American Journal of Primatology 73:1160–1168 (2011)

RESEARCH ARTICLE White Monkey Syndrome and Presumptive Copper Deficiency in Wild Savannah Baboons

A. CATHERINE MARKHAM1Ã, LAURENCE R. GESQUIERE1, JEAN-PHILIPPE BELLENGER2,3, SUSAN C. ALBERTS4,5, 1,5 AND JEANNE ALTMANN 1Department of Ecology and Evolutionary Biology, Princeton University, Princeton, New Jersey 2Department of Geosciences, Princeton Environmental Institute, Princeton University, Princeton, New Jersey 3De´partement de Chimie, Universite´ de Sherbrooke, Sherbrooke, Que´bec, Canada 4Biology Department, Duke University, Durham, North Carolina 5Institute for Primate Research, National Museums of Kenya, Nairobi, Kenya In immature wild savannah baboons (Papio cynocephalus), we observed symptoms consistent with copper (Cu) deficiency and, more specifically, with a disorder referred to as white monkey syndrome (WMS) in laboratory primates. The objectives of this study were to characterize this pathology, and test three hypotheses that (1) Cu deficiency may have been induced by zinc (Zn) toxicity, (2) it may have been induced by molybdenum (Mo) toxicity, and (3) cumulative rainfall during the perinatal period and particularly during gestation is an ecological factor distinguishing infants afflicted with WMS from non- WMS infants. During 2001–2009, we observed 22 instances of WMS out of a total 377 live births in the study population. Visible symptoms exhibited by WMS infants included whitening of the animal’s fur and/or impaired mobility characterized by an apparent ‘‘stiffening’’ of the hindlimbs. Occurrence of WMS did not vary significantly by gender. However, among individuals that survived at least 180 days, WMS males had a significantly lower survivorship probability than non-WMS males. Zn/Cu ratios assessed from hair samples of adult female baboons were higher in females who had produced at least one WMS offspring relative to females who had not had a WMS offspring. This was true even when the hair sample was collected long after the birth of the female’s afflicted infant. We consider this potentially indicative of a robust tendency for low Cu levels induced by elevated Zn intake in some individuals. No significant differences of Mo/Cu ratios were observed. Cumulative rainfall during gestation (179 days) was 50% lower for WMS infants relative to non-WMS infants. In contrast, rainfall for the two classes of infants did not differ in the 180 days before conception or in the 180 days following birth. This finding highlights the importance of prenatal ecological conditions in healthy fetal development with regard to WMS. Am. J. Primatol. 73:1160–1168, 2011. r 2011 Wiley Periodicals, Inc.

Key words: Amboseli; copper deficiency; Papio cynocephalus; white monkey syndrome; wild baboons; zinc toxicity

INTRODUCTION lower limb parathesias, and failure to thrive [reviewed in Culotta & Gitlin, 2000]. Trace minerals are naturally occurring elements Both genetic and environmental factors can lead essential in minute concentrations for optimal to Cu deficiency. For example, Menkes disease is an growth and development of living organisms. Dele- x-linked recessive disorder in humans resulting from terious effects are often observed with either excessive or deficient intake, and imbalances are considered risk factors for several diseases in a wide Contract grant sponsors: American Society of Primatologists; range of species. In particular, extensive research on Animal Behavior Society; International Primatological Society; NIA; Contract grant numbers: R01AG034513-01; Contract grant human and nonhuman subjects has focused on the sponsor: NSF; Contract grant numbers: IBN-0322613; IOS- causes and consequences of deficiencies in the trace 0919200; BCS-0851750; Contract grant sponsor: Sigma Xi. element copper (Cu). Cu is required for numerous ÃCorrespondence to: A. Catherine Markham, Department of cellular processes, including mitochondrial respira- Ecology and Evolutionary Biology, Princeton University, Guyot tion, antioxidant defense, neurotransmitter synth- Hall, Washington Road, Princeton, NJ 08544. E-mail: [email protected] esis, connective tissue formation, and tissue Received 19 January 2011; revised 4 June 2011; revision pigmentation [reviewed in Culotta & Gitlin, 2000]. accepted 19 June 2011 Cu deficiencies are associated with symptoms, DOI 10.1002/ajp.20983 including immune system dysfunction, anemia, Published online 6 September 2011 in Wiley Online Library (wiley pigmentation loss in the skin and hair, gait difficulty, onlinelibrary.com). r 2011 Wiley Periodicals, Inc. Baboon White Monkey Syndrome / 1161

disruptions to the body’s pathways for Cu absorption ratio in the same hair samples. For both hypotheses, and transport; severity of symptoms varies among we predicted that mineral ratios would be higher in patients and can include hypopigmentation, anemia, individuals associated with WMS (reflecting high Zn neurological defects, connective tissue defects, and or Mo concentration relative to the concentration of distinctively brittle hair [reviewed in de Bie et al., Cu) compared with the ratios in individuals with no 2007; Cox et al., 2002; Vonk et al., 2008]. Cu WMS association. Finally, because (1) rain in the deficiency may also be acquired owing to environ- semi-arid Amboseli ecosystem is a key environmen- mental factors, either (1) directly as a result of tal variable influencing the nutritional and repro- inadequacies in an organism’s habitat and/or nutri- ductive status of baboons [e.g. Alberts et al., 2005; tion or (2) indirectly as a result of Cu’s susceptibility Beehner et al., 2006] and (2) gestation is a vulnerable to transport or bioavailability interference from time to mineral imbalances [e.g. Adogwa et al., 1999; other trace elements, notably zinc (Zn) and molyb- Barone et al., 1998; Gambling & McArdle, 2004], we denum (Mo) in the circumstances of this study tested the hypothesis that rainfall in the perinatal [reviewed in McDowell, 2003]. Consequently, ele- period contributes to WMS. Specifically, we predicted vated levels of one or more of these other elements that infants afflicted with WMS experienced lower may result in a secondary Cu deficiency. rainfall regimes during gestation than did non-WMS In nonhuman primates, acquired Cu deficiency infants. induced by Zn toxicity has been associated with a disorder known as white monkey syndrome (WMS). In documented cases of Zn/Cu imbalances in rhesus METHODS macaques (Macaca mulatta) [Obeck, 1978] and All project protocols complied with regulations baboons (Papio spp.) [Frost et al., 2004], afflicted in Kenya (Republic of Kenya Research Permits individuals were juveniles characterized by alopecia, NCST/5/002/R/776 to J.A. and NCST/5/002/R/777 to dehydration, emaciation, cachexia, dermatitis, diar- S.C.A.) and in the United States (Princeton Uni- rhea, and whitening of hair, skin, and mucous versity IACUC 1649), and adhered to the American membranes. Researchers attributed the occurrence Society of Primatologists Principles for the Ethical of these symptoms to elevated Zn intake originating Treatment of Nonhuman Primates. from the galvanized cages in which the animals were We observed a population of wild savannah housed. Prolonged, untreated exposure to the toxic baboons living in the Amboseli basin, a semi-arid Zn levels resulted in death; symptoms were occa- short grass savannah located in east Africa at the sionally reversible if animals were relocated to northwestern base of Mt. Kilimanjaro. Alberts et al. nontoxic enclosures. Similar pigmentation loss in [2005] provide a thorough description of the Ambo- hair was also observed in a population of captive seli study site and the broader ecological region. squirrel monkeys (Saimiri sciureus) housed in out- Review of the area’s volcanic geochemistry and clay door pens with galvanized metal fencing and roofs at mineralogy are described by Hay et al. [1995] and the Sabana Seca Field Station of the Caribbean Stoessell and Hay [1978]. Maskall and Thornton Primate Research Center, University of Puerto Rico. [1996] provide an overview of the trace mineral However, no tests were performed to confirm the content of soils in Amboseli; elevated Mo concentra- conditions were directly attributable to Zn/Cu im- tion is of particular relevance. balances [Matthew Kessler, personal communica- Savannah baboons are a highly social species tion, February 3, 2011]. obligated to group living. Many aspects of their Here, we report the occurrence of symptoms sociality have been studied extensively, including consistent with WMS and, more broadly, with Cu patterns characteristic of typical infant care and deficiency in wild savannah baboons (Papio cynoce- development [e.g. Altmann, 1980; Altmann et al., phalus). Using long-term observational data avail- 1981; Rhine et al., 1985]. The species is omnivorous able through the Amboseli Baboon Research Project but seasonally available seeds, leaves, pods, and (ABRP), we offer novel documentation and insight fruits comprise the bulk of their diet [e.g. Alberts into the characterization of this pathology in wild et al., 2005; Altmann, 1998; Norton et al., 1987]. primates. Specifically, we tested two hypotheses All baboons within the study population were regarding the probable source of this pathology. individually identifiable by ABRP field researchers, First, we evaluated whether Cu deficiency was and each of the five social groups monitored by induced by Zn toxicity [as reported by Frost et al., ABRP was the focus of detailed observations several 2004; Obeck, 1978] by analyzing the Zn/Cu ratio in days each week. Consequently, conception dates, hair samples obtained from a subset of individuals ages, and the onset of visible pathological symptoms within our study population. Second, because of individuals born into the study population were previous research suggested Cu deficiency in typically accurate to within a few days. Complete east African wildlife may be secondary to toxic details on monitoring effort and data collection Mo concentrations in volcanic soils [Maskall & protocols can be accessed online (http://www.princeton. Thornton, 1991, 1996], we also analyzed the Mo/Cu edu/baboon/).

Am. J. Primatol. 1162 / Markham et al.

For this study, we used data on demography, Corporation, Matthews, NC]. The solution was pathology, and rainfall from 2001 to 2009, a time collected, diluted, spiked with two internal standards period during which we standardized recording of ([In] 5 [Sc] 5 2 ppb) and analyzed for its element observational data on infants/juveniles with regard concentration with an Inductively Coupled Plasma- to whitening of the fur and/or impaired hindlimb Mass Spectrometer (ICP-MS, Element 2, Thermo mobility. Any time a baboon showed pathological Finnigan, Bremen, Germany) at medium resolution. symptoms (e.g. malaise, diarrhea, limping in the absence of a wound), field observers recorded the Statistics animal’s ID, date and time of the observation, and a We evaluated the potential influence of gender description of the animal’s condition. All pathologies on the occurrence of WMS by performing a chi- were rechecked regularly to record any changes in square analysis in STASTISTICA v.5 [StatSoft Inc., the animal’s status. We queried this complete 1995], where expected WMS occurrences were dataset of recorded pathologies to extract the subset derived from observed births per gender class. One of observations used in this study that unambigu- non-WMS individual born in the study period was ously described individuals with whitening fur and/ excluded from this analysis because it died before or impaired hindlimb mobility. ‘‘Ambiguous’’ WMS researchers observed its gender. To test whether instances (N 5 4; fur color that was lighter than survivorship differed between WMS and non-WMS typical but deviation was within the known varia- individuals, we used Kaplan–Meier survival analysis bility in the coat color of ‘‘healthy’’ individuals) were with Mantel Cox log rank test in SPSS 17.0 [SPSS not included in analyses as either WMS or non-WMS. Inc., 2008]; owing to challenges in consistently Rainfall values (mm) were measured daily from a discerning WMS in very young baboons, survivorship rain gauge located centrally within the Amboseli analysis was focused on individuals who survived at basin at the ABRP base camp. least 180 days, an age when coat color changes and Adult hair samples were collected between 2006 extent of independent locomotion makes WMS and 2009, from female baboons darted using a identification reliable. We performed a one-way blowpipe to inject a syringe containing TelazolTM ANOVA in SPSS 17.0 to analyze variation in Zn/Cu (Fort Dodge Animal Health in Fort Dodge, IA) and Mo/Cu ratios for adult females categorized by (tiletamine hydrochloride and zolazepam); for de- whether any of their offspring born within the 9-year tails, see Altmann et al. [1993] and Sapolsky and study period were afflicted with WMS. Females Altmann [1991]. In order to avoid even unlikely categorized as mothers of WMS offspring were not complications affecting the viability of a fetus or necessarily caring for their WMS offspring at the dependent offspring, we selected individuals that, at time their hair was sampled, and they may have the time of darting, were not past the first trimester borne non-WMS offspring at some point as well. To of pregnancy and did not have a young nursing further probe variability in Zn/Cu ratios between infant. A small quantity of hair ( 0.25 g) was cut r females, we paired each female who had had a WMS from the distal flank of the animal’s hind leg; to offspring with a female who had not had a WMS ensure that the sample included the full length of the offspring. We minimized, to the extent possible, the hair shaft, we cut the hair as close to the animal’s time difference in the date of hair sample collection skin as possible. for paired females (median 5 days, range 0–105 Infant and young juvenile baboons were not days). We performed a paired-samples t-test using darted. However, one hair sample was opportunisti- SPSS 17.0 to examine whether Zn/Cu ratios differed cally collected post-mortem from an infant baboon significantly between these two categories of females. who had exhibited WMS symptoms when first Finally, we performed a series of one-way ANOVAs observed at approximately 2 days. The infant was in SPSS 17.0 to test for differences in rainfall during 11 days when she died and her corpse was carried by the 180 days before conception, during gestation her mother for several days post death. This (N 5 373 infants, mean 179 days770.3 SE, range behavior is commonly observed in the immediate 156–194 days), and during the 180 days post-birth for aftermath of a newborn’s death [Altmann, 1980]. WMS infants vs. non-WMS infants. Field observers were able to retrieve the body 3 days after the infant’s death when the mother left it unattended. RESULTS Hair samples were analyzed for total metal Of the 377 infants born in the study population content (Cu, Zn, Mo) by mass spectrometry. Samples during 2001–2009, researchers observed 22 (5.8%) were dried and reduced to powder in liquid nitrogen instances of unambiguous WMS. Occurrences of WMS (Airgas East in Philadelphia, PA) using a mortar and did not differ significantly by sex (males: N 5 9, 40.9%; pestle. Hair samples were then digested using a females: N 5 13, 59.1%; w2 5 1.03, df 5 1, P 5 0.31) microwave-assisted procedure (trace metal grade compared with the sex ratio of non-WMS infants born nitric acid, Fisher Scientific Company in Netwark during the study window. For 17 of 22 individuals DE, 2001C, 20 min) with a Mars 5 digester [CEM afflicted with WMS, detailed data existed on the first

Am. J. Primatol. Baboon White Monkey Syndrome / 1163

date on which the pathology was observed. For these Zn/Cu and Mo/Cu Concentrations in Hair 17 individuals, mean age of observed symptom onset Samples 7 was 245 days ( 42.2 SE, range 2–731 days). Figure 1 Hair of the WMS infant had the highest Zn/Cu shows pictures taken of the youngest documented case ratio (53.6) of any individual we sampled, followed by of WMS in our study; the photographed individual was females with WMS offspring (mean7SE: 37.273.81), born with white fur and was described as suffering and finally females without WMS offspring from poor/weak health, alopecia [Novak & Meyer, (23.472.27; Fig. 2). Hair of the WMS infant also had 2009], and difficulty using her hindlimbs to cling the highest Mo/Cu ratio (0.17); however, mean values ventrally to her mother. Careful visual observation were nearly identical for hair of females who had had confirmed that the infant had normal eye pigmenta- WMS offspring (0.0670.01) and females without WMS tion and was therefore not a ‘‘true’’ albino. Eleven of the WMS infants died before the end of the study period; probable cause of death 60 was pathology in four of these cases and predation 55 in seven. Mean age of death for all WMS infants that did not survive was 709 days (7145.6 SE, 50 range 11–1,597 days); mean age of death for those 45 who died of pathology was 577 days (7190.9 SE, range 11–820 days). Among individuals that survived to at 40 least 180 days, male WMS infants had a significantly 35 lower survivorship probability than non-WMS males Zn/Cu ratio (N 5 154; Mantel Cox log rank, w2 5 8.68, df 5 1, 30 P 5 0.003). We did not observe significant differences 25 in survivorship for WMS vs. non-WMS females 2 20 (N 5 145; Mantel Cox log rank, w 5 0.01, df 5 1, N=20 N=5 N=1 P 5 0.92) or for both sexes combined (N 5 299; Mantel 15 Cox log rank, w2 5 3.26, df 5 1, P 5 0.07). Females without Females with WMS Infant WMS offspring WMS offspring Three mothers had two offspring that developed Fig. 2. Zn/Cu ratio in analyzed hair samples for adult female WMS during the 9-year study period; all other baboons without white monkey syndrome (WMS) offspring mothers of individuals with WMS (N 5 16) only had (N 5 20), female baboons with WMS offspring (N 5 5), and an one afflicted offspring. All but one mother of WMS infant baboon with WMS (N 5 1). Points and error bars represent mean7SE. Zn/CU ratios differed significantly be- offspring (N 5 18) also had Zone non-WMS infant tween females with WMS offspring and females without WMS (range 2–6 infants). offspring (one-way ANOVA, F 5 7.80, df 5 24, P 5 0.01).

Fig. 1. Photos in left column depict an infant with white monkey syndrome (WMS); images are of the youngest documented case of WMS within the Amboseli baboon population. For comparison, photos in right column depict ‘‘healthy’’ infants (no observed pathologies, including WMS) of comparable age. Note that the individual featured in the top right was from a captive Papio spp. population (photo of an individual in the same position from study population was not available).

Am. J. Primatol. 1164 / Markham et al.

offspring (0.0770.00). Zn/Cu ratios differed signifi- cumulative rainfall during three time periods: 180 cantly between females with WMS offspring and days preconception, gestation (179 days), and 180 females without WMS offspring (one-way ANOVA, days post-birth. Cumulative rainfall differed signifi- F 5 7.80, df 5 24, P 5 0.01). In contrast, the compar- cantly between WMS infants and non-WMS infants able model predicting Mo/Cu ratios was not significant during gestation (one-way ANOVA, F 5 6.34, (one-way ANOVA, F 5 0.78, df 5 24, P 5 0.39). df 5 372, P 5 0.01), but not in the 180 days before To further probe variability in Zn/Cu values conception (one-way ANOVA, F 5 0.16, df 5 372, between females, we paired each female with WMS P 5 0.69) nor in the 180 days post-birth (one-way offspring with a female without WMS offspring (see ANOVA, F 5 0.57, df 5 372, P 5 0.45; Fig. 3). Methods for pairing details). Consistent with the results from the one-way ANOVA, we found significantly DISCUSSION higher Zn/Cu levels in females with WMS offspring (mean7SE 5 37.273.81) relative to females without Cu is a trace mineral so important for proper WMS offspring (24.975.67; t 5 5.25, df 5 4, P 5 0.006). growth and development that deficiencies significantly Table I provides the trace mineral concentra- compromise the health and survivorship of afflicted tions (Cu, Zn, and Mo) and ratios (Zn/Cu and Zn/Mo) individuals. Our results characterize a pathology in in hair samples of individual baboons in each of the wild savannah baboons that is consistent with three categories evaluated: an infant baboon with symptoms of a disorder known as WMS in laboratory WMS, adult female baboons with WMS offspring, primates and, more broadly, with Cu deficiency as and adult female baboons without WMS offspring. described in a range of species, including humans. We used field observations of physical symptoms and complemented them with a laboratory assessment of Perinatal Ecological Conditions (Cumulative trace mineral concentrations in hair samples. Our Rainfall) approach thus balanced long-term, noninvasive mon- We evaluated variability in ecological conditions itoring with quantitative measures of trace mineral predisposing an infant to WMS by comparing levels in a subset of individuals from our study

TABLE I. Trace Mineral Concentrations (Cu, Zn, and Mo) and Ratios (Zn/Cu and Zn/Mo) in Hair Samples of Baboons

Category, individual’s name Cu (mg/kg) Zn (mg/kg) Mo (mg/kg) Zn/Cu Mo/Cu

WMSÃ infant IDA 2.2 117.3 0.4 53.64 0.17 Adult female with WMSÃ offspring RWA 3.5 165.8 0.2 47.25 0.07 YAI 3.8 169.1 0.3 44.02 0.07 VET 4.0 149.1 0.3 37.27 0.08 HON 6.1 176.7 0.3 28.88 0.05 LOL 4.7 134.7 0.3 28.66 0.06 Adult female without WMSÃ offspring OXY 4.0 172.5 0.3 42.88 0.08 LAO 4.5 176.7 0.3 39.19 0.06 VOT 4.5 157.8 0.3 35.40 0.07 VIN 2.5 88.1 0.3 34.57 0.10 WIP 6.8 222.0 0.3 32.44 0.05 ABB 5.5 168.1 0.5 30.55 0.10 NUT 4.6 137.1 0.5 30.04 0.11 KOL 6.4 159.7 0.4 24.80 0.05 NAP 3.0 72.4 0.1 24.39 0.04 LYM 7.6 169.7 0.3 22.34 0.04 OPH 4.4 92.4 0.2 21.16 0.05 EVA 3.5 65.5 0.2 18.88 0.05 FAX 3.7 66.9 0.2 17.91 0.05 KIW 3.8 65.1 0.2 17.21 0.04 WON 4.0 68.5 0.2 17.11 0.04 SCE 4.7 79.6 0.1 17.10 0.03 DUN 15.0 242.9 0.4 16.15 0.03 WEN 6.2 71.1 0.3 11.49 0.05 OCT 17.0 154.0 0.4 9.03 0.03 NIK 12.1 74.1 0.5 6.15 0.04

ÃWMS, white monkey syndrome.

Am. J. Primatol. Baboon White Monkey Syndrome / 1165

250 [Altmann, 1980, 1998; Altmann et al., 1981]. How- Non-WMS Infants (N=351) 225 WMS Infants (N=22) ever, prior monitoring was of a much smaller study population, incidents were too infrequent to produce 200 a sample size adequate for characterization until 175 recently, and no earlier work associated the pathology with potential trace mineral deficiencies. In all cases, 150 symptoms described herein are consistent with those 125 listed in these prior accounts. 100 One aspect of the pathology that is particularly striking is the extent to which observations of Cumulative rainfall (mm) 75 symptoms were limited to young baboons. Mean 50 age of symptom onset (approximately 8 months) coincides with the age at which the natal pelage of 25 180 d prior Gestation 180 d post most infants transitions from black to golden brown/ to conception birth yellow [Altmann et al., 1981]. Individuals at this age Fig. 3. Cumulative rainfall (mm) during the 180 days before are still largely dependent on their mothers for conception, during gestation (179 days), and during the 180 days nutrition, protection, and long-distance transporta- post-birth for white monkey syndrome (WMS) infants (N 5 22) and non-WMS infants (N 5 351) born in the study population during tion; dietary independence and increasing explora- 2001–2009. Cumulative rainfall differed significantly between WMS tory behavior are achieved later in life when infants and non-WMS infants during gestation (one-way ANOVA, individuals are between 12–18 months [Altmann, F 5 6.34, df 5 372, P 5 0.01), but not in the 180 days before conception (one-way ANOVA, F 5 0.16, df 5 372, P 5 0.69) nor in 1980, 1998; Rhine et al., 1985]. Unweaned infant and the 180 days post-birth (one-way ANOVA, F 5 0.57, df 5 372, juvenile mammals are generally more susceptible P 5 0.45). Points and error bars represent mean7SE. than adults to Cu deficiencies, resulting from the low mineral content of milk [Blood et al., 1983; Salih population for whom samples were available, and et al., 1987] coupled with elevated Cu requirements therefore could be evaluated retrospectively. relative to the requirements of individuals in older Hair analysis has been widely used for trace age classes [Mills, 1966; Sakai et al., 2000]. When mineral studies in both humans and nonhumans milk is inadequate to provide sufficient Cu intake, [e.g. Aydin, 2008; Bencko, 1995; Combs, 1987; the onset of clinical symptoms coincides with deple- Holmes et al., 2003; Kim & Mahan, 2001; Marriott tion of fetal Cu stores [Blood et al., 1983]. The extent et al., 1986]. This tissue source has a number of to which fetal Cu stores can buffer subsequent advantages, including: (1) the stability of hair as a deficiency—and thus the duration following parturi- biological material, which facilitates storage and tion before symptoms of deficiency are evident—can transport; (2) the high concentrations of trace vary with the mother’s Cu intake during pregnancy. minerals usually found in hair samples and the Cu may be preferentially allocated to the developing resulting high probability of detection, compared fetus by mothers on a diet marginally Cu deficient with concentrations in blood and urine; and (3) the [Xin et al., 1993]. However, chronically low Cu intake capacity of hair to accumulate minerals, and thus during gestation can result in deficiency in both serve as a biomarker of mineral availability during mother and offspring [Barone et al., 1998]. Inade- extended time periods rather than a reflection of quate Cu nutrition of the mother during gestation short-term environmental challenges and nutritional can have a major impact on fetal [Keen et al., 1998] assessment [reviewed in Hambridge, 1982]. Despite and subsequent offspring growth and development these advantages, hair mineral testing can be [Barone et al., 1998; Blood et al., 1983; Gambling & sensitive to sampling procedures, variability in the McArdle, 2004; Prohaska & Bailey, 1993; Prohaska age and sex of sampled individuals, conditions at & Brokate, 2002]. time of collection (e.g. seasonal effects), and exogen- Cu deficiency has been documented in a variety ous contamination [reviewed in Bozsai, 1992; Combs of species in the wild [e.g. Flynn et al., 1977; Shen et al., 1982; Hambridge, 1982; Marriott et al., 1986]. et al., 2010], including some primate species [e.g. Furthermore, hair analyses performed in commer- Rode et al., 2003, 2006]. Deficiencies are commonly cial laboratories potentially yield inconsistent and identified by apparent dietary inadequacies, either unreliable results [e.g. Barrett, 1985; Seidel et al., through low Cu intake or high intake of minerals 2001]. We minimized these potential problems by that interfere with Cu absorption, such as Mo and Zn using consistent field methods for specimen collec- [reviewed in McDowell, 2003]. Trace mineral con- tion, by focusing our sampling and analysis on one centration in forage is influenced both by the mineral age–sex class, and by adhering to a single calibrated content of soils and species-characteristic variation laboratory protocol implemented at Princeton Uni- between plants in mineral uptake. Maskall and versity by the authors. Thornton [1996] suggest that soils of Amboseli may Several cases of WMS have been previously contain toxic levels of Mo, consistent with geochem- documented in the Amboseli baboon population ical analyses of volcanic soils in other savannah

Am. J. Primatol. 1166 / Markham et al.

ecosystems of east Africa [Maskall & Thornton, population. Higher Zn/Cu ratios are expected in 1991]. However, we found no differences in hair Cu-deficient individuals, and our analyses of hair Mo/Cu ratios between females with WMS offspring from a single WMS infant supports findings in and females without WMS offspring. This finding previous research that Cu deficiency induced by Zn does not support the prediction that, based on toxicity underlies WMS symptoms [Frost et al., 2004; Maskall and Thornton [1991, 1996], high Mo Obeck, 1978]. Because individuals afflicted with concentration in volcanic soils of east Africa is a WMS in our population were almost exclusively potential source of subsequent Cu deficiency in this nursing offspring, a strong correlation in trace species. In contrast, our results that hair Zn/Cu mineral toxicity/deficiency between mothers and ratios differed significantly between the same two infants is expected. This prediction is supported by categories of females is consistent with previous our results of higher Zn/Cu ratios in females who had research linking Zn intake with Cu deficiency in a WMS offspring during the study period relative to general [reviewed in McDowell, 2003] and in labora- females who did not. It is noteworthy that mothers of tory primates afflicted with WMS specifically [Frost WMS infants had higher Zn/Cu ratios, despite the et al., 2004; Obeck, 1978]. It further suggests that Zn fact that the hair samples we analyzed from these may be at toxic levels for baboons in this region’s mothers were not collected near the time of the WMS soils or vegetation. Further research is needed to offspring’s birth (a reflection of our post hoc analyses understand the underlying factors that contribute to of hair samples from the subset of animals targeted Cu deficiency in this ecosystem, in particular the as darting candidates for other ABRP research spatial–temporal bioavailability of Cu itself as well as objectives). This suggests the possibility of stable other trace minerals that can induce Cu deficiency. individual differences in vulnerability to low Cu Infants that developed WMS experienced, on levels induced by Zn toxicity, which may then average, half the rainfall during gestation that non- conditionally result in offspring pathology. Whether WMS infants experienced. This supports our hypoth- genetic variation within the population influences a esis that rainfall in the prenatal period distinguishes proclivity to Cu deficiency—and how this is affected WMS and non-WMS individuals, and highlights the by ecological conditions during gestation—is an importance of ecological conditions in healthy fetal important unanswered question. The extent to development with regard to WMS. In particular, which genetic and environmental factors together environmental challenges that the fetus experiences affect Cu deficiency is not known, but may explain may be detrimental even when buffered by favorable why (1) many females with WMS offspring also had conditions before conception, and may also prove non-WMS offspring and (2) some females with irreversible even under favorable postnatal condi- relatively high Zn/Cu ratios did not have WMS tions. Because rainfall is unlikely to have a direct offspring. effect on an animal’s trace mineral concentration, the One of the major limitations we faced in this suggested association between relative mineral con- study was an inability to detect subclinical Cu centrations in the baboon hair and rainfall probably deficiency. Although extreme cases of deficiency were results from an indirect effect propagated by envir- readily identifiable, some individuals may have had onmental factors that influence trace mineral bioa- low Cu levels in the absence of whitening fur and/or vailability. One plausible explanation is that soil impaired hindlimb mobility. Consequently, we may moisture affects the trace mineral concentration in not have visually identified all individuals that plants and soil [Howard et al., 1962; Shuman, 1980]. experienced Cu deficiency. However, our findings Particularly in tropical savannah ecosystems, highly that hair sample analysis may provide a reliable seasonal rainfall patterns can result in either severe biomarker for baboons susceptible to low Cu levels depletion of elements or accumulation to toxic levels suggest that broader application of this technique to [reviewed in Dissanayake & Chandrajith, 1999]. include more individuals may reveal important Thus, animals may be exposed to variable concentra- patterns of Cu limitation in this population. More tions of trace minerals under different rainfall generally, given the known impact Cu deficiency can conditions, even when making the same food-selec- have on immune system function [reviewed in tion decisions. Of particular relevance to savannah Bonham et al., 2002] and the effects of mineral habitats, such as Amboseli, Dharani et al. [2007] availability on population densities [McNaughton, found that Cu concentrations in Acacia xanthophloea 1988; Milewski, 2000; Oates et al., 1990], such were significantly higher in samples collected during additional studies would offer substantial contribu- wet season than dry season months. A. xanthophloea tion to the understanding of nutritional limitations is one of two dominant tree species in our study site on population dynamics in wild animals. and baboons within this population rely heavily upon the species as food resource, consuming its leaves, bark, sap, flowers, and pods [Altmann, 1998]. ACKNOWLEDGMENTS Our results also reveal intriguing interindivi- We are grateful to the government of the dual variability in Zn/Cu ratios within the study Republic of Kenya, to the Kenya Wildlife Services,

Am. J. Primatol. Baboon White Monkey Syndrome / 1167

the staff and wardens of Amboseli National Park, and Blood DC, Radostits OM, Henderson JA. 1983. Veterinary the local community of the Amboseli region. Tremen- medicine, 6th edition. London: Baillie`re Tindall. dous thanks go to ABRP researchers for their Bonham M, O’Connor JM, Hannigan BM, Strain JJ. 2002. The immune system as a physiological indicator of marginal contributions to data collection and dedication in the copper status? British Journal of Nutrition 87:393–403. field: R. Mututua, S. Sayialel, and J.K. Warutere. Bozsai G. 1992. Quality control and assurance in hair analysis. We also thank M. Akinyi, C. Fitzpatrick, N. Learn, Microchemical Journal 46:159–166. L. Roerish, J. Stroud, and J. Tung for their invaluable Combs DK. 1987. Hair analysis as an indicator of mineral assistance. We benefited from A. Williams’ prelimin- status of livestock. Journal of Animal Science 65:1753–1758. Combs DK, Goodrich RD, Meiske JC. 1982. Mineral concen- ary research on WMS in the Amboseli baboon trations in hair as indicators of mineral status: a review. population; her contribution enhanced our under- Journal of Animal Science 54:391–398. standing of this pathology in the region. Finally, Cox DW, Cullen LM, Forbes JR. 2002. Genetic susceptibility to we thank F. Morel for greatly contributing to the heavy metals in the environment. In: Sarkar B, editor. study’s success by making his laboratory available for Heavy metals in the environment. New York: Marcel Dekker, Inc. p 549–586. mineral analysis and by offering helpful suggestions Culotta VC, Gitlin JD. 2000. Disorders of copper transport. during numerous discussions with the authors, and In: Scriver CR, Sly WS, Childs B, Beaudet AL, Valle D, D.C. Dunbar and M.J. Kessler for their comments on Kinzler KW, Vogelstein B, editors. The metabolic and an earlier version of the manuscript. Financial support molecular bases of inherited disease, 8th edition. New York: was provided by American Society of Primatologists McGraw-Hill. p 3105–3126. de Bie P, Muller P, Wijmenga C, Klomp LWJ. 2007. Molecular (to A.C.M.), Animal Behavior Society (to A.C.M.), parthogenesis of Wilson and Menkes disease: correlation of International Primatological Society (to A.C.M.), NIA mutations with molecular defects and disease phenotypes. (R01AG034513-01 to J.A. and S.C.A.), NSF (IBN- Journal of Medical Genetics 44:673–688. 0322613 to J.A. and S.C.A., IOS-0919200 to S.C.A., Dharani N, Onyari JM, Maina DM, Mavuti KM. 2007. The and BCS-0851750 to J.A. and A.C.M.), and Sigma Xi distribution of Cu and Pb levels in soils and Acacia xanthophloea Benth. from Lake Nakuru National Park, (to A.C.M.). Kenya. Bulletin of Environmental Contamination and Toxicology 79:172–177. Dissanayake CB, Chandrajith R. 1999. Medical geochemistry REFERENCES of tropical environments. Earth-Science Reviews 47: Adogwa AO, Alleyne T, Mohammed A. 1999. Degenerative 219–258. changes in the cerebral cortex in swayback disease lambs. Flynn A, Franzmann AW, Arneson PD, Oldemeyer JL. 1977. Alzheimer’s Disease Review 4:19–22. Indications of copper deficiency in a subpopulation of Alberts SC, Hollister-Smith JA, Mututua RS, Sayialel SN, Alaskan moose. The Journal of Nutrition 107:1182–1189. Muruthi PM, Warutere JK, Altmann J. 2005. Seasonality Frost PA, Hubbard GB, Dammann MJ, Snider CL, Moore CM, and long term change in a savanna environment. In: Hodara VL, Giavedoni LD, Rohwer R, Mahaney MC, Brockman DK, van Schaik CP, editors. Seasonality in Butler TM, Cummins LB, McDonald TJ, Nathanielsz PW, primates: studies of living and extinct human and non- Schlabritz-Loutsevitch NE. 2004. White monkey syndrome in human primates. Cambridge: Cambridge University Press. infant baboons (Papio species). Journal of Medical Primatology p 157–195. 33:197–213. Altmann J. 1980. Baboon mothers and infants. Chicago: Gambling L, McArdle HJ. 2004. Iron, copper and fetal develop- University of Chicago Press. ment. Proceedings of the Nutrition Society 63:553 –562. Altmann SA. 1998. Foraging for survival: yearling baboons in Hambridge KM. 1982. Hair analyses: worthless for vitamins, Africa. Chicago: University of Chicago Press. limited for minerals. The American Journal of Clinical Altmann J, Altmann S, Hausfater G. 1981. Physical matura- Nutrition 36:943–949. tion and age estimates of yellow baboons, Papio cynocephalus, Hay RL, Hughes RE, Kyser TK, Glass HD, Liu J. 1995. in Amboseli National Park, Kenya. American Journal of Magnesium-rich clays of the meerschaum mines in the Primatology 1:389–399. Amboseli basin, Tanzania and Kenya. Clays and Clay Altmann J, Schoeller D, Altmann SA, Muruthi P, Sapolsky RM. Minerals 43:455–466. 1993. Body size and fatness of free-living baboons reflect food Holmes AS, Blaxill MF, Haley BE. 2003. Reduced levels of availability and activity levels. American Journal of Prima- mercury in first baby haircuts of autistic children. Interna- tology 30:149–161. tional Journal of Toxicology 22:277–285. Aydin I. 2008. Comparison of dry, wet and microwave Howard DA, Burdin ML, Lampkin GH. 1962. Variation in digestion procedures for the determination of chemical mineral and crude-protein content of pastures at Muguga in elements in wool samples in Turkey using ICP-OES the Kenya highlands. The Journal of Agricultural Science techniques. Microchemical Journal 90:82–87. 59:251–256. Barone A, Ebesh O, Harper RG, Wapnir RA. 1998. Placental Keen CL, Uriu-Hare JY, Hawk SN, Jankowski MA, Daston GP, copper transport in rats: effects of elevated dietary zinc on Kwid-Uribe CL, Rucker RB. 1998. Effect of copper deficiency fetal copper, iron and metallothionein. The Journal of on prenatal development and pregnancy outcome. The Nutrition 128:1037–1041. American Journal of Clinical Nutrition 67S:1003S–1011S. Barrett S. 1985. Commercial hair analysis: science or scam? Kim YY, Mahan DC. 2001. Effect of dietary selenium source, The Journal of the American Medical Association level, and pig hair color on various selenium indices. Journal 254:1041–1045. of Animal Science 79:949–955. Beehner JC, Onderdonk DA, Alberts SC, Altmann J. 2006. The Marriott BM, Smith Jr JC, Jacobs RM, Lee Jones AO, ecology of conception and pregnancy in wild baboons. Rawlins RG, Kessler MJ. 1986. Hair mineral content as an Behavioral Ecology 17:741–750. indicator of mineral intake in rhesus monkeys (Macaca Bencko V. 1995. Use of human hair as a biomarker in the mulatta). In: Rawlins RG, Kessler MJ, editors. The Cayo assessment of exposure to pollutants in occupational and Santiago macaques: history, behavior and biology. environmental settings. Toxicology 101:29–39. New York: State University of New York Press. p 219–231.

Am. J. Primatol. 1168 / Markham et al.

Maskall J, Thornton I. 1991. Trace element geochemistry of indicated by one-zero and instantaneous sampling of soils and plants in Kenyan conservation areas and implica- feeding. International Journal of Primatology 6:491–499. tions for wildlife nutrition. Environmental Geochemistry Rode KD, Chapman CA, Chapman LJ, McDowell LR. 2003. and Health 13:93–107. Mineral resource availability and consumption by colobus in Maskall J, Thornton I. 1996. The distribution of trace and major Kibale National Park, Uganda. International Journal of elements in Kenyan soil profiles and implications for wildlife Primatology 24:541–573. nutrition. In: Appleton JD, Fuge R, McCall GJH, editors. Rode KD, Chapman CA, McDowell LR, Stickler C. 2006. Environmental geochemistry and health with special refer- Nutritional correlates of population density across habitats ence to developing countries. Geological Society Special and logging intensities in redtail monkeys (Cercopithecus Publication No. 113. London: The Geological Society. p 47–63. ascanius). Biotropica 38:625–634. McDowell LR. 2003. Minerals in animal and human nutrition, Sakai T, Wariishi M, Nishiyama K. 2000. Changes in trace 2nd edition. London: Elsevier Press. element concentration in hair of growing children. Biological McNaughton SJ. 1988. Mineral nutrition and spatial concen- Trace Element Research 77:43–51. trations of African ungulates. Nature 334:343–345. Salih Y, McDowell LR, Hentges JF, Mason Jr RM, Wilcox CJ. Milewski A. 2000. Iodine as a possible controlling nutrient for 1987. Mineral content of milk, colostrum, and serum as elephant populations. Pachyderm 28:78–90. affected by physiological state and mineral supplementa- Mills CF. 1966. Trace element deficiency in livestock in tion. Journal of Dairy Science 70:608–612. Europe. World Review of Animal Production 2:51–57. Sapolsky R, Altmann J. 1991. Incidence of hyper- Norton GW, Rhine RJ, Wynn GW, Wynn RD. 1987. Baboon cortisolism and dexamathasone resistance increases with diet: a five-year study of stability and variability in the plant age among wild baboons. Biological Psychiatry 30: feeding and habitat of the yellow baboons (Papio cynoce- 1008–1016. phalus) of Mikumi National Park, Tanzania. Folia Prima- Seidel S, Kreutzer R, Smith D, McNeel S, Gilliss D. 2001. tologica 48:78–120. Assessment of commercial laboratories performing hair Novak MA, Meyer JS. 2009. Alopecia: possible causes and mineral analysis. The Journal of the American Medical treatments, particularly in captive nonhuman primates. Association 285:67–72. Comparative Medicine 59:18–26. Shen X, Li X, Zhang R. 2010. Studies of ‘‘unsteady gait Oates JP, Whitesides GH, Davies AG, Waterman PG, disease’’ of the Tibetan gazelle (Procapra picticaudata). Green SM, Dasilva GL, Mole S. 1990. Determinants of Journal of Wildlife Diseases 46:560–563. variation in tropical forest primate biomass: new evidence Shuman LM. 1980. Effects of soil temperature, moisture, and from West Africa. Ecology 71:328–343. air-drying on extractable manganese, iron, copper, and zinc. Obeck DK. 1978. Galvanized caging as a potential factor in the Soil Science 130:336–343. development of the ‘‘fading infant’’ or ‘‘white monkey’’ Stoessell RK, Hay RL. 1978. The geochemical origin of syndrome. Laboratory Animal Science 28:698–704. sepiolite and kerolite at Amboseli, Kenya. Contributions to Prohaska JR, Bailey WR. 1993. Persistent regional changes in Mineralogy and Petrology 65:255–267. brain copper, cuproenzymes and catecholamines following Vonk WIM, Wijmenga C, van de Sluis B. 2008. Relevance of perinatal copper deficiency in mice. The Journal of Nutri- animal models for understanding mammalian copper home- tion 123:1226–1234. ostasis. The American Journal of Clinical Nutrition Prohaska JR, Brokate B. 2002. The timing of perinatal copper 88S:840S–845S. deficiency in mice influences offspring survival. The Journal Xin Z, Waterman DF, Hemken RW, Harmon RK. 1993. Copper of Nutrition 132:3142–3145. status and requirement during the dry period and early Rhine RJ, Norton GW, Wynn GM, Wynn RD. 1985. Weaning lactation in multiparous Holstein cows. Journal of Dairy of free-ranging infant baboons (Papio cynocephalus)as Science 76:2711–2716.

Am. J. Primatol.