How to Detect an Elusive Aquatic Mammal in Complex Environments? a Study of the Endangered Antillean Manatee Trichechus Manatus Manatus in French Guiana

Home , Aquatic mammal

How to detect an elusive aquatic mammal in complex environments? A study of the Endangered Antillean manatee Trichechus manatus manatus in French Guiana

D ELMA N ATALY C ASTELBLANCO-MARTÍNEZ,VIRGINIE DOS R EIS and B ENOIT DE T HOISY

Abstract The Antillean manatee Trichechus manatus man- manatus manatus was present and seen regularly along the atus is a cryptic mammal that inhabits, among other areas, coast and up to  km inland, but may have been decreasing murky rivers and estuaries of Central and South America. in abundance. As the alluvial coastal plain is narrow, the The difficulty in detecting and counting manatees is an obs- manatee’s habitat is naturally restricted to nearby rocky tacle to monitoring their population status, as traditional shores and some large rivers, such as the Maroni, count sampling is unsuitable. We conducted a quantitative Approuague and Oyapock. With their secretive behaviour, analysis of the distribution and abundance of the manatee relatively undisturbed estuarine habitats and legal protec- across its known range in the rivers and estuaries of tion, the outlook for the future of manatees in French French Guiana, based on data from a range-wide line tran- Guiana is positive (Artigas et al., ; de Thoisy et al., sect visual survey combined with a side-scan sonar survey, ). Nevertheless, the Antillean manatee is categorized identification of feeding sites, and silent observations at as Endangered on the IUCN Red List (Self-Sullivan & fixed points. A total effective effort of  hours and Mignucci-Giannoni, ) because of its restricted area of , km of boat survey was completed. We used two relative occurrence and small population size. The manatee faces a abundance indices: the combined encounter rate, which number of threats throughout its range, including habitat combines encounters per km from either visual or side- degradation and loss, collisions with watercraft, entangle- scan sonar surveys, and the global detection index, which ment in fishing gear, pollution, human disturbance, natural is defined as the sum of all evidence per unit time. disasters, and hunting. However, estimates of abundance Manatees were detected in all nine study units. Niche mod- and habitat preferences are still unavailable on a country- elling was used to analyse the space selection by manatees, wide basis. Making management decisions based on incor- and helped to detect differential use of habitats according to rect categorization of conservation status could accelerate the season. The model predicted that coastal areas are used extinction of a species as a result of insufficient management more extensively during the dry season. In the absence of effort (Pool, ). Hence, a protocol for conducting a long- better techniques to detect wild manatees in complex habi- term and large-scale census of manatees, adapted to local tats this monitoring protocol may be relevant and replicable constraints, is required. in hydrological systems where manatee detectability is con- Aerial surveys have been used as a tool to assess distribu- strained by biogeographical characteristics. tion and abundance of sirenian populations since the s (Reynolds et al., ). However, they are unsuitable for Keywords Cryptic species, French Guiana, manatee, niche counting manatees inhabiting river basins in French modelling, relative abundance index, side-scan sonar, South Guiana, and more widely in the Guianas region, because America, Trichechus manatus manatus of the difficulty in detecting the animals from an aircraft. Coastal areas of the Guiana Shield, including French Guiana, are particularly rich in sediments and tannins as a Introduction result of sediment from the Amazon River (DeMaster et al., ; Sylvestre et al., ), and consequently the waters of n French Guiana preliminary social surveys (de Thoisy  rivers and estuaries are turbid. Boat transect surveys using Iet al., ) suggested the Antillean manatee Trichechus distance sampling have been particularly useful in deter- mining distribution and population size for some riverine   DELMA NATALY CASTELBLANCO-MARTÍNEZ (Corresponding author) Consejo cetaceans (Martin et al., ; Zhao et al., ). However, Nacional de Ciencia y Tecnología—Universidad de Quintana Roo, Boulevard they assume that the population distribution is random or Bahía s/n esq. Ignacio Comonfort, Col. del Bosque, 77019 Chetumal, Quintana Roo, Mexico. E-mail [email protected] uniform and visibility of the animals is unobstructed (Aragones et al., ) but the validity of these assumptions VIRGINIE DOS REIS and BENOIT DE THOISY Association Kwata, Cayenne Cedex,  French Guiana for sirenian populations is uncertain (Reynolds et al., ). Received  May . Revision requested  June . Boat- and land-based surveys can be useful for studying the Accepted  August . First published online  January . distribution, abundance, behaviour, feeding ecology and

Oryx, 2018, 52(2), 382–392 © 2017 Fauna & Flora International doi:10.1017/S0030605316000922 Downloaded from https://www.cambridge.org/core. IP address: 170.106.34.90, on 27 Sep 2021 at 16:02:16, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605316000922 The manatee in French Guiana 383

migration of manatees and dugongs but are difficult to im- () Approuague and () Oyapock. The same field collection plement over a large area (Aragones et al., ). We provide methods and approximately the same survey effort were a quantitative analysis of the distribution and abundance of used in each unit. Boat-based line transect surveys were con- the Antillean manatee across its known range in the rivers ducted at – km per hour from a  m long boat with a  hp and estuaries of French Guiana, based on data from a range- outboard motor. Manatees were detected visually or by side- wide, line transect visual survey combined with a side-scan scan sonar ( SI Combo, Humminbird, Eufaula, USA). sonar survey. To our knowledge this is the first attempt to Transects were selected strategically according to navigabil- conduct a large-scale survey of a population of manatees ity and the suitability of habitats for manatees. Surveys in murky rivers, a common habitat type throughout a sig- were conducted in the middle and lower sections of the nificant portion of the species’ distribution. rivers (,  km from the mouth), as exploratory surveys indicated that manatees rarely use the upper parts of the rivers. In narrow rivers (,  m width), transects followed the Study area middle and deepest area of the watercourse. For wider rivers the boat was maintained at a distance of c.  m from one of French Guiana, on the north-east coast of South America, is  the shores. the largest French overseas department (, km ). The Once a potential manatee was detected by the sonar, the Maroni River marks the political border with Surinam to image was captured, along with the date, time, geographical the west, and the Oyapock River the border with Brazil in coordinates, depth and speed. Immediately, the boat was the east. French Guiana lies on the Precambrian Guiana turned back and passed over the location of the record, Shield, one of the largest blocks of primary tropical forest maintaining the same speed and trajectory, to rule out and a region rich in biodiversity (Hammond, ). false identification of inanimate objects as manatees. If the Coastal and marine biodiversity in French Guiana is influ- object moved away or changed its position we assumed it enced by the Amazon River, which is a major factor defining was a manatee. In addition to boat transects, fixed-point the geological structure of estuarine, coastal and shelf mar- surveys were conducted by at least four observers scanning ine ecosystems. The coastal waters are turbid, and extensive ° around the boat. Fixed-point surveys were also con- mudflats occur along the coast (Artigas et al., ). The cli- ducted from the shore, from an elevated vantage point mate is equatorial, characterized by dry and wet seasons, such as a hill or a rock. Indirect signs of manatee presence, with a maximum variation in mean temperature of °C. such as signs of feeding or excrement, were also recorded. The dry season is July–November, and the rainy season Typical signs of manatee feeding on leaves and stalks in December–June, interrupted by a variable period of drought macrophyte communities are easily distinguishable by ex- during March–April (Lam-Hoai et al., ). The hydro- perienced observers. graphic network of French Guiana is characterized by high density as a result of the annual rains and the gently sloping terrain, with maximum flow rate during May– June and minimum generally in November, when the Data analysis flow rate of the Amazon River is also considerably lower.  The images recorded using side-scan sonar were analysed a The river basins vary in size (c. –, km ; Tejerina- posteriori to detect manatees. The images selected (on the Garro et al., ). The tide is semi-diurnal, with amplitude basis of shape, size and position) were then organized and of up to . m (spring tides, mesotidal regimes). The coastal placed in a catalogue. A blind, online peer-review system habitats are considered to be well preserved, with limited was created, and a number of international experts in the loss and degradation of estuary areas, swamps and man- use of side-scan sonar imaging in manatee research were in- groves during the last decade (Lefebvre & Verger, ). vited to participate. The researchers were asked to analyse and score each photograph – according to the criteria out-  #  Methods lined in Table . Photographs with a mean score of were not considered to be evidence of manatee presence and were Field data acquisition discarded from the analysis. We calculated the encounter rate in each unit by counting Our study covered the entire coastal area of French Guiana the total number of manatees detected either visually or (c.  km of coastline and up to  km upstream along using side-scan sonar, and dividing by the total survey effort the rivers), all at ,  m altitude, during March–April (length in km) in the unit. As both side-scan sonar and visual , October–December  and January–April . surveys are likely to miss manatees, we defined a combined The study area was divided into nine units, covering estuar- encounter rate to represent the minimum encounter rates ies, beaches and rivers: () Coswine, () Mana, () Iracoubo, along the survey transect, based on the highest encounter () Sinnamary, () Kourou, () Cayenne, () Kaw, rate at any point along the river from either visual or sonar

TABLE 1 Criteria used to define confidence values for photographs categorization (mangroves, high forests, flooded forests captured using side-scan sonar as evidence of manatee Trichechus and swamps) derived from habitat mapping (Guitet et al., manatus manatus presence. ); and the density of the freshwater network, obtained % confidence in the from digital maps. The species occurrence and the pre- photograph as evidence diction of more suitable habitats were found to rely on Score Criterion of manatee presence elevation, mangroves, high forests and flooded forests. The 1 I am certain that there is no 0 two remaining variables (swamps and the freshwater net- manatee in the photograph work) did not contribute to the distribution. The model − 2 I am not sure, but probably 5 was run with a convergence threshold of  , a maximum NOT of , iterations, and linear/quadratic regularization. 3 I am not sure at all 50 Cross-validation is important in testing model performance, 4 I am not sure, but probably YES 75 especially in situations where it is hazardous, costly or im- 5 I am certain that there is at least 99  one manatee in the photograph possible to collect additional samples (Li et al., ). The -fold cross-validation method and area under the receiver operating characteristic curve (AUC) method were used to records (Martin et al., ; Zhao et al., ). In the case of assess the accuracy of the MaxEnt model. The AUC index the fixed-point surveys the relative abundance was estimated was used as a measurement of the discriminatory capacity from the number of manatees sighted per hour in each of the of classification models, which is independent of the choice nine study units. To provide an indicator of the use of the of threshold (Jiménez‐Valverde, ). study area by manatees (Alvarez-Alemán et al., ), we also calculated a global detection index as the sum of all evidence (presence/absence of feeding tracks, number of sight- Results ings and number of sonar detections) divided by the search effort in hours. Multiple pairwise comparisons (Kruskal– We completed  days of fieldwork, with an effective sam- Wallis one-way analysis of variance on ranks) were used to pling time of . hours, .% devoted to boat surveys investigate the differences in relative abundance indices be- (n = ) and .% to fixed-point surveys (n = ). The tween units, to detect spatial trends in manatee distribution. duration of boat surveys was – minutes (mean = . Seasonal differences in the global detection index and com- minutes); surveys from fixed points lasted – minutes bined encounter rate were also explored using a normality (mean = . minutes). The boat transects using side- test (Shapiro–Wilk) and paired t-test or a Mann–Whitney scan sonar covered ,. km at a mean vessel speed of – rank sum test if the normality failed. . km hour , and the mean sonar range was . m. We developed a species distribution model by using spa- The mean depth of the surveyed area was . m (range – tial environmental data to make inferences about habitat . m). Fifty-nine instances of manatee sign were recorded suitability for manatees. Conceptually, such models are in- in total, eight by fixed-point surveys and  by boat-based tended to determine and predict components of a species’ surveys (Fig. , Table ). ecological niche through space and time (Kearney & Fifteen direct sightings of manatees were recorded, eight Porter, ). We used MaxEnt ..k (Phillips et al., during boat surveys and seven during fixed-point surveys. ) to estimate the likelihood of maximum entropy distri- All were of solitary individuals except one sighting of two bution of each environmental variable across the study area. individuals in the Kaw River, probably a mother–calf The algorithm used by MaxEnt has been shown to produce dyad. During the boat transects , photographs were cap- reliable predictions (Elith et al., ) and to process pres- tured by the sonar, of which  were considered to be suit- ence-only data and small data sets effectively (Wisz et al., able candidates for the blind peer-review analysis. Six ). As the objective was to investigate possible changes experts accepted our invitation to score the catalogue online. in distribution between habitats according to season, two All except one (KAG, who has used this method mostly in datasets of manatee records (visual and side-scan sonar shallow swamps) had experience of using side-scan sonar in detections) were considered in the analysis: a first set to ex- habitats similar to those found in French Guiana. After dis- plore habitat use during the wet season, including the  cussing the scores provided by the experts, and based on our data ( March– April, n = ) and the  data ( own field experience, we selected  photographs as evi- February– April, n = ), and a second set to explore habi- dence of manatee presence (Plate ). Both sightings and de- tat use during the dry season using the  data (– tections by sonar were confirmed in most cases by the October, n = ). The following predictive environmental observation of typical feeding tracks of manatees. Riparian data were used in a preliminary assessment of the variables vegetation with typical signs of consumption by manatees of importance: elevation (digital elevation model acquired was observed during  boat surveys, and consisted of by the shuttle radar topography mission); habitat patches of aquatic macrophytes such as Mucu-mucu

FIG. 1 (a) Distribution and abundance of Antillean manatees Trichechus manatus manatus in French Guiana. (b−d) Distribution of evidences of manatee presence in nine survey units: , Coswine; , Mana; , Iracoubo; , Sinnamary; , Kourou; , Cayenne; , Kaw; , Approuague; , Oyapock.

(Crinum sp.) and Gramineae (Echinochloa polystachya). No mean global detection index is significantly greater for the manatees were detected either by sonar or visually along dry season than for the wet season could not be rejected % of the surveyed length, or during % of the survey (P = .). time (. hours). The niche models had comparable relevance for both Manatees were detected in all units, resulting in a mean seasons (AUC = . and . for wet and dry seasons, re- global detection index of . ± SD . records per hour, spectively). The percentage contribution of variables to – with the highest values in Coswine (. hour ), Mana manatee occurrence differed among seasons as follows: – – (. hour ) and Kaw (. hour ; Fig. ). Boat surveys mangroves, % in wet season, % in dry season; flooded produced a combined encounter rate of . ± SD . forests, % in wet season, .% in dry season; dry forests, per km. Nevertheless, there was not a statistically significant % in wet season, % in dry season. These results indicate a difference among areas in either combined encounter rate higher likelihood of manatee presence in the rivers during (H = .;df=,P=.) or global detection index the dry season, and a higher likelihood of manatee presence (H = .;df=,P=.). The combined encounter rate in coastal areas during the wet season (Fig. ). was greater during the dry season than during the wet sea- – son (. ± SD . and . ± SD . encounters km , respectively), with non-normal distribution of the data Discussion (Shapiro–Wilk, P , .), but the variation was not sufficient to exclude the possibility that it was a result of chance (t = We aimed to explore and validate methodological and .;df=,P=.). In contrast, the global detection analytical tools for the assessment of seasonal variations index was lower during the dry season than during the wet in habitat selection and key areas for a secretive mammal season (. ± SD . and . ± SD . records per in a region of conservation importance. The indices of hour, respectively), with a normal distribution of the data detection (i.e. combined encounter rate and global detec- (Shapiro–Wilk, P = .). However, the hypothesis that the tion index) were higher in surveys conducted at Coswine,

Kourou and Oyapock, which may be indicative of a higher population density of manatees in those areas. However, the data were not sufficient to demonstrate any significant statistical difference among study units. Also, the type of information obtained was not sufficient to infer whether the differences in the indices are attrib- utable to manatee abundance or to factors affecting manatee detection. Despite recording evidence of manatee presence in all nine units, detection rates were low throughout the study area, with the highest ranks obtained in only a small proportion of the surveys (Fig. ); for example, combined encounter rates of . . per km were recorded for only .%of the total km surveyed, and global detection indices of . . records per hour were recorded in only .% of the total effort in time. Previous studies conducted blind transect surveys to evaluate the effectiveness of side-scan sonar in various environmental conditions, with preliminary detections of  and % of the manatees present in the study areas in Florida and Mexico, respectively (Gonzalez- Socoloske et al., ). River basins in French Guiana have similar characteristics to those of Tabasco (Mexico). We detected manatees  times with the sonar, and if we assume a detection rate of % we estimate the total number of manatees present was  individuals. This suggests the manatee population in French Guiana is small compared to other populations. Nonetheless it may be argued, based on interviews and on the availability of extensive high-quality habitat, that manatees are relatively abundant and well conserved in French Guiana, and we suggest that manatee ), obtained by direct sightings, side-scan sonar, and the presence or absence of feeding scars.  detection rates are lower in French Guiana than in other

Fig. areas. Our detection capabilities were probably hampered by the manatee’s elusive nature, and the noise of the engine is likely to have disturbed manatees in the vicinity of the boat, causing them to flee rapidly from the capture area of the side-scan sonar. There is evidence that manatees are capable of localizing the sounds produced by boats (Colbert-Luke et al., ), and of reacting to them; for example, disruption of a manatee’s call behaviour has

No. of records of manatee presence (No.Coswine of records per hour) Mana Iracoubo Sinnamary Kourou Cayenne Kaw Approuague Oyapockbeen Total demonstrated at boat–manatee distances of ,  m (Rivera-Chavarría et al., ). As a boat approaches, manatees respond by increasing their swim speed and orienting towards deeper waters (Nowacek et al., ). This bias is likely to be augmented in narrow rivers, where manatees will move downstream or upstream (i.e. not alongside the boat), beyond the capture range of the side-scan sonar. Corroboration of sonar captures in situ was not possible: we failed to visually confirm any of the images captured by the sonar. In French Guiana manatee habitats consist mostly of meandering rivers, and therefore the visible area of the water surface is limited, making it difficult to observe

2 Records of manatee presence in nine study units in French Guiana ( signs of manatees breathing far from the vessel. Also, the waters are generally turbid, making it impossible to visually Direct sightingsSide-scan sonar detectionsFeeding scarsFixed pointBoat based 5 (0.163) 2 (0.065) 3 (0.116) 7 (0.229) 2 (0.070) 1 (0.035) 5 (0.193) 1 (0.052) 2 (0.103) 1 (5.000) 1 (0.035) 13 (0.472) 1 1 (0.049) 4 (0.556) (0.194) 7 (0.291) 1 2 (0.027) 2 (0.324) (0.054) 2 (0.089) 1 4 (0.207) (0.104) 2 2 (0.052) (0.138) 1 (0.032) 2 (0.097) 7 (0.449) 3 (0.111) 9 (0.277) 3 (0.185) 4 6 (0.150) (0.156) 15 (0.060) 19 (0.077) 3 (0.095) 4 (0.307) 3 (0.500) 51 1 (0.247) (0.062) 25 (0.101) 8 (0.191) ABLE Type of evidence Survey method Total no. of records (GlobalSurvey detection effort index) (hours)Survey effort (km) 14 (0.457) 8 (0.309) 4 (0.139) 3 (0.155) 7 (0.340) 3 30.62 (0.082) 168.9 12 (0.312) 25.85 4 (0.127) 114.9 28.72 4 (0.247) 104.61 59 19.33 (0.238) 52.21 20.58 82.39 36.80 138.45 38.43 262.2 31.52 135.42 16.2 70.27 1,129.35 248.05 T detect manatees that are underwater.

PLATE 1 Examples of side-scan sonar images selected as positive for Antillean manatees in (a) Kourou,  October ; (b) Cayenne,  October ; (c) Sinnamary,  October ; and (d) Oyapock,  October  (Fig. ). M, manatee; S, shadow.

The main limiting factor in the use of side-scan sonar is its produce a signature shadow that trees, fishes, reptiles, rocks unknown detection rate (Gonzalez-Socoloske & Olivera- and branches do not; however, field experience and a trained Gomez, ). The detection capability can be affected by eye are needed to interpret the shape and shadow correctly. weather, boat speed, depth, current speed and direction, Manatees are identified in sonar images based on the follow- type of environment, and observer experience. The field biol- ing criteria: the manatee’s unique signature peanut shape, ogists who participated in the boat surveys and in the manipu- the paddle shape of the tail, a small head and flippers, and lation of the side-scan sonar had various levels of experience the signature shadow (Brice, ). Reptiles are also distinct- andskill.Finally,thedatawerecollectedovera-month per- ive to a trained observer (Davy & Fenton, ); for example, iod, with no quantified variations in general hydro-climatic a crocodile creates an elongated shape with an extended, conditions (e.g. depth, which may affect manatee detection). narrow snout. The exact length of an animal detected by As most of these factors cannot be controlled and are still un- sonar cannot be determined because the image is influenced known or poorly understood, it is impossible to correct the by body position, vessel speed and water-depth distortions. error resulting from detectability conditions in the field. However, the approximate length of a manatee’s acoustic Another factor to take into account when using side-scan response compared to other objects and the lateral sonar is the subjectivity of recognizing manatees in sonar range scale can be used as a size indicator (Brice, ). images. Following the precautionary principle, we tried to Identification was facilitated by the fact that the manatee avoid or reduce overestimation of manatee numbers. was the largest completely aquatic vertebrate in the surveyed However, our method was still subjective and limited by a system. number of human factors. Errors in image interpretation Comparison of relative abundance indices (combined en- may include mistaking an object or another large animal counter rate and global detection index) indicated that mana- for a manatee. Species of megafauna inhabiting water- tee surveys in turbid rivers are more efficient and productive courses and estuaries in French Guiana include the torche when several methods and evidence sources are combined catfish Brachyplatystoma filamentosum (Le Bail et al., (Fig. ). As none of the methods are infallible, collecting mul- ), the black caiman Melanosuchus niger (de Thoisy tiple types of evidence increases confidence in the presence/ et al., ), the spectacled caiman Caiman crocodilus absence data. Although we couldn’t confirm any visual detec- (Vasconcelos et al., ), the giant otter Pteronura brasi- tions with images from the side-scan sonar (or vice versa), liensis, the neotropical river otter Lontra longicaudis, and previous research has validated the use of side-scan sonar some species of turtles (e.g. Podocnemis expansa). The to detect manatees in turbid waters (Gonzalez-Socoloske giant otter has not been reported in habitats used by mana- et al., ; Gonzalez-Socoloske & Olivera-Gomez, ; tees (Huguin & de Thoisy, ). Fishes with soft bodies, Arévalo-González et al., ;Brice,). such as catfishes, may absorb too much of the sonar signal to be detected. Larger and more reflective (hard-bodied) Seasonality in habitat use fishes, and reptiles such as crocodiles and turtles, are more likely to be imaged by side-scan sonar (Flowers & Although the seasonal variation in combined encounter Hightower, ). Nevertheless, in general terms, manatees rates and global detection indices was not statistically

FIG. 2 Relative abundance of Antillean manatees in nine study units in French Guiana (Fig. ), estimated by (a) combined encounter rate and (b) global detection index. The percentages indicate the proportion of effort in distance (km, combined encounter rate) or time (hours, global detection index) where the index rank was recorded. The colour gradient represents the magnitude of the indices of relative abundance, with the darkest boxes representing the lowest values.

significant, our niche modelling predicted a greater likeli- when possible (Olivera-Gómez & Mellink, ). Hence, hood of manatee presence in rivers during the dry season, the temporal variation in freshwater availability may explain whereas during the wet season manatees seem to prefer the seasonality of habitat selection by manatees in French coastal areas, as noticed also by local people (B. de Thoisy, Guiana, where there is a strong seasonal fluctuation in the pers. obs.). Similar seasonality in habitat preference has relative contributions of marine and continental waters to been reported for coastal manatees in Belize, where there estuarine waters (e.g. during the dry season the estuarine was found to be a greater probability of manatee presence waters of the Kaw River comprise a mixture of approxi- in the cay habitat (marine environment) than in rivers mately equal amounts of freshwater and seawater (Lam- (freshwater) during the wet season (Auil-Gomez, ). Hoai et al., ), and manatees may need to travel In the tropics manatee distribution is influenced by season- upstream to find freshwater to drink). During the rainy sea- ality (i.e. rainy or dry; Pablo‐Rodríguez et al., ) rather son, manatees may remain in coastal areas as the salinity of than fluctuations in water temperature (i.e. summer or win- estuarine waters corresponds to that of fluvial waters origin- ter) as occurs in subtropical regions such as Florida, USA ating from the drainage of the river (Lam-Hoai et al., ). (Irvine, ). In tropical freshwater environments where a strong flood pulse occurs (e.g. the Amazon, Orinoco and Usumacinta basins) manatees tend to disperse over larger Conservation areas to feed during the rainy season but in the dry season, when the water level drops, they remain confined to smaller All the localities in French Guiana previously reported by areas (Arraut et al., ; Castelblanco-Martínez et al., ; interviews and surveys as being important areas for mana- Pablo-Rodríguez et al., ). However, when tropical tees were confirmed in the field by direct sightings, sonar manatees inhabit both coastal and marine areas, the acces- images and evidence of feeding. The recording of the sibility of freshwater may become a major limiting factor for three types of evidence in all of the basins surveyed indicates their distribution and habitat utilization. There is extensive that the population is distributed widely across the territory. evidence of the association between occurrence of the Additionally, previous interviews (de Thoisy et al., , Antillean manatee and freshwater sources (e.g. Powell Castelblanco-Martínez, , unpubl. data) revealed that en- et al., ; Jiménez, ; Olivera-Gómez & Mellink, counters with manatees in French Guiana are relatively ; Castelblanco-Martínez et al., a; Landero et al., common in comparison with other areas of the species’ dis- ), and tracking studies have found that Antillean mana- tribution, suggesting that the population there is relatively tees inhabiting salty environments travel several kilometers abundant. to the mainland to satisfy their need for freshwater Although narrow, given the restricted alluvial coastal (Castelblanco-Martínez et al., b). The manatee’s need plain, suitable habitats for manatees are widely available in to imbibe freshwater is unusual among marine mammals French Guiana, and include dense and well-preserved areas (MacAvoy et al., ), and although the species apparently of mangrove (Fromard et al., ). Habitat loss and modi- has adaptations to survive in salt water (Ortiz et al., ) the fication is a significant threat to manatees globally, but it use of such mechanisms could have a high energy cost, mak- does not seem to be a major concern in French Guiana, ing it advantageous to remain close to freshwater sources where . % of the forests are owned by the French state,

FIG. 3 Map of predicted potential distributions of manatees in French Guiana during the dry season, based on four environmental data sets: elevation, mangroves, high forests and flooded forests. Model predictions are illustrated on a spectrum from white (the model predicts absence) to black (the model predicts presence). The results indicate a higher likelihood of manatee presence in the rivers during the dry season (black areas).

and are protected against large-scale threats (de Thoisy et al., ). Only two coastal nature reserves, the Amana and Kaw Nature Reserves, provide protection for manatee habitats; however, the Conservatoire du Littoral, a govern- ment administrative unit, ensures protection of other coastal areas. Circa % of coastal and estuarine area is legally

protected, and European Union regulations, national de- FIG. 4 Percentage of the survey effort in (a) distance and crees and several laws provide protection outside govern- (b) time, with cumulative number of records of manatee ment-protected areas; for example, the Loi sur presence in French Guiana (Fig. ). The combined encounter rate l’Eau (Water Law, EU) and the Loi sur la Biodiversity (a) combines only visual and side-scan sonar detections from (Biodiversity Law, France) not only address species protec- boat surveys, whereas the global detection index (b) considers all tion but also incorporate a holistic approach towards the types of evidence: sightings (from boat and fixed-point surveys), side-scan sonar detections, and feeding scars. protection of ecological connectivity and networks. Manatee mortality as a result of entanglement or inten- tional hunting is apparently low, although awareness of the to manatees, such as contamination caused by gold mining species needs to be reinforced. However, collision with water activities, are still poorly understood and merit further vessels may be an increasing threat to manatees, especially consideration. in areas with higher human population density, such as Assessing manatee abundance and distribution in turbid Cayenne, and areas frequented by fishers, such as the estu- waters is one of the main obstacles to evaluating the conser- aries. Coswine and Oyapock are located on the international vation status of this species in the region. Our research has border (with Surinam and Brazil, respectively), where the demonstrated advances in the methodology for surveying threats to manatees, including illegal hunting, may be high- freshwater and estuarine manatees, particularly in geomor- er, uncontrolled or unknown. Telemetry studies have shown phically and hydraulically complex riverine and estuarine that Antillean manatees can travel up to hundreds of kilo- habitats, where distance sampling methods are generally in- metres (Castelblanco-Martínez et al., a; Normande appropriate. We developed indices of relative abundance et al., ), and it is likely that at least some individuals that may reflect spatial and temporal trends in population are crossing back and forth between nations. There is size and facilitate comparisons between locations and, in fu- therefore a need for a wider regional approach to manatee ture, years. Furthermore, we incorporated niche modelling conservation in northern South America. Other threats as a novel analytical tool to investigate manatee

distributions and seasonality in habitat use. Our results con- in Developing Countries (eds E.M. Hines, J.E. Reynolds, III, firmed the presence of manatees in all the rivers and estuar- L. Aragones, A.A. Mignucci-Giannoni & M. Marmontel), pp. –  ies of French Guiana (suggesting a temporal variation in the . University Press of Florida, Gainesville, USA. ARÉVALO-GONZÁLEZ, G.K., CASTELBLANCO-MARTÍNEZ, D.N., use of rivers and coastal areas, probably related to freshwater SÁNCHEZ-PALOMINO, P., LÓPEZ-ARÉVALO, H.F. & MARMONTEL, use and availability), and reinforced the information ob- M. () Complementary methods to estimate population size tained previously through interviews. As the conservation of Antillean manatees (Sirenia:Trichechidae) at Ciénaga de status of manatees in Amapá State (Brazil) and Surinam is Paredes, Santander, Colombia. Journal of Threatened Taxa, , –  poorly known, Oyapock and Coswine merit particular at- . ARRAUT, E.M., MARMONTEL, M., MANTOVANI, J.E., NOVO, E.M.L.M., tention in the development of management plans, and inter- MACDONALD, D.W. & KENWARD, R.E. () The lesser of two national effort is needed to coordinate actions for manatee evils: seasonal migrations of Amazonian manatees in the Western preservation. More information is needed regarding the Amazon. Journal of Zoology, , –. health status, genetic structure and movement of the mana- ARTIGAS, L.F., VENDEVILLE, P., LEOPOLD, M., GUIRAL,D.&  tee population in French Guiana to provide a better under- TERNON, J.-F. ( ) Marine biodiversity in French Guiana: estuarine, coastal, and shelf ecosystems under the influence of standing of its status, abundance and vulnerability and to Amazonian waters. Gayana, , –. inform the development of adequate management plans AUIL-GOMEZ, N.E. () Abundance and distribution trends of the for the species. West Indian manatee in the coastal zone of Belize: implications for conservation. MSc thesis. Texas A&M University, College Station, USA. Acknowledgements BRICE, C.E. () The detection of Amazonian manatees (Trichechus inunguis) using side-scan sonar and the effect of oil activities on their This study was funded jointly by the French agencies habitats in eastern Ecuador. MSc thesis. Nova Southeastern Direction de l’Environnement, de l’Aménagement et du University Oceanographic Center, Dania Beach, USA. CASTELBLANCO-MARTÍNEZ, D.N., BERMÚDEZ-ROMERO, A.L., Logement de Guyane and Parc National de La Guadeloupe. GÓMEZ-CAMELO, I.V., ROSAS, F.C.W., TRUJILLO,F.&ZERDA- The research benefited from technical support from the ORDOÑEZ,E.() Seasonality of habitat use, mortality and Marais de Kaw-Roura and Amana Nature Reserves, l’Office reproduction of the Vulnerable Antillean manatee Trichechus National de la Chasse et de la Faune Sauvage and the Protocol manatus manatus in the Orinoco River, Colombia: implications for  – Concerning Specially Protected Areas and Wildlife–Regional conservation. Oryx, , . Activity Centre. We thank Sébastien Barrioz, Billitis Le CASTELBLANCO-MARTÍNEZ, D.N., PADILLA-SÁLDIVAR, J., HERNÁNDEZ-ARANA, H.A., SLONE, D.H., REID, J.P. & MORALES- Guirriec, Sebastién Rives, Régis Gomes, Boris Lerebours VELA,B.(a) Movement patterns of Antillean manatees in and Mathieu Rhoné for their collaboration during field sur- Chetumal Bay (Mexico) and coastal Belize: a challenge for regional veys. Katherine Arévalo González, Leon David Olivera, Katie conservation. Marine Mammal Science, , –. Tripp and three anonymous researchers generously donated CASTELBLANCO MARTÍNEZ, D.N., POWELL, J., GALVES,J.&AUIL  valuable time to score and comment on the side-scan sonar GOMEZ,N.( b) Preliminary information from first tagged manatees in Turneffe Atoll (Belize) reveal regular travel patterns to images. The comments of two anonymous reviewers helped the mainland. Sirenews, , –. to improve the article. COLBERT-LUKE, D.E., GASPARD, J.C., REEP, R.L., BAUER, G.B., DZIUK, K., CARDWELL,A.&MANN, D.A. () Eight-choice sound localization by manatees: performance abilities and head Author contributions related transfer functions. Journal of Comparative Physiology A, , –. DNCM and BdT conceived, designed and managed the sur- DAVY, C.M. & FENTON, M.B. () Technical note: side-scan sonar veys. BdT and VdR coordinated the field work. VdR, BdT enables rapid detection of aquatic reptiles in turbid lotic systems.  – and DNCM collected data in the field. DNCM performed European Journal of Wildlife Research, , . DEMASTER, D.J., SMITH,JR, W.O., NELSON, D.M. & ALLER, J.Y. () statistical analysis and mapping. BdT developed the habitat Biogeochemical processes in Amazon shelf waters: chemical niche model. DNCM, BdT and VdR wrote the article. distributions and uptake rates of silicon, carbon and nitrogen. Continental Shelf Research, , –. DE THOISY, B., FAYAD, I., CLÉMENT, L., BARRIOZ, S., PETRIDOU, M., References POIRIER,E.&GOND,V.() Predators, prey and habitat structure: can we detect key conservation areas and early signs ALVAREZ-ALEMÁN, A., ANGULO-VALDÉS, J.A., GARCÍA ALFONSO, E., of population collapses in a neotropical forest? PLoS ONE , POWELL, J.A. & TAYLOR, C.R. () Occurrence of the Endangered e. Antillean manatee Trichechus manatus manatus in a marine DE THOISY, B., HRBEK, T., FARIAS, I.P., VASCONCELOS, W.R. & protected area, Isla de la Juventud, Cuba. Oryx, http://doi.org/. LAVERGNE,A.() Genetic structure, population dynamics, and /S. conservation of black caiman (Melanosuchus niger). Biological ARAGONES, L.V., LACOMMARE, K., KENDALL, S., CASTELBLANCO- Conservation, , –. MARTÍNEZ, D.N. & GONZÁLEZ-SOCOLOSKE,D.() Boat and DE THOISY, B., SPIEGELBERGER, T., ROUSSEAU, S., TALVY, G., VOGEL, land-based surveys. In Sirenian Conservation: Issues and Strategies I. & VIE, J.-C. () Distribution, habitat, and conservation status

of the West Indian manatee Trichechus manatus in French Guiana. fluviatilis) in the Central Amazon. Marine Mammal Science, , Oryx, , –. –. ELITH, J., GRAHAM, C.H., ANDERSON, R.P., DUDÍK, M., FERRIER, S., NORMANDE, I.C., LUNA, F.D.O., MALHADO, A.C.M., BORGES, J.C.G., GUISAN,A.etal.() Novel methods improve prediction of VIANA JUNIOR, P.C., ATTADEMO, F.L.N. & LADLE, R.J. () species’ distributions from occurrence data. Ecography, , –. Eighteen years of Antillean manatee Trichechus manatus manatus FLOWERS, H.J. & HIGHTOWER, J.E. () A novel approach to releases in Brazil: lessons learnt. Oryx, , –. surveying sturgeon using side-scan sonar and occupancy modeling. NOWACEK, S.M., WELLS, R.S., OWEN, E.C.G., SPEAKMAN, T.R., Marine and Coastal Fisheries: Dynamics, Management, and FLAMM, R.O. & NOWACEK, D.P. () Florida manatees, Ecosystem Science, , –. Trichechus manatus latirostris, respond to approaching vessels. FROMARD, F., PUIG, H., MOUGIN, E., MARTY, G., BETOULLE, J.L. & Biological Conservation, , –. CADAMURO,L.() Structure, above-ground biomass and OLIVERA-GÓMEZ, L.D. & MELLINK,E.() Distribution of the dynamics of mangrove ecosystems: new data from French Guiana. Antillean manatee (Trichechus manatus manatus) as a function of Oecologia, , –. habitat characteristics, in Bahia de Chetumal, Mexico. Biological GONZALEZ-SOCOLOSKE,D.&OLIVERA-GOMEZ, L.D. () Gentle Conservation, , –. giants in dark waters: using side-scan sonar for manatee research. ORTIZ, R.M., WORTHY, G.A. & BYERS, F.M. () Estimation of The Open Remote Sensing Journal, , –. water turnover rates of captive West Indian manatees (Trichechus GONZALEZ-SOCOLOSKE, D., OLIVERA-GOMEZ, L.D. & FORD, R.E. manatus) held in fresh and salt water. Journal of Experimental () Detection of free-ranging West Indian manatees Trichechus Biology, , –. manatus using side-scan sonar. Endangered Species Research, , PABLO‐RODRÍGUEZ, N., OLIVERA‐GÓMEZ, L.D., AURIOLES‐ –. GAMBOA,D.&VEGA‐CENDEJAS, M.E. () Seasonal differences GUITET, S., PÉLISSIER, R., BRUNAUX, O., JAOUEN,G.&SABATIER,D. in the feeding habits of the Antillean manatee population () Geomorphological landscape features explain floristic (Trichechus manatus manatus) in the fluvial‐lagoon systems of patterns in French Guiana rainforest. Biodiversity and Conservation, Tabasco, Mexico. Marine Mammal Science, , –. , –. PHILLIPS, S.J., ANDERSON, R.P. & SCHAPIRE, R.E. () Maximum HAMMOND, D.S. () Ancient land in a modern world. In Tropical entropy modeling of species geographic distributions. Ecological Forests of the Guiana Shield: Ancient Forests in a Modern World Modelling, , –. (ed. D.S. Hammond), pp. –. CABI, Wallingford, UK. POOL,M.() Incorrect use of the IUCN classification may lead HUGUIN,M.&DE THOISY,B.() Liste Rouge des Mammifères Non to bad management decisions. The case of the Trichechus Volants de Guyane. Association Kwata/DEAL Guyane/IUCN, manatus manatus in Suriname. Academic Journal of Suriname, , France. –. IRVINE, A.B. () Manatee metabolism and its influence on POWELL, J.A., BELITSKY, D.W. & RATHBUN, G.B. () Status of the distribution in Florida. Biological Conservation, , –. West Indian manatee (Trichechus manatus) in Puerto Rico. Journal JIMÉNEZ,I.() Heavy poaching in prime habitat: the conservation of Mammalogy, , –. status of the West Indian manatee in Nicaragua. Oryx, , –. REYNOLDS, J., MORALES-VELA, B., LAWLER,I.&EDWARDS,H.() JIMÉNEZ‐VALVERDE,A.() Insights into the area under the Utility and design of aerial surveys for Sirenians. In Sirenian receiver operating characteristic curve (AUC) as a discrimination Conservation: Issues and Strategies in Developing Countries (eds measure in species distribution modelling. Global Ecology and E.M. Hines, J.E. Reynolds, III, L. Aragones, A.A. Mignucci- Biogeography, , –. Giannoni & M. Marmontel), pp. –. University Press of KEARNEY,M.&PORTER,W.() Mechanistic niche modelling: Florida, Gainesville, USA. combining physiological and spatial data to predict species’ ranges. RIVERA-CHAVARRÍA, M., CASTRO,J.&CAMACHO,A.() The Ecology Letters, , –. relationship between acoustic habitat, hearing and tonal LAM-HOAI, T., GUIRAL,D.&ROUGIER,C.() Seasonal change of vocalizations in the Antillean manatee (Trichechus manatus community structure and size spectra of zooplankton in the Kaw manatus, Linnaeus, ). Biology Open, , –. River estuary (French Guiana). Estuarine, Coastal and Shelf Science, SELF-SULLIVAN,C.&MIGNUCCI-GIANNONI,A.() Trichechus , –. manatus ssp. manatus. The IUCN Red List of Threatened Species LANDERO, M.M., DE LOS ÁNGELES LICEAGA‐CORREA,M.& : e.TA. Http://dx.doi.org/./IUCN.UK.. MORALES‐VELA,B.() Ecological distribution of manatee RLTS.TA.en [accessed  September ]. (Trichechus manatus manatus) in Bahía de la Ascensión, Mexico. SYLVESTRE, F., GUIRAL,D.&DEBENAY, J.P. () Modern diatom Marine Mammal Science, , –. distribution in mangrove swamps from the Kaw Estuary (French LE BAIL, P.Y., COVAIN, R., JÉGU, M., FISCH-MULLER, S., VIGOUROUX, Guiana). Marine Geology, , –. R. & KEITH,P.() Updated checklist of the freshwater and TEJERINA-GARRO, F.L., DE MÉRONA, B., OBERDORFF,T.&HUGUENY, estuarine fishes of French Guiana. Cybium, , –. B. () A fish-based index of large river quality for French Guiana LEFEBVRE, J.P. & VERGER,G.() Suivi de l’occupation du sol et des (South America): method and preliminary results. Aquatic Living changements d’occupation du sol en Guyane française entre  et Resources, , –. . MEDDE, IGN, ONF, French Guiana. VASCONCELOS, W.R., HRBEK,T.,DA SILVEIRA, R., DE THOISY, B., LI, G., DU,S.&GUO,K.() Evaluation of limiting climatic factors MARIONI,B.&FARIAS, I.P. () Population genetic analysis of and simulation of a climatically suitable habitat for Chinese sea Caiman crocodilus (Linnaeus, ) from South America. Genetics buckthorn. PLoS ONE, (), e. and Molecular Biology, , –. MACAVOY, S.E., BACALAN, V., KAZANTSEVA, M., RHODES,J.&KIM, WISZ, M.S., HIJMANS, R.J., LI, J., PETERSON, A.T., GRAHAM, C.H. & K. () Sulfur isotopes show importance of freshwater primary GUISAN,A.() Effects of sample size on the performance production for Florida manatees. Marine Mammal Science, , of species distribution models. Diversity and Distributions, , – –. . MARTIN, A.R., DA SILVA, V.M.F. & SALMON, D.L. () Riverine ZHAO, X., WANG, D., TURVEY, S.T., TAYLOR,B.&AKAMATSU,T. habitat preferences of botos (Inia geoffrensis) and tucuxis (Sotalia () Distribution patterns of Yangtze finless porpoises in the

Yangtze River: implications for reserve management. Animal Regional Co-chair for South America of the IUCN Sirenian Conservation, , –. Specialist Group and is currently conducting a long-term megafauna monitoring project in the Caribbean. V IRGINIE DOS R EIS works for a wildlife conservation NGO in French Guiana and is Biographical sketches involved in conservation programmes for marine vertebrates. B ENOIT DE T HOISY has worked on the research and conservation D ELMA N ATALY C ASTELBLANC-MARTÍNEZ is a conservation of large vertebrates for  years. His main areas of expertise are biologist studying aquatic vertebrates in Latin America. Her work ecology, population and habitat management, taxonomy and gen- focuses on conservation issues, population dynamics, behaviour etics, and awareness issues. and ecology of aquatic mammals, particularly manatees. She is