A mix of community-based conservation and protected forests is needed for the survival of the Endangered pygmy hippopotamus Choeropsis liberiensis
A NNIKA H ILLERS,GRAEME M. BUCHANAN,JERRY C. GARTEH,SOLOMON M. TOMMY M OHAMED L. FOFANA and J EREMY A. LINDSELL
Abstract The contribution of protected areas to biodiversity Keywords Agricultural expansion, community-based con- conservation is well attested but many taxa in many regions servation, community forests, large mammals, protected remain dependent on the unprotected wider landscape. To area, pygmy hippopotamus, species distribution model, develop conservation plans for large mammals such as the Upper Guinea Forest ecosystem Endangered pygmy hippopotamus Choeropsis liberiensis of To view supplementary material for this article, please visit West Africa’s Upper Guinea Forests it is critical to under- http://dx.doi.org/./SX stand the importance of unprotected land. Despite being a conservation priority, little is known about the habitat asso- ciations of this species, or its distribution across its range. Through a combination of field surveys, species distribution Introduction models and community questionnaires we investigated the use of unprotected areas by the pygmy hippopotamus in the rotected areas are the foundation for global efforts to Sierra Leone–Liberia border region. We found signs of Pconserve biodiversity and natural ecosystem processes the species in of -km cells surveyed. Our analysis (Rodrigues et al., ; Laurance et al., ). However, ana- suggested that the species is reasonably widespread in this lysis of threatened birds, mammals, amphibians and reptiles region and is associated with major rivers. It occurred at both continental and global scales suggests that protected close to, but rarely within, large areas of intact forest, and areas offer inadequate coverage of the ranges of these groups .% of pygmy hippopotamus signs were recorded outside (De Klerk et al., ; Rodrigues et al., ; Beresford et al., protected areas. The expansion of the protected area net- a,b; Butchart et al., ). Multiple-use landscapes and work in this area is unrealistic in Sierra Leone and to unprotected areas are important for species survival some extent in Liberia, mainly because of anthropogenic (Gardner et al., ; Perfecto & Vandermeer, ), espe- pressure and the overlap of proposed protected areas with cially where the existing network of protected areas does not ’ mining and logging concessions. Thus pygmy hippopot- entirely cover species ranges and suitable habitats; e.g. for amus conservation activities in the region need to include some large mammalian carnivores (Forrest et al., ; programmes on community lands while maintaining a ro- Swanepoel et al., ) and migratory herbivores (Western bust network of protected forests. Community-based con- et al., ). In some cases the expansion of the protected servation of the pygmy hippopotamus may prove valuable area network is recommended for more effective protection for other threatened and endemic species that are not con- (Forrest et al., ; Tweh et al., ) but this may not always fined to protected areas in this region. be feasible (Butchart et al., ), especially in regions where conservation efforts compete with other land uses, such as logging, mineral extraction and agriculture. In many tropic- al regions, such land use conflicts represent a threat not only to remaining natural habitats outside protected areas but ANNIKA HILLERS* (Corresponding author) and GRAEME M. BUCHANAN RSPB even to the protected areas themselves (Forrest et al., ; Centre for Conservation Science, The Royal Society for the Protection of Laurance et al., ; Tweh et al., ). A rapidly increasing Birds, The Lodge, Sandy, Bedfordshire, SG19 2DL, UK E-mail [email protected] human population is predicted to result in a further major expansion of tropical agriculture and associated encroach- JERRY C. GARTEH Society for the Conservation of Nature of Liberia, Congotown, Monrovia, Liberia ment of protected areas, particularly in Sub-Saharan SOLOMON M. TOMMY and MOHAMED L. FOFANA, Gola Rainforest National Park, Africa and South America (Laurance et al., ). The per- Kenema, Sierra Leone sistence of large mammals in such regions will therefore de-
JEREMY A. LINDSELL A Rocha International, London, UK pend to a great extent on their ability to survive in such *Also at: Gola Rainforest National Park, Kenema, Sierra Leone mixed landscapes. Received July . Revision requested September . Within the forest zone of West Africa, protected areas Accepted February . cover c. % of the remaining Upper Guinean Forest
Oryx, Page 1 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 2 A. Hillers et al.
ecosystem (CEPF, ). In Liberia and Sierra Leone only in south-eastern Sierra Leone, and presumably also the and % of the total land area, respectively, is officially pro- Gola National Forest in western Liberia, may constitute an tected (Junker et al., ; UNEP, ). These figures sug- important core area for the species (Hillers, ). The dis- gest that the protected area coverage in the region is covery of what could be a substantial population highlights inadequate, and falls below the Convention on Biological how poorly known the species’ distribution is. Diversity Aichi Target of at least % (CBD, ). We used records of pygmy hippopotamus occurrence Meanwhile, forests in this region have been severely de- collected during field surveys and using community ques- pleted and fragmented in recent decades (Mittermeier tionnaires to assess the species’ distribution in the Gola re- et al., ). Where forests are cleared they are replaced gion. We used field data to develop species distribution with low-intensity rotational agriculture. models that predicted the potential distribution of the The remaining forests harbour a number of threatened pygmy hippopotamus, and described the species’ basic habi- large mammals, some of which are known to range beyond tat associations. We were thus able to identify the extent to the bounds of protected areas (Merz, ; Junker et al., which the pygmy hippopotamus might occur in protected , ). Of these, the pygmy hippopotamus Choeropsis and unprotected areas, and where suitable habitat was lo- liberiensis, categorized as Endangered on the IUCN Red List cated both inside and outside the protected area network. (Ransom et al., ), is the largest species that is entirely en- Besides advancing our understanding of the role of unpro- demic to the Upper Guinean Forest and is therefore tected areas in general biodiversity conservation in this re- amongst the most vulnerable to habitat loss and fragmenta- gion, these data are critical if targeted and effective tion in the region. The pygmy hippopotamus is restricted to conservation interventions are to be implemented for this four countries (Sierra Leone, Liberia, Guinea and Côte species in particular. d’Ivoire; Lewison & Oliver, ), all of which are among the least developed countries in the world (UNDP, ). In recent decades, forest loss and fragmentation have Study area contributed to a reduction in the occupied range of the We conducted field work in the Gola Rainforest National species (Lewison & Oliver, ; Mallon et al., ; Park and its surroundings (c. ,, ha) in south-eastern Hodgkinson et al., ). The pygmy hippopotamus is also Sierra Leone and in the Gola National Forest (initially – threatened by poaching and human wildlife conflicts (e.g. c. , ha) in western Liberia, the whole area hereafter Greengrass, ; Hillers & Muana, ; Koroma, ; referred to as Gola (Fig. ). After the completion of field Conway, ; Conway et al., ), although there are no work, the area of the Gola National Forest was reduced by detailed data concerning the extent of these threats. Little , ha and part of our survey sites that were initially in- is known about the ecology and behaviour of the pygmy side the protected area now fall within community land. hippopotamus, and there are no reliable population esti- Natural vegetation in Gola is lowland moist evergreen – mates. The most recent estimate is , , individuals forest, and annual rainfall is c. , mm, falling in one (Mallon et al., ). The prevailing view is that protected season. Dominating families of the woody vegetation are areas harbour the majority of the population (Mallon Leguminosae-Caesalpinoideae, Euphorbiaceae, Leguminosae- et al., ; IUCN, ) but the importance of habitats out- Mimosoideae and Sterculiaceae (Klop et al., ). The study side protected areas remains unclear. Although most recent area is drained by three major rivers: the Moa River to the west; records have come from within protected areas (Collen the Moro River, marking the boundary between Sierra Leone et al., ; Mallon et al., ) this may represent sampling and Liberia in the eastern part of Gola; and the Mano River, bias, as such sites are often targeted for biodiversity surveys. flowing through the Liberian Gola National Forest and along Furthermore, riverine habitats can be poorly represented in the international border in the western part of Gola. protected areas (Nel et al., ) but are presumably critical for the pygmy hippopotamus, and in fact protected area sta- tus can mean little in a region where many are poorly funded Methods and ineffective (Brugière, ; Henschel et al., ). We undertook surveys for the pygmy hippopotamus Data on the distribution of the pygmy hippopotamus were – both within and outside protected areas to clarify the im- collected by field surveys during and community portance of unprotected areas for this species. We con- questionnaires administered in . ducted this work in the western part of the species’ range within Sierra Leone and western Liberia (Lewison, ). Field surveys Distribution information from this area is patchy and out- dated (e.g. Grubb et al., ) but evidence has emerged that Distribution data were collected by walking along streams and besides a well-known site at Tiwai Island (Conway et al., riversides within and around the Gola Rainforest National ) the area around the Gola Rainforest National Park Park during July –June and May –April ,
Oryx, Page 2 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X
http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 The pygmy hippopotamus 3
FIG. 1 (a) Location of the Gola Forests in Sierra Leone and Liberia in West Africa, and the predicted distribution of the pygmy hippopotamus Choeropsis liberiensis in both countries. (b) The study area around the Gola Rainforest National Park and the Gola National Forest. Areas predicted by the model at the . threshold and above the . threshold are indicated by light and dark shading, respectively. searching carefully for signs of the species (footprints, trails, We recorded direct observations (sightings and auditory) dung, feeding sites, wallows, direct sightings), especially and signs (dung, prints, trails, wallows, nests, feeding sites) where the substrate might preserve footprints or conditions of all larger mammals (. c. kg) along straight transects were suitable for a wallow or foraging. We searched sites in Gola Rainforest National Park, seven transects in com- along the length of the Moro and Mano rivers, tributaries munity areas (of which five were initially inside the pro- draining the National Park both within and outside the pro- tected Gola National Forest before the protected area was tected area, and parts of the Moa River. These surveys com- reduced) and transects in Gola National Forest during plemented work conducted on and around Tiwai Island April –July . Transects were usually km long ( (Conway, ; Conway et al., ). transects were shorter, –, m, because of their loca- In and field surveys were partly informed by tion along rivers or borders of the National Park) and . community questionnaires (below) as the field teams also km apart, arranged on a systematic grid (Hillers, ). conducted surveys to confirm the presence of pygmy hippo- Opportunistic observations were recorded throughout potamus at places where respondents had stated they had the study area (Fig. ) when signs of pygmy hippopotamus recently seen the species. were observed outside the sampling framework of particular Where pygmy hippopotamus signs were found the pre- surveys (Hillers, ). dominant habitat type was recorded as () forest (healthy primary or secondary forest after selective logging), () farm bush (regrown vegetation after agricultural rotational Modelling the potential distribution of the pygmy activities), ( ) plantations of non-natural vegetation, such as hippopotamus using field survey data oil palm and cocoa, or () rivers, streams and swamps, with- out clear definition of the surrounding habitat type (Hillers All of the presence data recorded during the field surveys were & Muana, ; Hillers, ). used to train maximum entropy models in Maxent .. Focal survey data were supplemented with ad hoc obser- (Phillips et al., ; Phillips & Dudik, ). Community vations from general biodiversity survey data (Fig. ) from questionnaire data were not used in the modelling. –. A terrestrial mammal survey using camera Environmental conditions (Table ) in occupied -km cells traps covered km of Gola Rainforest National Park, (the resolution of the analysis) were compared to those at Gola National Forest and the intervening community , randomly generated pseudo absence points within lands during – (Hillers, ). Camera traps the model building area (Supplementary Fig. S), before ex- (PC Hyperfire, Reconyx Inc., Holmen, USA) were trapolation to the whole of Sierra Leone and Liberia. The placed for c. days each at the centre of × km grid method of data collection may have introduced some biases cells (located using a global positioning system), using into the data (e.g. footprints more likely to be detected in camera positions unbiased by animal signs (Wearn et al., soft substrates) but Maxent appears to be robust to data col- ). lected in such an ad hoc way (Franklin, ).
Oryx, Page 3 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 4 A. Hillers et al.
FIG. 2 Locations of transects, camera traps, and streams surveyed for the pygmy hippopotamus in and around the Gola Rainforest National Park in Sierra Leone and the Gola National Forest in Liberia.
Only variables that explained at least % of the model gain questioned. They were asked to answer the questions only were retained in the final model using backwards deletion in relation to the community where the questionnaire was (Table ). The potential of model overfitting was reduced being administered. Interviews were conducted in the by altering the beta multiplier. A value of produced smooth local language by National Park staff members. It was stated response curves. The final model was run with cross vali- clearly that the purpose of the questionnaires was not pros- dations (i.e. the model was re-run times with a different ecution or punishment and that no personal details of re- % of records removed each time). The mean logistic output spondents were recorded. Questionnaires were designed from all models was used to produce a binary map of po- and administered following the guidelines of the tential distribution, based on the mean training data equal Association of Social Anthropologists of the UK & sensitivity and specificity threshold. The number of pygmy Commonwealth (). hippopotamus in the Gola region was estimated by extrapo- Respondents were asked about encounters with pygmy lation of the extent of potentially suitable habitat and esti- hippopotamus (recent and historical) in their locality, behav- mates of home range size (Roth et al., ): females – iour, activity patterns, habitat preferences, diet, and changes ha (midpoint ha used), males . ha, although home in the abundance of pygmy hippopotamus. They were also ranges of individuals can overlap. asked about human–wildlife conflicts, hunting of pygmy hippopotamus, their general attitude towards pygmy hippo- Community questionnaires potamus, and any traditional uses and beliefs pertaining to the species. Initial questions about each of these subjects Semi-structured questionnaires were administered in could be answered either yes or no, with yes responses leading communities around Gola Rainforest National Park during to additional questions. Answers regarding traditional uses July–November . Communities within km of the and beliefs, general attitudes towards the pygmy hippopot- National Park boundary and/or km of major streams amus, and the species’ diet are not reported here. were selected, to increase the probability of encountering Responses about encounters with the pygmy hippopot- people with knowledge of the pygmy hippopotamus. amus were mapped and reports of the species’ presence Only respondents who could identify a pygmy hippopot- were investigated by field visits. Community localities amus from a set of pictures of large mammals were were categorized as those where the pygmy hippopotamus
Oryx, Page 4 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X
http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 The pygmy hippopotamus 5
FIG. 3 Distribution of communities in the Gola region (Fig. ) where questionnaires about the pygmy hippopotamus were administered, and locations where signs of pygmy hippopotamus were recorded during field surveys. Communities are categorized as () those where the pygmy hippopotamus was observed both in the past and at present and fresh signs were recorded, () those where the pygmy hippopotamus was observed both in the past and at present but no fresh signs were recorded, () those where the pygmy hippopotamus was observed in the past but is no longer present, or () those where the pygmy hippopotamus was not observed either in the past or at present. Drainage is shown for a portion of the region for which data were available.
TABLE 1 Explanatory variables used for entropy models.
Class Variable Product Source Date acquired Land cover Maximum monthly Normalized Difference SPOT (2010) VITO Earth Observation (2010) 13 Sep. 2013 Vegetation Index Euclidean distance to forest GLC2000 Mayaux et al. (2004) Euclidean distance to water DIVA DIVA-GIS (2013) 18 Sep. 2013 Topography Slope, elevation HYDRO1 K USGS (2013) 17 Sep. 2013 Climate Mean annual + monthly variation in Bioclim Hijmans et al. (2005) 13 Sep. 2013 temperature & precipitation
() had never been reported, () was reported in the past, () .% of signs were footprints or trails, .% were dung, .% was reported to be present but was not verified during field were feeding sites, .% were direct sightings or camera-trap surveys, or () was reported and verified as present. photographs, and .% were wallows. The remaining signs (.%) were not specified. In some cases multiple signs were recorded at the same location (.%). Results Habitat descriptions were available for record loca- tions, with the majority of records found in farm bush Observed and modelled distribution of the pygmy areas ( ; . %), followed by forest ( ; . %), and rivers, hippopotamus streams and swamps ( ; . %). Only one record of pygmy hippopotamus presence was from a plantation (cocoa). Field surveys, including camera trapping, covered -km Footprints (.%), feeding sites (.%) and dung grid cells. Pygmy hippopotamus and their signs were re- (.%) were mainly recorded in farm bush areas (Fig. ), corded at locations in of these cells (Fig. ). Eight of whereas direct sightings and camera-trap photographs these locations were verified following information received were mainly recorded in the forest (.%). Pygmy hippo- through the community questionnaires. At the locations, potamus trails were mostly recorded in farm bush areas
Oryx, Page 5 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 6 A. Hillers et al.
TABLE 2 Variables included in the model of pygmy hippopotamus percentage contribution to the model (Table ), perhaps in- Choeropsis liberiensis distribution, with percentage contribution dicating that both variables were of similar importance in and permutation importance. determining the distribution (although distance to forest Permutation might have been marginally more important given that it Variable % contribution1 importance2 makes a larger contribution under permutation testing). Temperature variation 23.2 15.3 The species distribution model across the cross- Annual rainfall 20.9 6.6 validation runs had a mean AUC (area under the curve) Distance to forest 15.6 29.9 of . ± SD ., indicating the model was a reasonable Distance to river 12 8.3 fit to the data. At the equal sensitivity and specificity thresh- Mean temperature 11 10.1 old (.) there was , km of potentially occupied habi- Feb. normalized difference 7.7 2.7 tat across Sierra Leone and Liberia. The model correctly vegetation index (NDVI) identified of the locations with sightings but it incor- Altitude 4.6 5.3 Rain variation 2.2 21.5 rectly predicted of the apparently unoccupied Apr. NDVI 1.7 0.2 squares (locations surveyed but at which no signs of Dec. NDVI 1.1 0.2 pygmy hippopotamus were recorded) to be occupied. This
can arise where suitable habitat is unoccupied or where In each iteration of the training algorithm the increase in regularized gain was added to the contribution of the corresponding variable, or subtracted pygmy hippopotamus were undetected in occupied sites, from it if the change to the absolute value of lambda was negative. or it may indicate that the model had high sensitivity but For each variable in turn, the values of that variable on training presence low specificity (i.e. it overpredicted potential distribution). and background data were randomly permuted. The model was reevaluated A comparison of likelihood of occupancy at these apparent on the permuted data, and the resulting drop in training AUC is shown, normalized to percentages. absences with the values at confirmed presences (rather than pseudo absences from the model building process) produced an ROC (receiver operating characteristic) to ., with a revised equal sensitivity and specificity threshold of .. At this threshold % of both presences and ab- sences were correctly predicted, and , km across Sierra Leone and Liberia were identified as potentially occu- pied by the pygmy hippopotamus. This indicates there may be habitat for c. , female and fewer than c. male pygmy hippopotamus across the two countries. Many of these patches are small and isolated (Fig. ), perhaps render- ing them unsuitable. By far the largest area of potentially suitable habitat is around the Gola region (Fig. ), where much of the suitable habitat falls outside Gola Rainforest National Park and Gola National Forest (Fig. b).
FIG. 4 Percentage of various types of pygmy hippopotamus signs recorded in various habitat types. Community perceptions of pygmy hippopotamus distribution (.%) and along rivers, streams and swamps (.%). However, the habitats surrounding rivers, streams and In the majority of communities () – respondents were swamps were not specified and could be either forest or interviewed. In the remaining communities only one re- farm bush areas. spondent was interviewed, based on the ability to identify a The majority of record locations (; .%) were out- pygmy hippopotamus. Of respondents in communi- side the boundary of the protected areas. The locations were ties only five communities reported that pygmy hippopot- clustered along the larger streams and rivers, such as the amus were not currently present or had never been Mano and Moro Rivers bordering Sierra Leone and present. Eighty-one communities reported that pygmy Liberia, and the Moa River. hippopotamus were present in the past but were no longer Modelled pygmy hippopotamus distribution, based on present, and reported that pygmy hippopotamus were the occupied -km cells, was most strongly associated present at the time of the study but signs of presence were with climatic variables and distance to rivers and closed for- found in only eight of these locations during follow-up sur- est, with response curves indicating that areas closer to riv- veys (Fig. ). Communities where the pygmy hippopotamus ers and forest were more likely to be occupied (Fig. ). was considered (or proven) to be extant were mainly located Distance to rivers and to closed forests made a similar along the Mano, Moro and Moa Rivers and nearer to the
Oryx, Page 6 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X
http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 The pygmy hippopotamus 7 border with Liberia (Fig. ). The majority of the reported ex- extrapolating the results to the rest of the pygmy hippopot- tinctions were along the more heavily settled northern amus’ range. Areas mapped as potentially suitable (such as boundary of Gola Rainforest National Park. However, around the coast of Sierra Leone) may be covered by habitat field surveys indicated that the species persists in part of that is broadly suitable for the species, yet it may be absent this area (Fig. ). Over % of respondents (; .%) be- for reasons such as hunting pressure or fragmentation. Of lieved that pygmy hippopotamus were less common than in perhaps greater concern is the failure of the model to iden- the past. Respondents attributed the decline to logging (es- tify areas in eastern Liberia (e.g. Sapo National Park, Grebo pecially the noise of power saw operation; respondents), National Forest) where the pygmy hippopotamus is known farming (), hunting (), expansion and creation of settle- to occur. The reasons for this are worth investigating but ments (), general habitat destruction and deforestation (), may be linked to differences in land cover (e.g. larger con- the sound of guns and shooting during the war (), mining nected forest areas in eastern Liberia). (), as well as natural deaths () and dry streams (). Our population estimate for the pygmy hippopotamus in Sierra Leone and Liberia alone was close to the estimated global population, which includes populations in Guinea ’ Discussion and Côte d Ivoire. There are a number of possible explana- tions for this apparent discrepancy between local and global Field surveys, species distribution models, and community population estimates. Firstly, our model may have overesti- questionnaires all indicate that in the Gola region the pygmy mated the potential extent of suitable habitat, even after as- hippopotamus is generally found in association with sessing the model based on known absences. Secondly, our streams or swampy areas, mostly outside protected areas, model only examined habitat and did not include data on and in close proximity to forests (including protected for- potential exploitation, such as hunting for bushmeat. ests). Closed forest, although of equal or perhaps greater im- Thus, although there might be suitable habitat to support portance than proximity to streams and rivers according to a high number of individuals, the actual numbers may be the models, was not the most frequented habitat according lower as a result of hunting. Under these circumstances, to field surveys. However, with the modelling approach used control of hunting could result in an increase in the popu- we were unable to definitively identify which was more im- lation. Finally, it is possible that the models were adequate portant. The outputs of the distribution model are sup- but that the density estimates from previous studies (Roth ported by the field observations and suggest that the most et al., ) were unrealistically high and that the previous suitable habitat is in the unprotected, community-owned population estimate is incorrect. With one exception the landscape. In particular, the area along the Moro and home range estimations of Roth et al. () were based Mano Rivers between Sierra Leone and Liberia was pre- on translocated individuals. It may be possible that in its dicted to contain extensive habitat suitable for the pygmy natural habitat the pygmy hippopotamus has larger home hippopotamus. Some of the apparent selection for streams ranges, which would affect our model and lead to lower es- and swampy areas, especially in the analysis of field surveys, timates of population sizes. Field surveys are needed in areas might have been because of the targeting of these areas for predicted to be occupied as well as those predicted to be un- surveys. These biases might have continued to have an influ- occupied, to test the model thoroughly, and improved dens- ence on the species distribution models but comparison of ity estimates in various habitat types are needed. Future field occupied areas with such a large number of pseudo absences surveys should be structured, using standardized methods. should reduce the impact of this potential bias. Larger The data used in this analysis were collected by multiple streams and lowland areas are more prevalent in Gola methods, in different years and by multiple observers. National Forest than Gola Rainforest National Park, Consequently there may be unquantified errors and biases which may explain the greater number of detections in the in our data. However, we modelled the data using an approach former, despite the higher hunting pressure there (Hillers, that has been shown to work well with ad hoc survey data ; Lahai, ). Elsewhere in its range the pygmy hippo- (Franklin, ),andatabroadspatialresolution( km cells). potamus is reported to use even small streams under forest The Conservation Strategy for the Pygmy Hippopotamus cover (Roth et al., ). (Mallon et al., ) indicates that most recent records of the The pygmy hippopotamus’ preference for mainly unpro- species are from protected forest areas, and puts consider- tected farm bush areas may also be linked to the presence of able emphasis on the importance of protected areas in man- the herbaceaous plant Triumfetta cordifolia (family aging this species. Expansion of the protected area network Tiliaceae; kPuhun in the local language; Hillers & Muana, to cover a greater area of forest should be encouraged, espe- ). Local people identified this plant as an important cially in light of the reported low coverage of Sierra Leone food plant for the pygmy hippopotamus. and Liberia by protected areas (Junker et al., ; UNEP, Our species distribution model is based on data from a ). Whereas forests in Sierra Leone, Guinea and Côte small area, and therefore caution should be applied in d’Ivoire are highly fragmented and under significant
Oryx, Page 7 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 8 A. Hillers et al.
anthropogenic pressure (Mallon et al., ), Liberia still has hippopotamus, are still being hunted in the Gola region large blocks of forest in the north-west and south-east (Koroma, ; Hillers, ) and in other areas in Liberia (Christie et al., ). The Liberian government committed (Greengrass, ). to the protection of % of remaining forests in and Apart from some areas in north-western and south- again in (Junker et al., ). However, proposed eastern Liberia, remaining forests in all pygmy hippopot- protected areas largely overlap with logging and mining amus range countries are highly fragmented (Mallon concessions, which jeopardizes further declarations of et al., ) and large-scale exploitation of timber and miner- protected areas. als, as well as the increasing need for agricultural land, are It may be unfeasible to expand the global protected area further threatening remaining natural habitats and their network to cover all sites of importance and capture species’ biodiversity (Laurance et al., ; Tweh et al., ). ranges adequately (Butchart et al., ). However, our ana- Therefore, it is likely that pressures on the pygmy hippopot- lysis showed that although the pygmy hippopotamus may amus will remain high. Consequently, in addition to the po- rely on forest to some unknown extent, protecting forests tential expansion of the protected area network in Liberia alone will not protect all of the pygmy hippopotamus’ habi- (Junker et al., ) community-based conservation activ- tats. Thus, in addition to maintaining a robust network of ities and community management of unprotected sites, in- protected forests, conservationists should also consider cluding effective land-use planning, are likely to become an community land close to protected areas, especially wet important tool for conservation in West Africa. Approaches areas along rivers. The complementary conservation value that empower communities as well as improving biodiver- of unprotected areas has previously been shown in various sity should be encouraged, especially with a view to meeting locations and for various species (e.g. Gardner et al., ; the Sustainable Development Goals (UN, ). Perfecto & Vandermeer, ; Forrest et al., ), and this and other recent surveys have indicated the importance of habitats outside protected areas for threatened species in Acknowledgements Sierra Leone and Liberia. More than half of the known breeding colonies of the Vulnerable white-necked pi- This work was funded by Basel Zoo, Switzerland, and the cathartes Picathartes gymnocephalus in the Gola region European Union. We thank partners of the Gola Rainforest — are located on community land around Gola Rainforest National Park and the project Across the River A National Park (Monticelli et al., ; Hillers, ), and in Transboundary Peace Park for Sierra Leone and Liberia:the Liberia a nationwide survey of large mammals revealed that Forestry Division of the Ministry of Agriculture, Forestry and the majority (%) of western chimpanzees Pan troglodytes Food Security, the Conservation Society of Sierra Leone, the verus and some of the most species-diverse mammal com- Liberian Forestry Development Authority, the Society for the munities occur outside protected areas (Tweh et al., ). Conservation of Nature of Liberia, BirdLife International, There are successful examples of community conserva- Vogelbescherming Nederland and the Royal Society for the tion initiatives that have been applied in various forms Protection of Birds. Andrew Muana, Sheku Massaquoi, and could also work in Sierra Leone and Liberia, including Mohamed Kallon, Muana M. Konneh, Umaru Bockarie, community sensitization and outreach programmes Lahai Kanneh, Trokon S. Grimes, Alex T. Johnson, (Trewhella et al., , Steinmetz et al., ), community- Mawehn B. Koenig, Aaron Dayeker, Daniel K. Harris, based wildlife monitoring (Danielsen et al., ) and Mohamed Koroma, Mohamed F. Kallon and Prince Soriba community-based natural resource management (e.g. helped with field data collection, with contributions from through community conservancies and wildlife sanctuaries; additional Gola Rainforest National Park research techni- Brown & Bird, ; Garnier et al., ). In Ghana, at the cians, local guides, and Jessica Ganas. This paper is dedicated eastern edge of the Upper Guinea Forest ecosystem, the es- to our late colleagues Andrew Muana and Sheku Massaquoi. tablishment of a Community Resource Management Area process has proven to be a successful and promising tool for managing forest resources (Asare et al., ). References Such community-level approaches may also be more suc- ASARE, R.A., KYEI,A.&MASON, J.J. () The community resource cessful in addressing threats unrelated to habitat, such as management area mechanism: a strategy to manage African forest hunting. Within the pygmy hippopotamus’ range several resources for REDD+. Philosophical Transactions of the Royal environmental education and outreach programmes focus- Society B, , . ing on the species’ conservation have been conducted ASSOCIATION OF SOCIAL ANTHROPOLOGISTS OF THE UK & (Steck, ). However, no follow-up research has investi- COMMONWEALTH ( ) Ethical Guidelines for Good Research Practice. Http://www.theasa.org/ethics/guidelines.shtml [accessed gated the impact and success of these programmes so far, March ]. and there is evidence that despite past and ongoing outreach BERESFORD,A.E.,BUCHANAN, G.M., D ONALD, P.F., B UTCHART, S.H.M., programmes threatened species, including the pygmy FISHPOOL,L.D.C.&RONDININI,C.(a) Poor overlap between the
Oryx, Page 8 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X
http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 The pygmy hippopotamus 9
distribution of protected areas and globally threatened birds in HENSCHEL, P., COAD, L., BURTON, C., CHATAIGNER, B., DUNN, A., Africa. Animal Conservation, , –. MACDONALD, D. et al. () The lion in West Africa is Critically BERESFORD,A.E.,BUCHANAN, G.M., D ONALD, P.F., B UTCHART, S.H.M., Endangered. PLoS ONE, (), e. FISHPOOL,L.D.C.&RONDININI,C.(b) Minding the HIJMANS, R.J., CAMERON, S.E., PARRA, J.L., JONES, P.G. & JARVIS,A. protection gap: estimates of species’ range sizes and holes in the () Very high resolution interpolated climate surfaces for global protected area network. Animal Conservation, , –. land areas. International Journal of Climatology, , –. BROWN,J.&BIRD,N.() Sustainable Natural Resource HILLERS,A.() Final Report on the Activities of the Research Unit, Management in Namibia: Successful Community-Based Wildlife Across the River—A Transboundary Peace Park for Sierra Leone and Conservation. ODI Publications, London, UK. Liberia, to . Unpublished report for BirdLife International/ BRUGIÈRE,D.() Identifying priority areas for the conservation of RSPB. antelopes in the Republic of Guinea, West Africa, using the HILLERS,A.&MUANA,A.() Pygmy Hippopotamus Conservation complementarity approach. Oryx, , –. Project Within the “Across the River—A Transboundary Peace Park BUTCHART, S.H.M., CLARKE, M., SMITH, R.J., SYKES, R.E., for Sierra Leone and Liberia” Project (ARTP). Unpublished report SCHARLEMANN, J.P.W., HARFOOT, M. et al. () Shortfalls and for Zoo Basel, Basel, Switzerland. solutions for meeting national and global conservation area targets. HODGKINSON, C., MALLON, D., VOGT,T.&RANSOM,C.() First Conservation Letters, , –. national conservation strategy for the pygmy hippopotamus in CBD (CONVENTION ON BIOLOGICAL DIVERSITY)() COP Liberia. Oryx, , –. Decision X/. Strategic plan for biodiversity –. Http://www. IUCN () The IUCN Red List of Threatened Species v. .. cbd.int/decision/cop/?id= [accessed October ]. Http://www.iucnredlist.org [accessed October ]. CEPF (CRITICAL ECOSYSTEM PARTNERSHIP FUND)() Upper JUNKER, J., BLAKE, S., BOESCH, C., CAMPBELL, G., DU TOIT, L., Guinean Forest Ecosystem of Guinean Forests of West Africa DUVALL, C. et al. () Recent decline in suitable environmental Biodiversity Hotspot. CEPF, Washington, D.C., USA. conditions for African great apes. Diversity and Distributions, , CHRISTIE, T., STEININGER, M.K., JUHN,D.&PEAL,A.() –. Fragmentation and clearance of Liberia’s forests during –. JUNKER, J., BOESCH, C., FREEMAN, T., MUNDRY, R., STEPHENS,C.& Oryx, , –. KÜHL, H.S. () Integrating wildlife conservation with conflicting COLLEN, B., HOWARD, R., KONIE, J., DANIEL,O.&RIST,J.() Field economic land-use goals in a West African biodiversity hotspot. surveys for the Endangered pygmy hippopotamus Choeropsis Basic and Applied Ecology, , –. liberiensis in Sapo National Park, Liberia. Oryx, , –. KLOP, E., LINDSELL, J.A. & SIAKA,A.() Biodiversity of Gola CONWAY, A.L. () Conservation of the pygmy hippopotamus Forest, Sierra Leone. Unpublished report. RSPB and CSSL. (Choeropsis liberiensis) in Sierra Leone, West Africa. PhD thesis. KOROMA,P.Y.() Analysis of the distribution of large mammals University of Georgia, Athens, USA. around the Gola Rainforest National Park (Malema Chiefdom) CONWAY, A.L., HERNANDEZ, S.M., CARROLL, G.P., GREEN, G.T. & based on a camera trap study, with a focus on key bushmeat species. LARSON,L.() Local awareness of and attitudes towards the BSc thesis. Eastern Polytechnic, Kenema, Sierra Leone. pygmy hippopotamus Choeropsis liberiensis in the Moa River Island LAHAI,K.() The distribution of key species in the Gola Forests – a Complex, Sierra Leone. Oryx, , –. comparison between the Gola Rainforest National Park in Sierra DANIELSEN, F., JENSEN, P.M., BURGESS, N.D., ALTAMARINO, R., Leone and the Gola National Forest in Liberia. A case study in Gola ALVIOLA, P.A., ANDRIANANDRASANA, H. et al. ()A Central (Nomo, Gaura, Malema and Tunkia Chiefdoms). BSc thesis. multicountry assessment of tropical resource monitoring by local Eastern Polytechnic, Kenema, Sierra Leone. communities. BioScience, , –. LAURANCE, W.F., SAYER,J.&CASSMAN, K.G. () Agricultural DE KLERK, H.M., FJELDSÅ, J., BLYTH,S.&BURGESS, N.D. () Gaps expansion and its impacts on tropical nature. Trends in Ecology & in the protected area network for threatened Afrotropical birds. Evolution, , –. Biological Conservation, , –. LAURANCE, W.F., USECHE, C., RENDEIRO, J., KALKA, M., BRADSHAW, DIVA-GIS () Http://www.diva-gis.org/gdata [accessed C.J.A., SLOAN, S.P. et al. () Averting biodiversity collapse in September ]. tropical forest protected areas. Nature, , –. FORREST, J.L., BOMHARD, B., BUDIMAN, A., COAD, L., COX, N., LEWISON, R.L. () Family Hippopotamidae (hippopotamuses). In DINERSTEIN, E. et al. () Single-species conservation in a Handbook of the Mammals of the World, Volume (eds D.E. Wilson multiple-use landscape: current protection of the tiger range. & R.A. Mittermeier), pp. –. Lynx Edicions, Barcelona, Spain. Animal Conservation, , –. LEWISON,R.&OLIVER,W.(IUCNSSCHIPPO SPECIALIST FRANKLIN,J.() Mapping Species Distributions. Cambridge SUBGROUP)() Choeropsis liberiensis.InThe IUCN Red List of University Press, Cambridge, UK. Threatened Species v. .. Http://www.iucnredlist.org [accessed GARDNER, T.A., CARO, T., FITZHERBERT, E.B., BANDA,T.&LALBHAI, October ]. P. () Conservation value of multiple-use areas in East Africa. MALLON, D., WIGHTMAN, C., DE ORNELLAS,P.&RANSOM, C. (eds) Conservation Biology, , –. () Conservation Strategy for the Pygmy Hippopotamus. IUCN GARNIER, J., HILL, N., GUISSAMULO, A., SILVA, I., WITT,M.& Species Survival Commission, Gland, Switzerland & Cambridge, GODLEY,B.() Status and community-based conservation of UK. marine turtles in the northern Querimbas Islands (Mozambique). MAYAUX, P., BARTHOLOMÉ, E., FRITZ,S.&BELWARD,A.()A Oryx, , –. new land-cover map of Africa for the year . Journal of GREENGRASS,E.() Exploring the Dynamics of Bushmeat Hunting Biogeography, , –. and Trade in Sapo National Park. Unpublished report. Fauna & MERZ,G.() The status of the forest elephant Loxodonta africana Flora International, Monrovia, Liberia. cyclotis, Matschie, , in the Gola Forest reserves, Sierra Leone. GRUBB, P., JONES, T.S., DAVIES, A.G., EDBERG, E., STARIN, E.D. & Biological Conservation, , –. HILL, J.E. () Mammals of Ghana, Sierra Leone and the Gambia. MITTERMEIER, R.A., ROBLES GIL, P., HOFFMAN, M., PILGRIM, J., The Tendrine Press, Zennor, UK. BROOKS, T., GOETTSCH MITTERMEIER, C. et al. () Hotspots
Oryx, Page 9 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188 10 A. Hillers et al.
Revisited: Earth’s Biologically Richest and Most Endangered TWEH, C.G., LORMIE, M.M., KOUAKOU, C.Y., HILLERS, A., KÜHL, H.S. Terrestrial Ecoregions. University of Chicago Press, Chicago, USA. &JUNKER,J.() Conservation status of chimpanzees Pan MONTICELLI, D., SIAKA, A., BUCHANAN, G.M., WOTTON, S., troglodytes verus and other large mammals in Liberia: a nationwide MORRIS, T., WARDILL, J.C. & LINDSELL, J.A. () Long term survey. Oryx, , –. stability of white-necked picarthartes population in south-east UN (UNITED NATIONS)() UN Transforming Our World: The Sierra Leone. Bird Conservation International, , –. Agenda for Sustainable Development. Resolution adopted by NEL, J.L., ROUX, D.J., MAREE, G., KLEYNHANS, C.J., MOOLMAN, J., the General Assembly on September . Https:// REYERS, B. et al. () Rivers in peril inside and outside protected sustainabledevelopment.un.org/post/transformingourworld areas: a systematic approach to conservation assessment of river [accessed April ]. ecosystems. Diversity and Distributions, , –. UNDP (UNITED NATIONS DEVELOPMENT PROGRAMME)() PERFECTO,I.&VANDERMEER,J.() The agroecological matrix as Human Development Report . Sustaining Human Progress: alternative to the land-sparing/agricultural intensification model. Reducing Vulnerabilities and Building Resilience. United Nations Proceedings of the National Academy of Sciences of the United States Development Programme. New York, USA. of America, , –. UNEP (UNITED NATIONS ENVIRONMENT PROGRAMME)() The PHILLIPS, S.J., ANDERSON, R.P. & SCHAPIRE, R.E. () Maximum World Database on Protected Areas. Http://www.unep-wcmc.org entropy modeling of species geographic distributions. Ecological [accessed October ]. Modelling, , –. USGS (U.S. GEOLOGICAL SURVEY)() HYDRO K data. Https:// PHILLIPS, S.J. & DUDIK,M.() Modelling of species distributions lta.cr.usgs.gov/HYDROK [accessed September ]. with Maxent: new extensions and a comprehensive evaluation. VITO EARTH OBSERVATION () SPOT vegetation programme. Ecography, , –. Http://www.spot-vegetation.com/ [accessed September ]. RANSOM, C., ROBINSON, P.T. & COLLEN,B.() Choeropsis WEARN, O.R., ROWCLIFFE, J.M., CARBONE, C., BERNARD,H.& liberiensis. The IUCN Red List of Threatened Species : EWERS, R.M. () Assessing the status of wild felids in a e.TA. Http://dx.doi.org/./IUCN.UK.-. highly-disturbed commercial forest reserve in Borneo and the RLTS.TA.en [accessed March ]. implications for camera trap survey design. PLoS ONE, (), e. RODRIGUES, A.S.L., ANDELMAN, S.J., BAKARR, M.I., BOITANI, L., WESTERN, D., RUSSELL,S.&CUTHILL,I.() The status of wildlife BROOKS, T.M., COWLING, R.M. et al. () Effectiveness of the in protected areas compared to non-protected areas of Kenya. PLoS global protected area network in representing species diversity. ONE, (), e. Nature, , –. ROTH, H.H., HOPPE-DOMINIK, B., MÜHLENBERG, M., STEINHAUER-BURKART,B.&FISCHER,F.() Distribution and status of the hippopotamids in the Ivory Coast. African Zoology, , –. Biographical sketches STECK,B.() International Studbook for the Pygmy Hippopotamus . Zoo Basel, Basel, Switzerland. A NNIKA H ILLERS’ work focuses on the sustainable management of STEINMETZ, R., SRIRATTANAPORN, S., MOR-TIP,J.&SEUATURIEN,N. protected areas and the conservation of threatened species in West () Can community outreach alleviate poaching pressure and Africa. G RAEME B UCHANAN works primarily on the use of spatial recover wildlife in South-East Asian protected areas? Journal of data to answer applied conservation questions, from local to global Applied Ecology, , –. scales. J ERRY G ARTEH is a biologist with a strong interest in practical SWANEPOEL, L.H., LINDSEY, P., SOMERS, M.J., VAN HOVEN,W.& conservation and human–wildlife interactions in Liberia. S OLOMON DALERUM,F.() Extent and fragmentation of suitable leopard T OMMY has a special interest in community involvement for the ef- habitat in South Africa. Animal Conservation, , –. fective protection of natural habitats and threatened species in Sierra TREWHELLA, W.J., RODRIGUES-CLARK, K.M., CORP, N., ENTWISTLE, Leone. M OHAMED F OFANA works to facilitate the monitoring of wild- A., GARRETT, S.R.T., GRANEK, E. et al. () Environmental life populations to contribute to the effective management of protected education as a component of multidisciplinary conservation areas and their biodiversity in Sierra Leone. J EREMY L INDSELL works programs: lessons from conservation initiatives for Critically on a range of practical tropical forest conservation projects and is in- Endangered fruit bats in the western Indian Ocean. Conservation terested in the role of protection and the impact of degradation on Biology, , –. tropical forests and their threatened species.
Oryx, Page 10 of 10 © 2016 Fauna & Flora International doi:10.1017/S003060531600020X
http://journals.cambridge.org Downloaded: 08 Jun 2016 IP address: 108.171.128.188