcentrations in the CCA (F1,110 = 0.253, p = 0.616;Fig. 1) and year did not interact with age or sex (year× sex, F1,108 = 2.778, p = 0.098; year × age, F2,104 =0.249, p = 0.780). Therefore, in subsequent analyses, weused the average FGM value within a year for individuals


Ahlering et al. 573

Figure 1. Fecal glucocorticoid metaboliteconcentrations for savanna elephants in thecommunity conservation area in southern Kenya byyear (2007 and 2008) (lower and upper lines ofboxes, 25th and 75th percentiles, respectively;horizontal line, median; whiskers, maximum andminimum values; points, potential outliers).

Table 1. Model effects from the analysis of variance testing age, sex,site (Amboseli National Park, Maasai Mara National Park, and the Com-munity Conservation Area), and year (2007 and 2008) on fecal gluco-corticoid metabolite concentrations of savannah elephants.

Effect Levels F p

Age 3 1.160 0.316Sex 2 5.302 0.022Site 3 15.865 <0.001Year 2 0.655 0.419

for which we collected multiple samples; all individualswere represented equally in these analyses.

Sex and site significantly affected FGM concentrations(Table 1). Males had significantly higher FGM concentra-tions than females (p = 0.025), and Amboseli elephantshad significantly lower FGM concentrations than both theCCA elephants (p < 0.001) and Maasai Mara elephants(p < 0.001) (Fig. 2). The difference between the sexeswas most pronounced in Amboseli (Fig. 2). Although wetested for interaction effects among all the factors, nonewere significant.

Neither group size of elephants (r [SE] = −0.012[0.059], p = 0.845) nor average pairwise genetic relat-edness (r [SE] = 0.188 [1.218], p = 0.884) were relatedto the average FGM concentration of a group.

Discussion

Overall, our results suggest that elephants living in thehuman-livestock-dominated CCA did not have levels of

Figure 2. Fecal glucocorticoid metaboliteconcentrations for savanna elephants in southernKenya by site (AM, Amboseli; MM, Maasai Mara; OK,Olkiramatian Community Conservation Area) andsex (F, female; M, male) (lower and upper lines ofboxes, 25th and 75th percentiles, respectively;horizontal line, median; whiskers, maximum andminimum values; points, potential outliers).

FGM concentrations above those found for elephantsin surrounding national protected areas, except whencompared with elephants in Amboseli. Specifically, FGMconcentrations in elephants within the CCA were similarto those observed in Maasai Mara elephants. Althoughsite and sex in the full ANOVA model did not interact,it appeared that the sex and site differences may havebeen affected primarily by low FGM concentrations inthe Amboseli females.

Vegetation structure and composition, and thereforediet, were similar among the 3 locations. Therefore, thesevariables were not a likely cause of the difference ob-served in location and sex, but without data to directlycompare habitats, it cannot be ruled out entirely. So-cial status can affect FGM levels (Goymann & Wingfield2004), but because our samples were collected noninva-sively, we do not have direct information on social statusof the individuals. However, in elephants, dominance andsocial status are often correlated with age (Bradshaw etal. 2005), and we found no significant difference in FGMconcentrations on the basis of bolus size (Jachmann &Bell 1984). Given that the difference between Amboseliand the other 2 locations is driven by the low FGM valuesin the Amboseli females, differences between males andfemales are likely contributing to the between locationdifferences. Male elephants have higher FGM levels thanfemales in some locations (Woolley et al. 2008), but it



is unclear why females at Amboseli might exhibit lowerFGM concentrations than females at the other 2 locations.

One possible explanation is that the differences inFGMs are not due to chronic stress but to differencesin baseline levels of GC hormones at these locations. Thepatterns we detected among the different locations inthis study are similar to baseline GC levels reported forelephants at other locations. There was no difference inFGM concentrations between the sexes in Maasai Maraor the CCA. This finding is consistent with what others,using the same assay we used, have found in forest ele-phants (Loxodonta cyclotis) (Munshi-South et al. 2008).The elephants in Maasai Mara and the CCA also behavesimilarly to forest elephants; they are both somewhat elu-sive and quick to retreat. The male and female differencein FGM concentrations detected at Amboseli was similarto what others, also using the corticosterone assay weused, have found in savanna elephants in Pilanesberg Na-tional Park, South Africa (Woolley et al. 2009). Amboselielephants behave similarly to the savannah elephants inSouth Africa. They spend most of their time in open areas.The behavioral differences and different patterns in FGMconcentrations between the sexes suggest there couldbe innate differences in physiology among locations thatresult in different baselines. However, it is also possi-ble that the different patterns in FGM concentrationsare due to current or past differences in disturbances,such as poaching, among the locations. In particular, thearea around Maasai Mara has a much longer history ofheavy poaching (D. Western, personal communication).To what extent these FGM differences are innate versusenvironmentally dependent remains to be determinedand deserves further examination.

The FGM concentrations within individuals in the CCAdid not differ significantly within or between years,which is similar to the few differences found in individualforest elephants in Gabon (Munshi-South et al. 2008).Seasonal FGM variation has been reported for savannaelephants in Tanzania (Foley et al. 2001), but not in SouthAfrica (Woolley et al. 2009). Our sampling periods didnot span more than one dry season in either year, andalthough the sampling intervals did not overlap betweenthe 2 years, they were only 1 month apart, which likelyaccounted for the lack of FGM differences between years.Within the CCA, female group size, average group relat-edness, and average FGM concentration were not related.In contrast, adult female African elephants from groupswith disrupted social structure in Mikumi National Park,Tanzania, had significantly higher FGM concentrations(Gobush et al. 2008). The female groups in the CCA,however, had relatedness patterns similar to what hasbeen reported for a relatively undisturbed population anddid not show signs of a disrupted family structure (Archieet al. 2006; Ahlering et al. 2012).

The FGM concentrations of the elephants in the CCAwere not significantly different from elephants in one

of the major protected areas in Kenya, the Maasai MaraNational Game Reserve. During human-elephant con-flict, such as crop raiding (Ahlering et al. 2011) andillegal killing of matriarchs (Gobush et al. 2008), ele-phants may exhibit elevated FGM concentrations, butwithout these encounters, individuals living outside pro-tected areas in this CCA appear to cope well with thepresence of humans and livestock. Elephants may haveadapted to dealing with the threats and pressures exertedon them in human-dominated landscapes (Munshi-Southet al. 2008). These results have important implicationsfor conservation strategies promoting the expansion oflarge, wide-ranging species outside protected areas. Ourresults also suggest that if different factors and pressuresoperating among locations affect FGM concentrations,then elephants in parks may be just as susceptible topressures resulting from human conflict as those thatinhabit human-dominated landscapes. Further studies in-vestigating the relation between the human and animaluse of a landscape and the effects of their interactions onstress physiology are needed to provide better solutionsfor the protection of species that are increasingly expand-ing into the human-dominated matrix outside traditionalprotected areas.

Acknowledgments

We thank staff at the African Conservation Centre andthe South Rift Association of Land Owners for support,especially A. Muchiru, K. Mwathe, J. Kamanga, J. Muriuki,J. Nyamu, and J. Worden. We especially thank A. Kuseyo,P. Meiponyi, and S. Russell for field help and A. Johnsonand the Eggert lab for lab assistance. This study was sup-ported by the University of Missouri Research Board, theCenter for Conservation and Evolutionary Genetics, andthe Center for Species Survival at the Smithsonian Con-servation Biology Institute, the Smithsonian PostdoctoralFellowship Program, Friends of the National Zoo, theAmerican Philosophical Society, and the Liz Claiborneand Art Ortenberg Foundation.

Literature Cited

Ahlering, M. A., J. E. Maldonado, R. C. Fleischer, D. Western, and L. S.Eggert. 2012. Fine-scale group structure and demography of Africansavannah elephants recolonising lands outside protected areas. Di-versity and Distributions 18:952–961.

Ahlering, M. A., J. J. Millspaugh, R. J. Woods, D. Western, and L. S.Eggert. 2011. Elevated levels of stress hormones in crop-raidingmale elephants. Animal Conservation 14:124–130.

Archie, E. A., C. J. Moss, and S. C. Alberts. 2003. Characterization oftetranucleotide microsatellite loci in the African savannah elephant(Loxodonta africana africana). Molecular Ecology Notes 3:244–246.

Archie, E. A., C. J. Moss, and S. C. Alberts. 2006. The ties that bind:genetic relatedness predicts the fission and fusion of social groups inwild African elephants. Proceedings of the Royal Society of London– Series B: Biological Sciences 273:513–522.


Ahlering et al. 575

Armbruster, P., and R. Lande. 1993. A population viability analysis forAfrican elephant (Loxodonta africana): How big should reservesbe? Conservation Biology 7:602–610.

Blanc, J. J., C. R. Thouless, J. A. Hart, H. T. Dublin, I. Douglas-Hamilton,C. G. Craig, and R. F. W. Barnes. 2003. African elephant statusreport 2002: an update from the African elephant database. Inter-national Union for the Conservation of Nature (IUCN), Species Sur-vival Commission, African Elephant Specialist Group. IUCN, Glad,Switzerland.

Bradshaw, G. A., A. N. Schore, J. L. Brown, J. H. Poole, and C. J. Moss.2005. Elephant breakdown. Nature 433:807.

Comstock, K. E., N. Georgiadis, J. Pecon-Slattery, A. L. Roca, E. A. Os-trander, S. J. O’Brien, and S. K. Wasser. 2002. Patterns of moleculargenetic variation among African elephant populations. MolecularEcology 11:2489–2498.

Comstock, K. E., S. K. Wasser, and E. A. Ostrander. 2000. Polymor-phic microsatellite DNA loci identified in the African elephant (Lox-odonta africana). Molecular Ecology 9:1004–1005.

Creel, S., J. E. Fox, A. Hardy, J. Sands, B. Garrott, and R. O. Peterson.2002. Snowmobile activity and glucocorticoid stress responses inwolves and elk. Conservation Biology 16:809–814.

Douglas-Hamilton, I. 1987. African elephants: population trends andtheir causes. Oryx 21:11–24.

Eggert, L. S., M. A. Ahlering, and S. Manka. 2007. Lessons from geneticcensuses of forest elephants. Proceedings of the 2007 InternationalElephant Conservation and Research Symposium. International Ele-phant Foundation, Orlando, Florida.

Eggert, L. S., U. Ramakrishnan, N. I. Mundy, and D. S. Woodruff. 2000.Polymorphic microsatellite DNA markers in the African elephant(Loxondonta africana) and their use in the Asian elephant (Elephasmaximus). Molecular Ecology 9:2223–2225.

Ellenberg, U., A. N. Setiawan, A. Cree, D. M. Houston, and P. J. Seddon.2007. Elevated hormonal stress response and reduced reproductiveoutput in yellow-eyed penguins exposed to unregulated tourism.General and Comparative Endocrinology 152:54–63.

Foley, C. A. H., S. Papageorge, and S. K. Wasser. 2001. Noninvasivestress and reproductive measures of social and ecological pressuresin free-ranging African elephants. Conservation Biology 15:1134–1142.

Gadd, M. E. 2005. Conservation outside of parks: attitudes of localpeople in Laikipia, Kenya. Environmental Conservation 32:50–63.

Gobush, K. S., B. M. Mutayoba, and S. K. Wasser. 2008. Long-term im-pacts of poaching on relatedness, stress physiology, and reproduc-tive output of adult female African elephants. Conservation Biology22:1590–1599.

Goymann, W., and J. C. Wingfield. 2004. Allostatic load, social status andstress hormones: the costs of social status matter. Animal Behaviour67:591–602.

Graham, M. D., B. Notter, W. M. Adams, P. C. Lee, and T. NyumbaOchieng. 2010. Patterns of crop-raiding by elephants, Loxodontaafricana, in Laikipia, Kenya, and the management of human-elephant conflict. Systematics and Biodiversity 8:435–445.

Jachmann, H., and R. H. V. Bell. 1984. The use of elephant droppingsin assessing numbers, occupance and age structure: a refinement ofthe method. African Journal of Ecology 22:127–141.

Keay, J. M., J. Singh, M. C. Gaunt, and T. Kaur. 2006. Fecal gluco-corticoids and their metabolites as indicators of stress in variousmammalian species: a literature review. Journal of Zoo and WildlifeMedicine 37:234–244.

Miller, C., P. Joyce, and L. P. Waits. 2002. Assessing allelic dropout andgenotype reliability using maximum likelihood. Genetics 160:357–366.

Millspaugh, J. J., T. Burke, G. U. S. Van Dyk, R. O. B. Slotow, B. E. Wash-burn, and R. J. Woods. 2007. Stress response of working Africanelephants to transportation and safari adventures. Journal of WildlifeManagement 71:1257–1260.

Moss, C. J. 2001. The demography of an African elephant (Loxodontaafricana) population in Amboseli, Kenya. Journal of the ZoologicalSociety of London 255:145–156.

Munshi-South, J., L. Tchignoumba, J. L. Brown, N. Abbondanza, J. E. Mal-donado, A. Henderson, and A. Alonso. 2008. Physiological indicatorsof stress in African forest elephants (Loxodonta africana cyclotis) inrelation to petroleum operations in Gabon, Central Africa. Diversityand Distributions 14:993–1001.

Newmark, W. D., D. N. Manyanza, D. -G. M. Gamassa, and H. I. Sariko.1994. The conflict between wildlife and local people living adja-cent to protected areas in Tanzania: human density as a predictor.Conservation Biology 8:249–255.

Nyakaana, S., and P. Arctander. 1998. Isolation and characterizationof microsatellite loci in the African elephant, Loxodonta africana.Molecular Ecology 7:1436–1437.

Palme, R., S. Rettenbacher, C. Touma, S. M. EL-Bahr, and E. Mostl. 2005.Stress hormones in mammals and birds: comparative aspects regard-ing metabolism, excretion, and noninvasive measurement in fecalsamples. Annals of the New York Academy of Sciences 1040:162–171.

Park, S. D. E. 2001. Trypanotolerance in West African cattle and thepopulation genetic effects of selection. PhD dissertation. Universityof Dublin, Dublin.

Queller, D. C., and K. F. Goodnight. 1989. Estimating relatedness usinggenetic markers. Evolution 43:258–275.

R Development Core Team. 2008. R: a language and environmentfor statistical computing. R Foundation for Statistical Computing,Vienna.

Raymond, M., and F. Rousset. 1995. GENEPOP (version 1.2): popula-tion genetics software for exact tests and ecumenicism. Journal ofHeredity 86:248–249.

Rousset, F. 2008. Genepop 2007: a complete reimplementation of theGenepop software for Windows and Linux. Molecular Ecology Re-sources 8:103–106.

Stokes, E. J., et al. 2010. Monitoring great ape and elephant abun-dance at large spatial scales: measuring effectiveness of a con-servation landscape. PLoS Biology 5: DOI:10.1371/journal.pone.0010294.

Tabarelli, M., A. Venceslau Aguiar, M. Cezar Ribeiro, J. Paul Metzger,and C. A. Peres. 2010. Prospects for biodiversity conservation inthe Atlantic Forest: lessons from aging human-modified landscapes.Biological Conservation 143:2328–2340.

Taberlet, P., S. Griffin, B. Goossens, S. Questiau, V. Manceau, N.Escaravage, L. P. Waits, and J. Bouvet. 1996. Reliable genotypingof samples with very low DNA quantities using PCR. Nucleic AcidResearch 24:3189–3194.

Tarlow, E. M., and D. T. Blumstein. 2007. Evaluating methods to quantifyanthropogenic stressors on wild animals. Applied Animal BehaviourScience 102:429–451.

Waits, L. P., G. Luikart, and P. Taberlet. 2001. Estimating the probabilityof identity among genotypes in natural populations: cautions andguidelines. Molecular Ecology 10:249–256.

Wasser, S. K., K. E. Hunt, J. L. Brown, K. Cooper, C. M. Crockett,U. Bechert, J. J. Millspaugh, S. Larson, and S. L. Monfort. 2000. Ageneralized fecal glucocorticoid assay for use in a diverse array ofnondomestic mammalian and avian species. General and Compara-tive Endocrinology 120:260–275.

Woolley, L.-A., J. J. Millspaugh, R. J. Woods, S. Janse van Rensburg,R. L. Mackey, B. Page, and R. O. B. Slotow. 2008. Populationand individual elephant response to a catastrophic fire in Pi-lanesberg National Park. PLoS ONE 3: DOI:10.1371/journal.pone.0003233.

Woolley, L.-A., J. J. Millspaugh, R. J. Woods, S. J. Van Rensburg, B.R. Page, and R. Slotow. 2009. Intraspecific strategic responses ofAfrican elephants to temporal variation in forage quality. Journal ofWildlife Management 73:827–835.


conservation outside protected areas and the effect of human-dominated landscapes on stress hormones...

Documents