sign in sign up
<<
Home , Archaeolemur, Cryptoprocta spelea, Hadropithecus, Mesopropithecus, Mouse lemur

American Journal of Primatology 73:25–37 (2011)

RESEARCH ARTICLE A Glance to the Past: , Stable Isotopes, Seed Dispersal, and Loss in Southern

1 2Ã 3 BROOKE E. CROWLEY , LAURIE R. GODFREY , AND MITCHELL T. IRWIN 1Department of Anthropology, University of Toronto, Toronto, Ontario, Canada 2Department of Anthropology, University of Massachusetts, Amherst, Massachusetts 3Redpath Museum, McGill University, Montreal, Quebec, Canada The Spiny Thicket Ecoregion (STE) of Southern and southwestern Madagascar was recently home to numerous giant and other ‘‘megafauna,’’ including pygmy hippopotamuses, giant tortoises, elephant birds, and large euplerid carnivores. Following the arrival of humans more than 2,000 years ago, dramatic occurred. Only one-third of the lemur species which earlier occupied the STE survive today; other taxa suffered even greater losses. We use stable isotope biogeochemistry to reconstruct past diets and habitat preferences of the recently extinct lemurs of the STE. We show that the extinct lemurs occupied a wide range of niches, often distinct from those filled by coeval non- . Many of the now-extinct lemurs regularly exploited habitats that were drier than the gallery forests in which the remaining lemurs of this ecoregion are most often protected and studied. Most fed predominantly on C3 plants and some were likely the main dispersers of the large seeds of native C3 trees; others included CAM and/or C4 plants in their diets. These new data suggest that the recent extinctions have likely had significant ecological ramifications for the communities and ecosystems of Southern and southwestern Madagascar. Am. J. Primatol. 73:25–37, 2011. r 2010 Wiley-Liss, Inc.

Key words: seed dispersal; stable isotope; lemur; Spiny Thicket Ecoregion

INTRODUCTION Winter, 1979]. Additionally, there are patches of endemic or native C4 grasses [Bond et al., 2008; If there is any tropical location on earth where Bosser, 1969]. C3 trees also thrive in the STE; these one might be surprised to see nonhuman primates too may have extreme adaptations to aridity, such as that place would be the ‘‘Spiny Thicket Ecoregion’’ swollen trunks or roots for water storage, or reduced (STE) of Southern and southwestern Madagascar leaves. Whereas many of the CAM plants have tiny [Burgess et al., 2004]. This ‘‘spiny desert’’ covers seeds that are adapted for wind dispersal, some C3 44,300 km2, extending from the Mangoky River on trees have large seeds that are adapted for dispersal the west coast to the Anosyennes Mountain chain in via endozoochory. the southeast. Owing to an extreme rain shadow Today, half a dozen lemur species belonging to effect, the STE dry season lasts up to 11 months of four families inhabit the STE. Four of these, ranging the year, and in the driest areas, annual rainfall may in size from 60 g to slightly more than 3 kg, are not exceed 350 mm. Its vegetation is diverse, and common. In the recent past, there were eight varies depending on water sources and the porosity additional lemur species, ranging in size from just and chemistry of the soils [DuPuy & Moat, 1996]. It over 10 kg ( globiceps and Pachylemur includes grasslands, spiny bush, bushy scrub, and insignis)toaround85kg( edwardsi). denser (but usually narrow) gallery or riparian forests bordering streams or rivers, few of which are perennial. The more open ‘‘spiny’’ forests are usually Contract grant sponsor: NSF; Contract grant number: 0129185; only 3–6 m in height, with rare emergent trees Contract grant sponsors: David and Lucille Packard Foundation; exceeding 10 m, creating a discontinuous canopy. John Simon Guggenheim Foundation; Nancy Skinner Clark This ecoregion is renowned for its unusual flora Vassar Graduate Fellowship. with special adaptations to store water or minimize ÃCorrespondence to: Laurie R. Godfrey, Department of Anthro- water loss. Plant is higher here than pology, 240 Hicks Way, University of Massachusetts, Amherst, MA 01003. E-mail: [email protected] anywhere else in Madagascar [Burgess et al., 2004]. Received 17 August 2009; revised 26 January 2010; revision The most dominant plants are succulents (Didier- accepted 27 January 2010 eaceae, Crassulaceae, and Euphorbiaceae) that use DOI 10.1002/ajp.20817 the Crassulacean Acid Metabolism (CAM) photosyn- Published online 4 March 2010 in Wiley Online Library (wiley thetic pathway [Kluge et al., 1991, 1995, 2001; onlinelibrary.com). r 2010 Wiley-Liss, Inc. 26 / Crowley et al.

These species belonged to Madagascar’s famed ‘‘mega- Many lighter elements, such as carbon and nitrogen, fauna,’’ known through their bony remains at ‘‘sub- have isotopes that are stable and do not radioactively fossil’’ (late and Holocene) sites decay. A standardized ‘‘d’’ notation is used to refer to throughout the island. The megafauna, which also the relative proportion of heavy-to-light isotopes, 13 12 included giant ratites, giant tortoises, and pygmy where d 5 ((Rsample/Rstandard)-1)Ã1,000, and R 5 C/ C hippopotamuses, disappeared after humans arrived or 15N/14N. Values are measured in parts per roughly around 2,000 years ago [Burney et al., 2004]. thousand (%). The standards for carbon and nitro- All Madagascar’s extant lemur families (Cheirogalei- gen, respectively, are Pee Dee Belemnite (a marine dae, Lepilemuridae, , , and Dau- carbonate fossil) and air. The different number of bentoniidae) and extinct families (Megaladapidae, neutrons in isotopes of the same element results in Palaeopropithecidae, and Archaeolemuridae) are re- different masses, causing each isotope to behave presented in subfossil deposits throughout the STE. slightly differently during chemical and physical Burney et al. [2003] have shown that pre- reactions; these differences lead to relatively more settlement megafaunal biomass was high in what or less heavy isotope in different materials. When were then open areas, likely reflecting an abundance relatively more of the heavy isotope is present, the of giant tortoises, pygmy hippopotamuses, and ‘‘d’’ value is higher; when relatively less is present, it elephant birds. However, the remains of giant is lower. lemurs have been found alongside those of non- primates, and the extent to which primates were key elements of the various ecosystems of the extreme Carbon South and Southwest is unknown. Although extant Carbon isotope ratios have been used to differ- species have mainly been studied in gallery forests entiate C3 and C4 plants, or the relative contribution (e.g. Beza Mahafaly Special Reserve Parcel I; of these plant groups to diets. Carbon isotope Berenty), lemurs survive today in the STE outside ratios for C3 and C4 plants range from À21 to À35% the gallery forests, and giant lemurs may have done (average À28%), and from À10 to À20% (average the same. Furthermore, whereas there is clear À14%), respectively [O’Leary, 1988]. Plants utilizing anatomical evidence that even the largest-bodied the CAM photosynthetic pathway have intermediate extinct lemurs were adept tree-climbers [Godfrey d13C values, depending on the degree of succulence et al., 2006], none could have been strictly arboreal, and water stress [Ehleringer & Monson, 1993]. even in gallery forests. At Beza Mahafaly, which Nonphotosynthetic plant tissues, such as has both gallery and spiny forest habitats, three and flowers, tend to have d13C values 1–1.5% higher of the four locally surviving lemur species (Lemur than leaves [Codron et al., 2005; Cernusak et al., catta, ring-tailed lemur; Propithecus verreauxi, 2009], but this pattern has not been observed in Verreaux’s ; and Microcebus griseorufus, Madagascar’s dry forests [Crowley et al., submitted; grey-) regularly descend to the Dammhahn & Kappeler, 2009]. Collagen d13C values ground [Brockman et al., 2008; Youssouf Jacky & in herbivores are roughly 5% higher than those in Rasoazanabary, 2008]. Only the , consumed plants, and consumers that eat animal Lepilemur leucopus, is strictly arboreal. matter should have slightly higher d13C values than Here, we use stable isotope biogeochemistry to sympatric herbivores [Koch, 1998]. explore lemur paleoecology in the STE. Specifically, Average foliar d13C values for C3 and CAM we investigate whether the extinct lemurs of this plants from the gallery forest at Beza Mahafaly are region occupied niches likely not filled by sympatric À29.672.6% and À15.370.9%, respectively, and species and how their loss may have affected plant their ranges do not overlap (from À37.3 to À23.5% communities. Stable isotope ratios can be expected to for C3 plants, and from À17.0 to À13.2% for CAM differ among belonging to different trophic plants) [Crowley et al., submitted]. Carbon isotope groups, consuming plants using different photosyn- ratios have not been measured for any C4 plants thetic pathways or living in different habitats. We from the Beza region, but their d13C values should be suggest that, by carefully combining what have higher than those in CAM plants. Assuming a hitherto been rather separate fields—subfossil lemur fractionation of ca. 5% between lemurs and plants paleobiology, isotope geochemistry, lemur ecology, [Crowley et al., submitted; Koch, 1998; Krueger & and floristic surveys of Madagascar’s habitats—we Sullivan, 1984], lemurs eating pure C3 plant diets in might be able to unravel traces of the still-reverber- gallery forest habitat should on average have d13C ating effects of the lemur extinctions. values ca. À24.6%, whereas lemurs incorporating some CAM or C4 foods into their diets should have more positive d13C values. BACKGROUND: INTERPRETING ISOTOPIC However, animals living in water-stressed envir- SIGNALS onments would be expected to have higher d13Cvalues When an element has variable numbers of than animals living in more mesic environments neutrons, the different phases are called ‘‘isotopes.’’ [Ehleringer, 1989; Farquhar et al., 1989]. Although

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 27

mean d13C values for CAM plants (À14.771.4% with 1991; Heaton, 1987; Sealy et al., 1987; Swap et al., arangeofÀ17.4 to À11.4%) from the spiny scrub 2004; Vitousek et al., 1989]. Finally, nitrate-bearing west of the gallery forest at Beza Mahafaly are similar sea spray can increase foliar d15N values [Heaton, to those from the gallery forest [Crowley et al., 1987]. submitted], C3 plants from the spiny scrub do, indeed, The results of moisture stress in animals are have average d13C values that are higher than those in similar to those in plants. can be classified the gallery forest (À26.472.8%, with a range of À30.9 as ‘‘obligate drinkers’’ or ‘‘drought-tolerant’’ to À22.8%). Thus, the difference in the mean carbon [Ambrose & DeNiro, 1986]. Obligate drinkers take isotope ratios between C3 and CAM plants is reduced in adequate amounts of water and do not recycle in the spiny scrub but still substantial. Even in the their urea to the same extent as drought-tolerant arid South of Madagascar, C3 plants have d13C values animals. Consequently, they excrete more 15N and lower than ca. À23%. This is ecologically important their tissues tend to be 15N-depleted relative to because, in Madagascar’s arid South (and likely drought-tolerant animals [Ambrose, 1991]. Severe elsewhere), it seems that all plant species that nutritional stress can lead to animals catabolizing produce nutritious fruits and, therefore, likely depend their own tissues, which also increases tissue d15N on endozoochory for seed dispersal also use the C3 values [Hobson & Clark, 1992; Hobson et al., 1993; photosynthetic pathway [Crowley et al., submitted; Voigt & Matt, 2004]. Thus, although elevated d15N Crowley, unpublished data]. Thus, carbon isotope values could be the result of consumption of animal ratios may provide a means to distinguish consumers matter, they also could reflect differences in nutri- of plants that depend on endozoochory from con- tional or moisture stress experienced by animals, or sumers of plants that use other dispersal methods. isotopic differences in consumed plants. Modern lemurs with d13C values higher than À21.4% There are other possible explanations for varia- (i.e. À26.4% plus 5%) likely incorporate some CAM or tion in d15N values. For example, fire frequency is C4 foods into their diets. inversely correlated with foliar d15N values in semi- Because of 13C changes in atmospheric carbon arid and mesic Australian savannas [Cook, 2001]. dioxide owing to the burning of fossil fuels over Additionally, increased animal biomass could result the past 150 years, subfossil d13C values are in elevated d15N values. Trampling and ammonia approximately 1.2% higher than modern values volatilization associated with urine input (which [Chamberlain et al., 2005]. We, therefore, use a results in preferential loss of 14N from the soil) can conservative cut-off of À20.2% (1.2% higher than affect soil, and consequently plant, d15N values the À21.4% usedformodernanimalsinxericenviron- [e.g. Frank & Evans, 1997]. However, there is ments) to differentiate subfossil species that must have some debate about the degree to which herbivores consumed at least some CAM or C4 matter from those affect plant d15N values and even the directionality that need not have included such plants in their diets. of the presumed effect [e.g. Frank & Evans, 1997; Schulze et al., 1999]. Finally, consumed plant part could also affect d15N values; mean values in Nitrogen can be higher than those in leaves [Codron et al., In most terrestrial systems, plants obtain their 2005; Crowley et al., submitted; Dammhahn & nitrogen from soil nitrate and ammonium, and their Kappeler, 2009]. d15N values are higher than air (which is 0%) In summary, subfossil lemurs that lived in drier [Muzuka, 1999; Schmidt & Stewart, 2003]. C3 plants localities and consumed more fruit or animal matter with symbiotic nitrogen-fixing bacteria have should have higher d15N values than lemurs that d15N values close to 0% because the bacteria fix lived in moist localities and ate less meat or fruit. nitrogen directly from the atmosphere [e.g. Schmidt & Stewart, 2003]. Nitrogen isotope ratios are useful for inferring trophic relationships within a commu- METHODS nity, as herbivore collagen d15N values are 3–5% We sampled d13C and d15N values for 141 higher than the plants they consume, and secondary individuals (104 primates and 37 non-primates) from consumers have collagen nitrogen isotope ratios that subfossil sites in the STE. To our own data, we added are 3–5% higher than herbivores [DeNiro & Epstein, earlier reported d13C and d15N values for 2 indivi- 1981; Fox-Dobbs et al., 2007]. It follows that d15N duals and d13C values for an additional 11 individuals values in a consumer’s tissues should increase with from the STE [Burney et al., 2004; Burney, personal increased faunivory, and decrease with the consump- communication]. Four extant and all eight extinct tion of plants that have symbiotic nitrogen-fixing lemur species were sampled from sites spanning the bacteria [Ambrose, 1991; Schoeninger et al., 1997]. STE from its easternmost (Andrahomana) to its However, these general principles tend to break down southern (e.g. Anavoha) and western (e.g. Andolo- in arid and coastal systems. Decreasing rainfall and nomby) borders. One subfossil site, Taolambiby, is increased salinity often correlate with increasing located well inland (in a region today characterized plant d15N values in non-N fixing plants [Ambrose, by both spiny thicket and gallery forests); all other

Am. J. Primatol. 28 / Crowley et al.

TABLE I. Species Sampled at Each Subfossil Site sites are on or near the coast. Among extinct STE Included in This Study primates, neither stenognathus nor Daubentonia robusta has been found at Taolambiby, Site Specimens sampled (with sample size) although both have been found at inland sites in Anavoha 1 Daubentonia robusta, other ecoregions. Only one specimen of Mesopro- 2 Hadropithecus stenognathus, pithecus has been found at Taolambiby, and it was 3 Megaladapis edwardsi, not sampled. Thus, these three species were sampled 1 M. madagascariensis only at coastal sites. Table I and Figure 1 summarize Andolonomby 1 Lepilemur leucopus,7Archaeolemur our samples by taxon and location. majori,3Lemur catta, Collagen from subfossil bone was isolated 2 M. edwardsi, and analyzed following standardized procedures 5 M. madagascariensis, [Crowley & Godfrey, 2009; Fox-Dobbs et al., 2006]. 2 Pachylemur insignis, Approximately 200 mg of subfossil bone were 14 ingens Andrahomana 2 H. stenognathus,10L. catta, cleaned, fragmented, decalcified in 0.5 M EDTA, 6 Microcebus spp., 3 M. edwardsi, rinsed 10 times in ultrapure water, and gelatinized 2 P. insignis,10Geochelone spp., in 0.01 N HCl at 571C. The gelatin solution was 3 Hypogeomys australis filtered using a 1.5 mm glass-fiber filter and dried Ankazoabo Grotte 6 L. leucopus,2Mesopropithecus under vacuum. After preparation, 0.7 mg of collage- globiceps,2P. ingensus, nous extracts were weighed into tin boats, com- 1 Cryptoprocta spelea busted, and analyzed for d13C and d15N values on a Beavoha 1 M. edwardsi ThermoElectron (Finnigan) Delta1XP continuous SW near Nosy Ve 1 Geochelone spp. flow system connected to an Elemental Analyzer at Taolambiby 2 A. majori,3L. catta,4L. leucopus, the University of California, Santa Cruz Stable 2 M. edwardsi,7M. Isotope Lab. Analytical precision (71 SD) based on madagascariensis,4P. insignis, 7 8 P. ingens,9Propithecus verreauxi, 21 IAEA Acetanilide replicates was À29.1 0.08% 3 ecaudatus, for carbon and 1.170.1% for nitrogen. Collagen 7 Geochelone sp., 8 Hippopotamus integrity was determined using a combination of lemerlei,2C. spelea collagen yield, C:N ratios, and d13C and d15N values Tsiandroina 1 M. globiceps [Ambrose, 1991; Fox-Dobbs et al., 2006].

Fig. 1. Map of the Spiny Thicket Ecoregion of Madagascar, showing the subfossil sites sampled for this study (circles), protected forests (squares), and main cities (triangles).

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 29

Statistical analyses were conducted using if D. robusta, M. globiceps,andH. stenognathus are SPSS 14. For multiple group comparisons, we used removed.) ANOVA with Tukey’s tests of Honestly Significant Ordinal differences in d15N values are highly Differences (HSD), to test the significance of differ- significant, whether considered across the entire STE, ences between particular pairs. Independent sample inland only (at Taolambiby) or along the coast t-tests were also used for comparisons of two (ANOVA for all sites: F 5 7.1, df 5 5, 128, Po0.001; samples. Significance was set at a 5 0.05. for inland only: F 5 2.8, df 5 4, 51, Po0.05; for coastal sites only: F 5 8.9, df 5 3, 74, Po0.001). In general, d15N values are highest at coastal sites (Fig. 2C and D). H. stenognathus has the highest d15Nvalues,buthigh RESULTS values are also found in many other extinct lemurs as Table II presents descriptive statistics for each well as carnivorans and (Table II). At the species sampled. Looking first at differences among inland site (Taolambiby), tenrecs and carnivorans have orders (primates, tortoises, hippos, carnivorans, the highest d15Nvaluesandhipposandtortoiseshave tenrecs, and rodents) across the entire STE, we the lowest; primates are intermediate. Post hoc tests found no significant differences in d13C values. across orders reveal that carnivorans have significantly Carbon isotope ratios in primates overlap those of higher d15N values than primates, tortoises, hippos, non- groups because three extinct species, and rodents but not tenrecs, and primates have D. robusta, M. globiceps, and H. stenognathus significantly higher d15Nvaluesthanhippos.Atcoastal have very high d13C values, equaling or (for sites, carnivorans have the highest d15Nvaluesand H. stenognathus) well surpassing those of non- rodents and tortoises the lowest, with primates inter- primates (Table II, Fig. 2B). All other extinct mediate (post hoc tests confirm significant differences primates in the STE ( majori, between carnivorans and both rodents and tortoises), M. edwardsi, M. madagascariensis, P. insignis,and and primates have significantly higher d15Nvalues Palaeopropithecus ingens)haved13C values at or below than rodents (no hippos were sampled at coastal sites). our À20.2% cut-off for distinguishing diets, including To better understand trophic variation, we CAM or C4 foods. Additionally, variability in d13C grouped tortoises, hippos, and rodents as herbivores, values is greater in coastal sites than inland (cf. and tenrecs and carnivorans as faunivores. Fig. 2A, B) because D. robusta, M. globiceps,and A consistent pattern emerges, whether at coastal or H. stenognathus were sampled only at coastal sites. At inland sites (Table III; Fig. 3): d13C values are lower the inland site (Fig. 2A), primates have significantly in primates than in non-primate herbivores and lower d13C values than hippos, tenrecs, and carnivorans faunivores, although differences among groups (ANOVA: F 5 5.1, df 5 4, 53, Po0.01). (The same approach significance only inland, at Taolambiby applies to primates vs. non-primates at coastal sites (ANOVA: F 5 2.9, df 5 2, 55, P 5 0.06). Variation in

TABLE II. Isotope Data for Subfossil Lemurs (Extinct and Extant) and Associated Fauna

d13C d15N

Species Order NX7SD NX7SD

Extant Lepilemur leucopus Lepilemuridae Primates 7 À20.072.0 7 10.372.2 Lemur catta Lemuridae Primates 14 À18.671.2 14 10.972.4 Propithecus verreauxi Indriidae Primates 9 À19.471.4 9 8.672.0 Microcebus griseorufus Primates 6 À17.871.0 6 11.071.3 Tenrec ecaudatus Tenrecidae Afrosoricida 2 À17.770.9 2 12.171.9 Extinct Archaeolemur majori Archaeolemuridae Primates 9 À20.371.2 9 12.570.9 Daubentonia robusta Daubentoniidae Primates 1 À16.3 1 12.5 Hadropithecus stenognathus Archaeolemuridae Primates 4 À9.270.6 3 15.671.4 Megaladapis edwardsi Megaladapidae Primates 6 À20.372.0 6 11.772.3 M. madagascariensis Megaladapidae Primates 13 À20.271.1 12 11.471.5 Mesopropithecus globiceps Palaeopropithecidae Primates 2 À16.773.7 2 13.272.6 Pachylemur insignis Lemuridae Primates 8 À20.871.2 8 13.071.2 Palaeopropithecus ingens Palaeopropithecidae Primates 26 À20.470.7 24 13.372.0 Cryptoprocta spelea 6 À18.771.4 6 14.771.1 Geochelone spp. Testudinidae Testudines 18 À18.972.8 16 10.471.4 Hippopotamus lemerlei Hippopotamidae Artiodactyla 8 À18.571.2 8 8.972.1 Hypogeomys australis Nesomyidae Rodentia 3 À18.972.0 2 7.970.1

Am. J. Primatol. 30 / Crowley et al.

Fig. 2. Box and whiskers plots of d13C and d15N values for subfossil primates and associated fauna at inland (5Taolambiby) and coastal sites in the STE. Note the extremely high d13C values of Hadropithecus stenognathus (circled, panel B).

TABLE III. Trophic Group Comparisons, with ANOVAs

Primates Herbivores Faunivores ANOVA Comparison Isotope NX7SD NX7SD NX7SD F (df), P

Entire STE Carbon 104 À19.472.5 29 À18.872.3 8 À18.571.3 1.0, (2,138), NS Nitrogen 100 11.972.4 26 9.771.8 8 14.171.7 14.4, (2,131), o0.001 Coastal only Carbon 65 À19.073.0 14 À17.972.3 4 À18.671.8 0.8, (2,80), NS Nitrogen 63 12.872.0 11 10.471.7 4 15.370.7 10.8, (2,75), o0.001 Inland only Carbon 39 À20.171.2 15 À19.772.1 4 À18.370.9 2.9, (2,55), 0.06 Nitrogen 37 10.372.2 15 9.271.7 4 12.871.4 5.2, (2,53), o0.01

15 d N values is highly significant, both inland and on DISCUSSION the coast, with primates consistently falling between herbivores (which have lower values on average) and Did the Extinct Lemurs of the STE Occupy faunivores (which have higher values). Pairwise Niches Likely not Filled by Other Taxa? comparisons of these groups are all significant according to Tukey’s HSD (P 5 0.02 for primates It is evident, on the basis of d13C values that vs. faunivores; Po0.001 for primates vs. non- most of the now-extinct lemurs of the STE exploited primate herbivores and for faunivores vs. herbi- C3 foods (despite the abundance of CAM plants in vores). Mean d15N values are higher in all three the STE), and that the greater exploitation by groups at coastal than inland sites, although sig- lemurs of these resources distinguished these species nificantly so only for primates (Po0.001) and from most sympatric species (herbivores and omni- faunivores (P 5 0.02). vores alike). Among the extinct lemurs, committed

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 31

TABLE IV. Possible Explanations of Elevated d15N Values in Subfossil Lemurs

Explanation Notes and caveats

1. Lemurs were extremely Unlikely. There is little protein stressed and were reason to believe that all catabolizing their own lemurs sampled from tissues subfossil communities would have been protein deprived 2. Lemurs occupied habitats Debatable. Although animal with high animal biomass biomass was likely much higher in the past than it is today, how this would have affected nitrogen isotope ratios is uncertain; furthermore, this cannot account for differences between lemurs and other Fig. 3. Isotopic separation of trophic groups at coastal and inland species sites, showing means and standard deviations for individuals. 3. Fire frequency in the STE Cannot account for was low in the past differences between C3 consumers apparently included P. insignis, lemurs and other species A. majori, P. ingens, M. edwardsi, and M. madagas- 4. Lemur diets were rich in Unlikely, with some possible animal matter exceptions (Daubentonia, cariensis (these taxa all had d13C values À20.2%). o Archaeolemur) Both A. majori and P. insignis have been earlier 5. Lemur diets were rich in Plausible. Anatomical reconstructed as likely large-seed dispersers fruit evidence suggests that [Godfrey et al., 2008a]. However, any of these leaves were the primary large-bodied species could have played an important staple for many of the role in large seed dispersal. extinct lemurs of the STE. Across the STE, all non-primate taxa have mean However, at least three d13C values that are higher than À20.2%. This extinct species, P. insignis, pattern holds for Cryptoprocta spelea and Tenrec A. majori, and D. robusta, ecaudatus, whose d13C values likely reflect some likely consumed large quantities of fruit animal consumption, and Hippopotamus lemerlei, 6. Lemurs were able to thrive Plausible. The degree to Hypogeomys australis, and Geochelone spp., whose 13 in habitats where access to which members of d C values likely reflect some CAM or C4 plant water was limited different orders may have consumption. However, some subfossil lemurs also been drought tolerant 13 have mean d C values above À20.2%. This is true needs further investigation for the extinct D. robusta, M. globiceps, and 7. Lemur diets were rich in Plausible. At inland sites, H. stenognathus as well as the still-extant L. catta, plants that grew in dry water-stress related to dry M. griseorufus, and P. verreauxi. Higher d13C values and/or saline soils; they soils is more likely than are unsurprising for subfossil M. griseorufus and consumed relatively few water-stress related to D. robusta, which were both undoubtedly omnivor- plants with symbiotic salinity. However, soil nitrogen-fixing bacteria salinity may help to ous, as Microcebus spp. and D. madagascariensis are explain very elevated today [Martin, 1972; Harding, 1981; Sauther et al., nitrogen isotope values at 1999; Sussman, 1999]. Modern L. catta are also coastal sites somewhat omnivorous [Sauther et al., 1999], and in certain localities (e.g. Tsimanampesotse and Cap St. What remains to be explained is why extinct Marie), regularly consume CAM plants [Loudon primates of the STE tend to have elevated d15N et al., 2008; Kelley, personal communication]. values relative to co-occurring herbivorous non- M. globiceps’ dental anatomy suggests it was highly primates, such as tortoises, hippos, and rodents. folivorous and it is, therefore, likely that this species’ Table IV offers possible explanations for this phe- elevated d13C values reflect utilization of C4 or CAM nomenon, some of which we regard as more likely plants rather than ingestion of animal matter. than others. Increased soil salinity resulting in The extremely high d13C values of H. stenognathus water-stressed plants may explain high d15N values (whose teeth also suggest dominant herbivory) in animals at coastal sites. But elevated d15N is not indicate it was likely an obligate C4 or CAM limited to coastal sites; lemurs living far from the plant consumer [Godfrey et al., 2005, 2008b; Ryan ocean (e.g. Taolambiby) also have elevated d15N et al., 2008]. values, and the isotopic differences between taxa are

Am. J. Primatol. 32 / Crowley et al.

consistent across sites. It is also unlikely that severe open habitats but less so in riparian forests nutritional stress elevated d15N values in all subfossil [Sussman & Rakotozafy, 1994]. The C3 plants of the lemurs sampled. spiny thickets themselves, through their anachron- Charcoal records attest to large fires across isms, reveal the habitat preferences of certain giant Madagascar for at least the past 40,000 years, but lemurs; many have special adaptations to counter human arrival was accompanied by increased fire overexploitation by climbing animals. Furthermore, it [Burney, 1987; Burney et al., 2004]. Assuming that is likely that large-seeded C3 trees located far from the subfossil communities sampled here largely water must have depended on climbing animals for predate human arrival, an increase in fire frequency seed dispersal. We discuss several examples below. following human arrival could help to explain the 15 elevated d N values in subfossil animals. It is also Effects of Lemur Species Loss possible that a greater megafaunal biomass on the Primates likely have a positive overall effect on landscape in the past is partly responsible for the plant community diversity [Cowlishaw & Dunbar, elevated d15N values in subfossils (problems with 2000]. However, their effects on individual plants this argument are discussed above). However, and plant species can be destructive (predatory) or neither of these explanations can account for the beneficial (assisting reproduction). Folivory, or the observed differences between primates and hippos at consumption of other structural parts or under- the same subfossil sites. ground storage organs, can reduce a plant’s ability to The d15N difference between subfossil lemurs reproduce by forcing a diversion of resources to the and non-primate herbivores might be explained if replacement of lost structures and/or depleting lemurs routinely consumed more fruit and animal banked resources. Consumption of flowers can also matter; certainly, this may account for some of the be detrimental, as it reduces the plant’s seed set, and observed differences. However, elevated d15N values granivory directly destroys seeds. Beneficial effects characterize lemurs with very different reconstructed include pollination (which increases seed set) and diets, including those that likely depended on leaves seed dispersal (transport and deposition of viable from C3 plants (Palaeopropithecus, Megaladapis), C4 seeds, sometimes with increased germination rates). or CAM plants (Hadropithecus, Mesopropithecus), Ecosystem-level benefits also exist. For example, and fruit and animal matter (Pachylemur, Archaeo- consumption of terrestrial herbs and grasses can lemur, Daubentonia). Alternatively or additionally, decrease fire frequency or intensity by reducing the observed isotopic differences could reflect physio- uncropped vegetation [Flannery, 1990; Johnson, logical differences across orders, i.e. that extinct 2009], particularly in dry, fire-prone habitats. The lemurs were as capable as, or more capable than, effects of primate on individual plant non-primate herbivores with regard to surviving in species will, therefore, include both the removal of dry, non-riparian habitats, without frequent drink- destructive and beneficial activities. Because the net ing. For example, African hippos are obligate drin- effect will certainly vary among plant species, kers [Sealy et al., 1987], and one can surmise that extinctions will alter plant community balance. Madagascan hippos would have been the same. In Demonstrating effects of primate extinctions on general (though not universally), obligate drinkers STE plant communities is difficult, fundamentally have lower d15N values than sympatric drought- because we are left with no ‘‘control plots’’ in which tolerant species [Ambrose & DeNiro, 1986; Sealy giant lemurs were spared from extinction. In addi- et al., 1987]. Higher d15N values in extinct lemurs tion, plant species’ extinctions which followed lemur could, therefore, reflect drought tolerance—or an extinctions may be so far undetected. One strategy is ability to survive in very dry habitats. If Madagascan to look for ecological ‘‘mismatches’’ (or ‘‘missing hippos were more dependent on water and vegetation matches’’) in extant ecosystems. near water, one could expect that their d15N values would be lower than those in lemurs that survive on food resources located further from water. Seed Dispersal: Ecological Mismatch in an The notion that extinct lemurs may have thrived Ecosystem Service? in very dry (non-riparian) habitats gains further Seed dispersal is often cited as one of the most support from data collected on the fur of modern important ecosystem services provided by primates mouse lemurs across Madagascar [Crowley et al., [Chapman & Russo, 2006]. Unfortunately, interpre- submitted]. As a rule, variation in mouse lemur d15N tations of temporal changes in seed dispersal in the values is pronounced, with individuals living in drier STE are hampered by the fact that there have been habitats having more positive d15N values than very few systematic studies of seed dispersal by individuals living in moist environments. Differences modern lemurs in this ecoregion. In general, how- across habitats in mouse lemur fur d13C values are ever, compared with primates in other tropical much smaller. Moreover, d13C values for giant forests, Madagascar’s lemurs play a disproportio- lemurs suggest that at least some were feeding on nately large role in seed dispersal owing to the CAM or C4 foods, which are dominant elements in relative lack of frugivorous birds [Voigt et al., 2004].

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 33

Whereas this dynamic may have been different in the Jacky & Rasoazanabary, 2008]. Given the large body past when the elephant birds were still alive, lemurs masses of all the extinct lemurs of the Spiny Thicket (including the giant species) likely played a major (ca. 10–85 kg), there can be no question that all of role in dispersing seeds of plants that did not attract them, including the most arboreal, would have done elephant birds. We know that, in addition to the the same. Indeed, both A. majori and H. stenog- ‘‘wire-plants’’ that are likely adapted to defend their nathus have been reconstructed as largely terres- leaves against consumption by elephant birds, there trial. This means that small bushes as well as tall are many succulent and broad-leaved trees in the trees are candidates for lemur dispersal. STE, some of which are structurally adapted to Second, most reviews of primate seed dispersal defend their leaves against consumption by climbing consider mainly endozoochory [e.g. Chapman & animals [Bond & Silander, 2007]. Because plants Russo, 2006], which seems to be a major mode of build defenses against actual consumers, one can dispersal in most primate communities. However, in infer that many plants with defenses (such as spines) Madagascar’s STE, epizoochory must be considered against climbing animals are today (or were in the as well. The generally water-stressed environment past) exploited by lemurs, and that some of them, likely translates into an increased relative cost of despite defending their leaves against predation, may producing fleshy fruits adapted for endozoochory; actually depend (or have depended) on lemurs to perhaps as a result, some small trees and shrubs in disperse their seeds. this ecoregion [e.g. Uncarina spp., all endemic to A list of STE tree genera that could have Madagascar; Humbert, 1971] have developed fruits depended on giant lemurs for seed dispersal adapted for epizoochory, with large spines and hooks might include Adansonia, Cedrelopsis, Commiphora, for attachment to hair or feathers. Whereas strictly Delonix, Diospyros, Grewia, Pachypodium, Salvadora, arboreal locomotion often involves contact with bare Strychnos, Tamarindus,andUncarina. C3 photo- trunks and branches, semi-terrestrial locomotion synthesis has been confirmed for most of these genera likely involves more incidental contact with leaves, [Crowley, unpublished data; Kluge, personal commu- flowers, and fruit in the forest understory and in nication], but the seed dispersal biology for only a few open areas with abundant shrubs. This would is known. Some (such as Cedrelopsis) are known to be facilitate the attachment of fruits to primates’ hair. primate-dispersed outside the STE [Voigt et al., 2004]. Indeed, it is common to see Uncarina fruits attached Others (such as Delonix) have protein-rich edible pods to the hair of extant lemurs [L. catta at Beza [Grant et al., 1995]. The seeds of the spiny-trunked Mahafaly; Sauther, personal communication]. Pachypodium are today dispersed by wind, but may Athough extant lemurs still certainly disperse have had other means of dispersal in the past seeds (both via endo- and epizoochory), there are [Rapanarivo et al., 1999]. Adansonia, which has fruit indications that the quality of seed dispersal and the with large seeds and nutritious pulp, may have been plant species affected have changed. First, home dispersed by A. majori or P. insignis [Baum, 1995; ranges of extant lemurs are small in the STE Godfrey et al., 2008a]. compared with other ecoregions. The home ranges of Regarding extant lemurs of the STE, Simmen even the two largest STE species are small by primate et al. [2006] showed that L. catta consume and pass standards [L. catta, less than 10 ha, Sussman, 1974; whole seeds of ripe Tamarindus indica fruit and that P. verreauxi, 1–7 ha, Richard, 1978; Jolly, 1966] and seeds in feces germinate. This is unsurprising, as that of the smallest species (M. griseorufus)canbeas comparative data from neighboring ecoregions con- small as 1 ha [Ge´nin, 2008]. The larger extinct lemurs firm that other lemurid species disperse seeds, and must have had larger home ranges. However, the that larger lemurid species tend to disperse larger contrast might not be that great—the largest STE seeds [Ganzhorn et al., 1999; Ralisoamalala, 1996]. lemurs (M. edwardsi at 85 kg and P. ingens at Among the other lemurs, Microcebus spp. and 40 kg) were likely dedicated folivores and not Cheirogaleus spp. disperse small-seeded species. important seed dispersers, and the subfossil lemurs P. verreauxi, a folivore with long gut passage time, that were likely the most important seed dispersers likely destroys most seeds but may occasionally (P. insignis and A. majori) weighed only 11.5 kg and disperse seeds. 18 kg, respectively [Jungers et al., 2008]. Godfrey Understanding the seed dispersal services of and Irwin [2007] reconstructed home ranges of lemurs in the STE is complicated by two factors that between 20 and 40 ha for both M. edwardsi and differ from most primate communities elsewhere. A. majori (the largest folivore and frugivore), using First, the discontinuous canopy necessitates a great Milton and May’s [1976] regressions based on body size deal of terrestrial travel. Some extant species are and diet. These may be overestimates, as giant lemurs quite devoted to terrestrial locomotion [L. catta were less agile than the anthropoids for which Milton spend 30–35% of their time on the ground; Sussman, and May [1976] generated their equations [Godfrey 1974], whereas most other lemur species regularly et al., 2006; Walker et al., 2008], but Archaeolemur was come to the ground to cross open areas [P. verreauxi, the most monkey-like and likely the most agile of all Brockman et al., 2008; M. griseorufus, Youssouf extinct lemurs [Walker et al., 2008].

Am. J. Primatol. 34 / Crowley et al.

Second, and probably more importantly, the past suggests that: (1) the extinct lemurs occupied a small size of extant lemurs relative to extinct lemurs wide range of niches, often distinct from those filled means that some fruits adapted for primate endo- by non-primate taxa; (2) some extinct lemur species zoochory may currently be too big for dispersal by included CAM and/or C4 foods in their diets, but any extant species. We know from studies in rain- most fed predominantly on C3 foods, and may have forest sites that Varecia variegata can swallow seeds been the main dispersers of seeds of endemic and with diameters of more than 30 mm [Dew & Wright, native C3 trees; and (3) many extinct lemurs 1998] and Eulemur fulvus can swallow seeds more regularly exploited arid habitats and were likely than 20 mm diameter [Birkinshaw, 2001]. As an not restricted to gallery forests in which the surviv- animal of more than twice the mass of V. variegata,it ing extant lemur species are most often protected is likely that Pachylemur would have easily swal- and studied today. The combination of elevated d15N lowed seeds in this range—perhaps, considerably values coupled with low d13C values can best be larger [Godfrey et al., 2008a]. Thus, those plant explained as resulting from exploitation of C3 plants species adapted for endozoochory and bearing large- (rather than CAM plants) in dry habitats. seeded fruits are likely candidates for disruption of Deciphering the specific effects of recent lemur dispersal services following lemur extinctions. Spe- extinctions will require further research. The follow- cies with smaller fruits but adapted for endozooch- ing hypotheses, when tested, will refine our under- ory, or with fruits adapted for epizoochory, may be standing of the ecological roles of extinct lemurs: (1) experiencing reduced effectiveness or distance of Plants whose seeds were mainly dispersed by extinct dispersal as well [Johnson, 2009]. lemurs, and whose dispersal has not been co-opted by There is some evidence that trees in western introduced animals, should be rare and/or declining Madagascar currently experience limited dispersal relative to species that use modes of dispersal other services compared with other ecoregions. Voigt et al. than endo- or epizoochory, or whose seeds are [2009] used methods based on population genetics to dispersed either by still-extant lemurs or by intro- determine that the endozoochorous Madagascan dry duced animal species; and (2) such plants should also forest tree Commiphora guillaminii exhibits high have signatures of reduced dispersal distances in their genetic differentiation among subpopulations at the population genetic structure. Specifically, there should local scale relative to a congener in South Africa, but be a discordance between scales, with genetic similar- both species have similar genetic differentiation ity of populations on a regional scale (representing among populations at the regional scale. The high past dispersals) showing relatively low interpopulation local differentiation suggests a much shorter dis- differences, but with genetic differentiation on the persal distance in the recent past for the Madagascan local level (e.g. short parent–offspring distances within species, but the similar differentiation at the larger a single site) reflecting relatively short dispersal scale suggests similar levels of historic gene flow. distances in recent generations. This observed pattern might at least partially reflect Finally, it is important to point out that much the fact that the South African species is still more research attention has been paid to the effects dispersed by the large animals with which it is of habitat changes on lemurs than the opposite— coadapted (frugivorous birds, including hornbills), lemurs’ effects on habitats [see Dausmann et al., but the Madagascan species may have lost its main 2008]. Undoubtedly, the uncertainty facing efforts to dispersers (giant lemurs). Similar studies for other reconstruct the effects of past lemur extinctions species will be crucial for identifying plant species derive largely from the lack of comparative studies whose dispersal relied mainly on extinct lemurs and today. Because of the unique of Madagascar’s is now disrupted. STE flora, we urgently need more quantitative Introduced animal species might act as potential studies of seed dispersal (gut passage time, dispersal dispersers and offset the losses of primate extinc- distance, and seed sizes) in this region before we are tions. Cattle, sheep, and goats regularly graze to make more specific inferences about the seed natural habitats. The wild Potamochoerus larva- dispersal services of extinct lemurs. One potentially tus, thought to be introduced or recently arrived productive area of research is in using local extirpa- from Africa, is also a likely disperser of some seeds. It tion of extant species as an analogy for extinction. will thus be important to consider the attributes of It will be imperative, not just for academic interest native fruits that might predispose them to dispersal but for conservation of imperiled floras, to compare by introduced and domesticated animals. Unfortu- similar habitats with and without extant lemur nately, it is also important to consider mounting species. It will also be important to increase our evidence that introduced animals, especially ungu- knowledge of ecological differences between lemurs lates, tend to facilitate the dispersal of invasive surviving in gallery forests and drier spiny thicket. plants to the detriment of native species [Vavra With an improved understanding of how primate et al., 2007]. extirpations change ecosystem processes, such as In summary, our research on the stable isotope herbivory, pollination, and seed dispersal, we can biogeochemistry of primates in the STE in the recent develop better models for community change which

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 35

we can then apply to past lemur extinctions. This Reich PB, Ellsworth DS, Dawson TE, Griffiths HG, would be especially informative if coupled with Farquhar GD, Wright IJ. 2009. Viewpoint: why are non- studies of the effects of grazing by domestic animals, photosynthetic tissues generally C-13 enriched compared such as cattle, sheep, and goats. with leaves in C3 plants? Review and synthesis of current hypotheses. Functional Plant Biology 36:199–213. Chamberlain CP, Waldbauer JR, Fox-Dobbs K, Newsome SD, ACKNOWLEDGMENTS Koch PL, Smith DR, Church ME, Chamberlain SD, Sorenson KJ, Risebrough R. 2005. Pleistocene to recent Samples were accessed and analyzed with dietary shifts in California condors. Proceedings of the appropriate permission from the curators of the National Academy of Sciences 102:16707–16711. following collections: Oxford Natural History Mu- Chapman CA, Russo SE. 2006. Primate seed dispersal: linking seum; U Antananarivo; Vienna Naturhistorisches behavioural ecology and forest community structure. In: Campbell CJ, Fuentes AF, MacKinnon KC, Panger M, Museum; and U Mass, Amherst. This research Bearder S, editors. Primates in perspective. Oxford: Oxford benefited from discussions with Michelle L. Sauther University Press. p 510–525. and Frank P. Cuozzo and the comments of anon- Codron J, Codron D, Lee-Thorp JA, Sponheimer M, Bond WJ, ymous reviewers. Funding was provided by NSF de Ruiter D, Grant R. 2005. Taxonomic, anatomical, and 0129185 (to David A. Burney, William L. Jungers, spatio-temporal variations in stable carbon and nitrogen isotopic compositions of plants from an African savanna. and L. R. G.), the David and Lucille Packard Journal of Archaeological Science 32:1747–1772. Foundation (Nathaniel Dominy), the John Simon Cook GD. 2001. Effects of frequent fires and grazing on stable Guggenheim Foundation (L. R. G.), and a Nancy nitrogen isotope ratios of vegetation in northern Australia. Skinner Clark Vassar Graduate Fellowship (B. E. C.). Austral Ecology 26:630–636. Cowlishaw G, Dunbar RIM. 2000. Primate conservation This research adhered to the American Society of biology. Chicago: University of Chicago Press. 498p. Primatologists principles for the ethical treatment of Crowley BE, Godfrey LR. 2009. Isotopic variability and lemur dieta nonhuman primates and the laws of Madagascar. dry Malagasy forest: a cautionary tale about vegetation sampling schemes. In: Masters JC, Gamba M, Ge´nin F, editors. Leaping ahead: advances in prosimian biology. New York: Springer. REFERENCES Dammhahn M, Kappeler PM. 2009. Scramble or contest Ambrose SH. 1991. Effects of diet, climate and physiology on competition over food in solitarily foraging mouse lemurs nitrogen isotope abundances in terrestrial foodwebs. Journal (Microcebus spp.): new insights from stable isotopes. of Archaeological Science 18:293–317. American Journal of Physical Anthropology 141:181–189. Ambrose SH, DeNiro MJ. 1986. The isotopic ecology of East DOI: 10.1002/ajp.21129. African mammals. Oecologia 69:395–406. Dausmann KH, Glos J, Linsenmair KE, Ganzhorn JU. 2008. Baum DA. 1995. A systematic revision of Adansonia (Bomba- Improved recruitment of a lemur-dispersed tree in Malagasy caceae). Annals of the Missouri Botanical Garden 82: dry forests after the demise of in forest 440–471. fragments. Oecologia 157:307–316. Birkinshaw CR. 2001. Fruit characteristics of species dis- DeNiro MJ, Epstein S. 1981. Influence of diet on the persed by the black lemur (Eulemur macaco) in the Lokobe distribution of nitrogen isotopes in animals. Geochimica Forest, Madagascar. Biotropica 33:478–486. Cosmochimica Acta 45:341–351. Bond WJ, Silander JA. 2007. Springs and wire plants: Dew JL, Wright PC. 1998. Frugivory and seed dispersal by primates anachronistic defences against Madagascar’s extinct ele- in Madagascar’s eastern rainforest. Biotropica 30:425–437. phant birds. Proceedings of the Royal Society B: Biological DuPuy D, Moat J. 1996. A refined classification of the primary Sciences B 274:1985–1992. vegetation of Madagascar based on the underlying geology: Bond WJ, Silander JA, Ranaivonasy J, Ratsirarson J. 2008. using GIS to map it distribution and to assess its conserva- The antiquity of Madagascar’s grasslands and the rise of the tion status. In: Lourenco WR, editor. Proceedings of the C4 grassy biomes. Journal of Biogeography 35:1743–1758. international symposium on biogeography of Madagascar. Bosser J. 1969. Gramine´es des paturages et des cultures a` Paris: Orstrom. p 205–218. Madagascar. Paris: ORSTOM. 440p. Ehleringer JR. 1989. Carbon isotope ratios and physiological Brockman DK, Godfrey LR, Dollar LJ, Ratsirarson J. processes in aridland plants. In: Rundel PW, Ehleringer JR, 2008. Evidence of invasive Felis sylvestris predation on Nagy KA, editors. Stable isotopes in ecological research. Propithecus verreauxi at Beza Mahafaly Special Reserve, New York: Springer. p 41–54. Madagascar. International Journal of Primatology 29: Ehleringer JR, Monson RK. 1993. Evolutionary and ecological 35–152. aspects of photosynthetic pathway variation. Annual Review Burgess N, D’Amico Hales J, Underwood E, Dinerstein E, of Ecology and Systematics 24:411–439. Olson D, Itoua I, Schipper J, Ricketts T, Newman K. 2004. Farquhar GD, Ehleringer JR, Hubrick KT. 1989. Carbon Terrestrial ecoregions of Africa and Madagascar: a con- isotopic discrimination and photosynthesis. Annual servation assessment. Washington, DC: Island Press. 501p. Review of Plant Physiology and Molecular Biology 40: Burney DA. 1987. Late quaternary stratigraphic charcoal 503–537. records from Madagascar. Quaternary Research 28:274–280. Flannery TF. 1990. Pleistocene faunal loss: implications of the Burney DA, Robinson GS, Burney LP. 2003. Sporormiella and after-shock for Australia’s past and future. Archaeology in the late Holocene extinctions in Madagascar. Proceedings of Oceania 25:45–67. the National Academy of Sciences, United States of America Fox-Dobbs K, Stidham TA, Bowen GJ, Emslie SD. 2006. 100:10800–10805. Dietary controls on extinction versus survival among avian Burney DA, Burney LP, Godfrey LR, Jungers WL, megafauna in the late Pleistocene. Geology 34:685–688. Goodman SM, Wright HT, Jull AJT. 2004. A chronology Fox-Dobbs K, Bump JK, Peterson RP, Fox DL, Koch PL. 2007. for late prehistoric Madagascar. Journal of Human Evolu- Carnivore-specific stable isotope variables and variation in tion 47:25–63. the foraging ecology of modern and ancient popula- Cernusak LA, Tcherkez G, Keitel C, Cornwell WK, tions: case studies from Isle Royale, Minnesota, and La Santiago LS, Knohl A, Barbour MM, Williams DG, Brea. Canadian Journal of Zoology 85:458–471.

Am. J. Primatol. 36 / Crowley et al.

Frank DA, Evans RD. 1997. Effects of native grazers on Kluge M, Razanoelisoa B, Brulfert J. 2001. Implications of grassland N cycling in Yellowstone National Park. Ecology genotypic diversity and phenotypic plasticity in the eco- 78:2238–2248. phyiological success of CAM plants, examined by studies on Ganzhorn JU, Fietz J, Rakotovao E, Schwab D, Zinner D. the vegetation of Madagascar. Plant Biology 3:214–222. 1999. Lemurs and the regeneration of dry deciduous forest Koch PL. 1998. Isotopic reconstruction of past continental in Madagascar. Conservation Biology 13:794–804. environments. Annual Review of Earth and Planetary Ge´nin F. 2008. Life in unpredictable environments: first Sciences 26:573–613. investigation of the natural history of Microcebus griseorufus. Krueger HW, Sullivan CH. 1984. Models for carbon isotope International Journal of Primatology 29:303–321. fractionation between diet and bone. In: Turnland JF, Godfrey LR, Irwin MT. 2007. The evolution of extinction Johnson PE, editors. Stable isotopes in nutrition, ACS risk: past and present anthropogenic impacts on the symposium series 258. Washington, DC: American Chemical primate communities of Madagascar. Folia Primatologica Society. p 205–222. 78:405–419. Loudon JE, Whitelaw DC, Sponheimer M, Sauther ML, Godfrey LR, Semprebon GM, Schwartz GT, Burney DA, Cuozzo FP. 2008. Lemurs eating isotopes: a stable Jungers WL, Flanagan EK, Cuozzo FP, King SJ. 2005. isotope analysis of ring-tailed lemurs (Lemur catta) New insights into old lemurs: the trophic adaptations of the and their menu at the Beza Mahafaly Special Reserve. Archaeolemuridae. International Journal of Primatology American Journal of Physical Anthropology Supplement 46: 26:825–854. 142. Godfrey LR, Jungers WL, Schwartz GT. 2006. Ecology and Martin RD. 1972. A preliminary field study of the lesser mouse extinction of Madagascar’s subfossil lemurs. In: Gould L, lemur (Microcebus murinus J.F. Miller 1777). Zeitschrift fur Sauther M, editors. Lemurs: ecology and adaptation. Tierpsychologie Supplement 9:43–89. New York: Springer. p 41–64. Milton K, May ML. 1976. Body weight, diet and home range Godfrey LR, Jungers WL, Schwartz GT, Irwin MT. 2008a. area in primates. Nature 25:459–462. Ghosts and orphans: Madagascar’s vanishing ecosystems. Muzuka A. 1999. Isotopic compositions of tropical East African In: Fleagle JG, Gilbert CC, editors. Elwyn Simons: a search flora and their potential as source indicators of organic for origins. New York: Springer. p 361–395. matter in coastal marine sediments. Journal of African Godfrey LR, Crowley BE, Muldoon KM, King SJ, Burney DA. Earth Sciences 28:757–766. 2008b. The Hadropithecus conundrum. American Journal of O’Leary MH. 1988. Carbon isotopes in photosynthesis. Physical Anthropology Supplement 46:105. BioScience 38:328–336. Grant G, More LJ, McKenzie NH, Dorward PM, Buchan WC, Ralisoamalala RC. 1996. Role de Eulemur fulvus rufus Telek L, Pusztai A. 1995. Nutritional and hemagglutination (Audeberg, 1799) et de Propithecus verreauxi verreauxi properties of several tropical seeds. Journal of Agricultural (A. Grandidier 1867) dans la dissemination des graines. Science 124:437–445. In: Ganzhorn JU, Sorg JP, editors. Go¨ttingen, Germany. Harding RSO. 1981. An order of , non-human Ecology and economy of a tropical dry forest in Madagascar. primate diets in the wild. In: Harding RSO, Teleki G, Primate Report 46–1:285–293. editors. Omnivorous primates: gathering and hunting in Rapanarivo SHJV, Lavranos JJ, Leeuwenberg AJM, Ro¨o¨sli W. human evolution. New York: Columbia University Press. 1999. Pachypodium (Apocynaceae): , habitats and p 191–214. cultivation. A.A. Balkema: Rotterdam, Brookfield. Heaton THE. 1987. The 15N/14N ratios of plants in South Richard AF. 1978. Behavioral variation: case study of a Africa and Namibia: relationship to climate and coastal/ Malagasy lemur. In: Candland DK, editor. Cranbury, NJ: saline environments. Oecologia 74:236–246. Associated University Presses. 213p. Hobson KA, Clark RG. 1992. Assessing avian diets using stable Ryan TM, Burney DA, Godfrey LR, Go¨hlich UB, isotopes II: factors influencing diet-tissue fractionation. Jungers WL, Vasey N, Ramilisonina, Walker A, Weber GW. Condor 94:189–197. 2008. A reconstruction of the Vienna of Hadropithecus Hobson KA, Alisauskas RT, Clark RG. 1993. Stable-nitrogen stenognathus. Proceedings of the National Academy isotope enrichment in avian tissues due to fasting and Sciences, United States of America 105:10699–10702. nutritional stress: implications for isotopic analyses of diet. Sauther ML, Sussman RW, Gould L. 1999. The socioecology of Condor 95:388–394. the ringtailed lemur: thirty-five years of study. Evolutionary Humbert H. 1971. Flore de Madagascar et des Comores. 179e Anthropology 8:120–132. Famille Pedaliacees. Paris: Muse´um National d’Histoire Schmidt S, Stewart GR. 2003. d15N values of tropical savanna Naturelle. 5–46p. and monsoon forest species reflect root specializations and Johnson CN. 2009. Ecological consequences of late quaternary soil nitrogen status. Oecologia 134:569–577. extinctions of megafauna. Proceedings of the Royal Society Schoeninger MJ, Iwaniec UT, Glander KE. 1997. Stable B: Biological Sciences 276:2509–2519. isotope ratios indicate diet and habitat use in New World Jolly A. 1966. Lemur behavior: a Madagascar field study. Monkeys. American Journal of Physical Anthropology Chicago: University of Chicago Press. 187p. 103:69–83. Jungers WL, Demes B, Godfrey LR. 2008. How big were the Schulze E-D, Farquha GB, Miller JM, Schulze W, Walker BH, ‘‘giant’’ extinct lemurs of Madagascar? In: Fleagle JG, Williams RJ. 1999. Interpretation of increased foliar d15N Gilbert CC, editors. Elwyn Simons: a search for origins. in woody species along a rainfall gradient in northern New York: Springer. p 343–360. Australia. Australian Journal of Plant Physiology 26: Kluge M, Brulfert J, Ravelomanana D, Lipp J, Ziegler H. 1991. 296–298. Crassulacean acid metabolism in Kalanchoe¨ species col- Sealy JC, van der Merwe NJ, Lee-Thorp JA, Lanham JL. 1987. lected in various climatic zones of Madagascar: a survey by Nitrogen isotopic ecology in southern Africa: implications d13C analysis. Oecologia 88:407–414. for environmental and dietary tracing. Geochimica Cosmo- Kluge M, Brulfert J, Rauh W, Ravelomanana D, Ziegler H. chimica Acta 51:2707–2717. 1995. Ecophyiological studies on the vegetation of Simmen B, Sauther ML, Soma T, Rasamimanana H, Madagascar: a d13C and dD survey for incidence of Sussman RW, Jolly A, Tarnaud L, Hladik A. 2006. Plant Crassulacean Acid Metabolism (CAM) among orchids from species fed on by Lemur catta in gallery forests of the montane forests and succulents from the exerophytic thorn- Southern Domain of Madagascar. In: Jolly A, Sussman RW, bush. Isotopes in Environmental and Health Studies Koyama N, Rasamimanana H, editors. Ringtailed lemur 31:191–210. biology. New York: Springer. p 55–68.

Am. J. Primatol. Isotopic Niches and Lemur Species Loss in Southern Madagascar / 37

Sussman RW. 1974. Ecological distinctions in sympatric Voigt CC, Matt F. 2004. Nitrogen stress causes unpredictable species of Lemur. In: Martin RD, Doyle GA, Walker AC, enrichments of 15N in two -feeding species. editors. Prosimian biology. London: Gerald Duckworth & Journal of Experimental Biology 207:1741–1748. Co. p 75–108. Voigt FA, Bleher B, Fietz J, Ganzhorn JU, Schwab D, Sussman RW. 1999. Primate ecology and social structure, Bo¨hning-Gaese K. 2004. A comparison of morphological volume 1: , lemurs, and tarsiers. Boston: Pearson and chemical fruit traits between two sites with different Press. 284p. frugivore assemblages. Oecologia 141:94–104. Sussman RW, Rakotozafy A. 1994. Plant diversity and Voigt FA, Arafeh R, Farwig N, Griebeler EM, Bo¨hning-Gaese structural analysis of a tropical dry forest in southwestern K. 2009. Linking seed dispersal and genetic structure of Madagascar. Biotropica 26:241–254. trees: a biogeographical approach. Journal of Biogeography Swap RJ, Aranibar JN, Dowty PR, Gilhooly III WP, Macko SA. 36:242–254. 2004. Natural abundance of 13C and 15NinC and C Walker A, Ryan TM, Silcox MT, Simons EL, Spoor F. 2008. 3 4 The semicircular canal system and locomotion: the case of vegetation of southern Africa: patterns and implications. extinct lemuroids and lorisoids. Evolutionary Anthropology Global Change Biology 10:350–358. 17:135–145. Vavra M, Parks CG, Wisdom MJ. 2007. , Winter K. 1979. d13C values of some succulent plants from exotic plant species, and herbivory: the good, the bad, Madagascar. Oecologia 40:103–112. and the . Forest Ecology and Management 246: Youssouf Jacky IA, Rasoazanabary E. 2008. Discovery of 66–72. Macrotarsomys bastardi at Beza Mahafaly Special Reserve, 15 Vitousek PM, Shearer G, Kohl DH. 1989. Foliar N natural southwest Madagascar, with observations on the dynamics abundance in Hawaiian rainforest: patterns and possible of small interactions. Madagascar Conservation & mechanisms. Oecologia 78:383–388. Development 3:31–37.

Am. J. Primatol.