Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution,...

12
Aquatic Mammals 2002, 28.3, 312–323 Distribution and abundance of the river dolphin (Inia georensis) in the Tijamuchi River, Beni, Bolivia Enzo Aliaga-Rossel Bolivian Collection of Fauna, Institute of Ecology, Universidad Mayor de San Andres, PO Box 3182, La Paz, Bolivia Abstract Very few studies have been conducted in Bolivia regarding the distribution, behavior, or ecology of the Amazon river dolphin (Inia georensis). The only published studies of bufeo in Bolivia are from Pilleri (1969) and Pillieri & Gihr (1977) and were not quantitative investigations of river dolphin abundance or distribution. The work presented here consists of an estimate of abundance and descrip- tion of the seasonal distribution of the bufeo in the Tijamuchi River, Beni, Bolivia. The study was conducted during January 1998–September 1999 and represents four hydro-climatic seasons (i.e., low, high, rising, and falling waters). Two hundred and twelve h were spent in survey eort. The total study area was approximately 185 km along the river. Strip transects were used to survey for dol- phins. Dolphin distributions among three habitats were compared; these habitats were riverine- blackwaters, riverine-mixed waters (black and white), and oxbow lakes of the river system. Group size and age structure were recorded. Any dead dolphins were necropsied. On average, 208 bufeos were observed in the Tijamuchi River, with an average encounter rate of 1.12 dolphins/linear km. Dolphins were seen most frequently during low and falling water (56% of total observations) and least often during high waters (22% of total observa- tions). These seasonal dierences were statistically significant. Dolphins were seen most often in oxbow lakes, and next often in confluences. Dierences in dolphin abundance according to water colour were not statistically significant. There was some evi- dence of associations between group size and sea- son, and group size and water colour. Forty-two per cent of observations were of solitary dolphins, 32% were of pairs, and maximum group size was 19. Calves were seen most often during falling and low waters. Causes of mortality of dolphins in the study area were identified as entanglement in fishing nets, intra-specific aggression, and collisions with outboard motors. Key words: river dolphin, Inia georensis, bufeo, boto, abundance, distribution, mortality, photo- identification, Tijamuchi river, Bolivia. Introduction There are few studies that have been conducted in Bolivia regarding the distribution, behaviour, or ecology of the boto or Amazon river dolphin ‘bufeo’, Inia georensis. The only published work consists of that of Pilleri (1969) and Pilleri & Gihr (1977). This work was descriptive in nature and lacked quantitative analysis of river dolphin abun- dance. Despite this, Pilleri & Gihr (1977) concluded that their survey of river dolphin populations in various rivers and lagoons of Bolivia revealed a visible reduction in river dolphin population size in some regions, which they attributed to changes in the ecosystem or to anthropogenic pressure. Based on this, they recommended that immediate action be taken to establish nature reserves to actively protect the Bolivian bufeo. One additional study of river dolphins in Bolivia was conducted by Yan ˜ez (1999) and consisted of a description of the behav- ior and general ecology of the bufeo in the Iten ˜ez River and the Paragua River of the Noel KempMercado National Park. Laws that explicitly protect the bufeo do not currently exist in Bolivia. Some protection is aorded by the Veda General Indefinida (D.S. 25458), enacted in July of 1999, which is a general decree that prohibits the harassment, capture, harvest, and training of wild animals and their derivatives. Debate exits over the eectiveness of this law, but it continues to be in eect in the absence of more stringent or specific laws. Throughout South America, Inia georensis is classified as ‘vulnerable’ by the IUCN. Uncertainty exists with respect to the taxonomy of Inia georensis. The genus Inia is con- sidered monospecific with three subspecies: Inia g. humboldtiana found in the Orinoco Basin, I. g. georensis in the Amazon Basin, and I. g. boliviensis 2002 EAAM

Transcript of Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution,...

Page 1: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

Aquatic Mammals 2002, 28.3, 312–323

Distribution and abundance of the river dolphin (Inia geoffrensis) inthe Tijamuchi River, Beni, Bolivia

Enzo Aliaga-Rossel

Bolivian Collection of Fauna, Institute of Ecology, Universidad Mayor de San Andres, PO Box 3182, La Paz, Bolivia

Abstract

Very few studies have been conducted in Boliviaregarding the distribution, behavior, or ecologyof the Amazon river dolphin (Inia geoffrensis). Theonly published studies of bufeo in Bolivia are fromPilleri (1969) and Pillieri & Gihr (1977) and werenot quantitative investigations of river dolphinabundance or distribution. The work presented hereconsists of an estimate of abundance and descrip-tion of the seasonal distribution of the bufeo in theTijamuchi River, Beni, Bolivia. The study wasconducted during January 1998–September 1999and represents four hydro-climatic seasons (i.e.,low, high, rising, and falling waters). Two hundredand twelve h were spent in survey effort. The totalstudy area was approximately 185 km along theriver. Strip transects were used to survey for dol-phins. Dolphin distributions among three habitatswere compared; these habitats were riverine-blackwaters, riverine-mixed waters (black andwhite), and oxbow lakes of the river system. Groupsize and age structure were recorded. Any deaddolphins were necropsied. On average, 208 bufeoswere observed in the Tijamuchi River, with anaverage encounter rate of 1.12 dolphins/linear km.Dolphins were seen most frequently during low andfalling water (56% of total observations) and leastoften during high waters (22% of total observa-tions). These seasonal differences were statisticallysignificant. Dolphins were seen most often in oxbowlakes, and next often in confluences. Differences indolphin abundance according to water colour werenot statistically significant. There was some evi-dence of associations between group size and sea-son, and group size and water colour. Forty-twoper cent of observations were of solitary dolphins,32% were of pairs, and maximum group size was 19.Calves were seen most often during falling and lowwaters. Causes of mortality of dolphins in the studyarea were identified as entanglement in fishingnets, intra-specific aggression, and collisions withoutboard motors.

Key words: river dolphin, Inia geoffrensis, bufeo,boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia.

Introduction

There are few studies that have been conducted inBolivia regarding the distribution, behaviour, orecology of the boto or Amazon river dolphin‘bufeo’, Inia geoffrensis. The only published workconsists of that of Pilleri (1969) and Pilleri & Gihr(1977). This work was descriptive in nature andlacked quantitative analysis of river dolphin abun-dance. Despite this, Pilleri & Gihr (1977) concludedthat their survey of river dolphin populations invarious rivers and lagoons of Bolivia revealed avisible reduction in river dolphin population size insome regions, which they attributed to changes inthe ecosystem or to anthropogenic pressure. Basedon this, they recommended that immediate actionbe taken to establish nature reserves to activelyprotect the Bolivian bufeo. One additional study ofriver dolphins in Bolivia was conducted by Yanez(1999) and consisted of a description of the behav-ior and general ecology of the bufeo in the ItenezRiver and the Paragua River of the Noel KempffMercado National Park.

Laws that explicitly protect the bufeo do notcurrently exist in Bolivia. Some protection isafforded by the Veda General Indefinida(D.S. 25458), enacted in July of 1999, which is ageneral decree that prohibits the harassment,capture, harvest, and training of wild animals andtheir derivatives. Debate exits over the effectivenessof this law, but it continues to be in effect in theabsence of more stringent or specific laws.Throughout South America, Inia geoffrensis isclassified as ‘vulnerable’ by the IUCN.

Uncertainty exists with respect to the taxonomyof Inia geoffrensis. The genus Inia is con-sidered monospecific with three subspecies: Inia g.humboldtiana found in the Orinoco Basin, I. g.geoffrensis in the Amazon Basin, and I. g. boliviensis

# 2002 EAAM

Page 2: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

found in the rivers of the Bolivian Amazon (Best &Da Silva, 1989a; Da Silva, 1993). The Bolivianbufeo may be isolated from dolphin populations inthe rest of the Amazon Basin due to geographicbarriers, such as the waterfalls that are found in thenorth of Bolivia. These falls extend more than200 km between Guayanamerin and Porto Velho inBrazil. These geographic barriers may have led tothe separation of the I.g. humboldtiania and I. g.geoffrensis on one side, and I.g. boliviensis on theother. In addition, this theory is supported by thefact that I. g. boliviensis differs from the othersubspecies of Inia in some morphological character-istics, such as number of teeth, body size, and skullsize (Anderson, 1997; Emmons & Feer, 1997).Based on these morphological differences and thegeologic barriers formed during the late Pliocene,which supposedly isolated the Bolivian bufeo, sometaxonomists proposed the classification of twospecies, Inia geoffrensis and Inia boliviensis. Thisdistinction of a new species has been disputed basedon the fact that the characteristics used to differen-tiate the species are highly variable and also thesmall sample size. Currently, the systematics of Iniahave been revised with molecular studies, whichsuggest that the genus of Inia found in Bolivia is adistinct species, Inia boliviensis (Banguera et al.,2000). The recognition of a new river dolphinspecies endemic to Bolivia increases the importanceand necessity of protecting this species (Klinowska,1991).

The current conservation status of the bufeo inthe Amazon Basin of Bolivia is poorly known, andbaseline population information is lacking. Thework presented here estimates the abundanceand distribution of the bufeo in a section of theTijamuchi River of the Bolivian Amazon, takinginto account three ecosystem types (i.e., black-waters, mixed white and blackwaters, and oxbowlakes of the river system), four habitat types(i.e., river channel, confluences, bends/curves/andlagoons) and four hydro-climatic seasons (i.e., high,low, rising and falling waters). This work con-tributes to the knowledge of the bufeo in Bolivia,and provides important baseline informationthat can be used to aid in their conservation, thecreation of management plans, and the initiation ofactive protection of the bufeo in this country.Standardized methods were used in the presentstudy, and these can be replicated and used forfuture investigation in Bolivia and elsewhere.

Materials and Methods

Study areaThe study was conducted on the Tijamuchi River,Moxos province, Department of Beni, Bolivia.

The study area is located from where the San-Borja-Trinidad Road crosses the Tijamuchi River towhere the Tijamuchi River empties into theMamoré River (Fig. 1), and is approximately185 linear km. Three oxbow lakes adjoining theTijamuchi were also surveyed.

The area surrounding the Tijamuchi Riverbelongs to the life zone of the subtropical humidforest. Gallery forests border the river, and areinterspersed with savannas, some of which arenatural and others are a result of deforestation fromcattle ranches. The mean temperature in the Llanosde Mojos is 26.5%C and the annual precipitationvaries between 1200 mm and 2400 mm per year.The months of November, December, January, andFebruary are always rainy, with maximum rainfallof 100–500 mm per month. The hydrologic regimeis tightly linked to precipitation, and displays aunimodal curve, with the highest water levels occur-ring between December and April, and the lowestwater levels from June to October (Loubens et al.,1992; Guyot, 1993; Hanagarth, 1993).

The Tijamuchi River is born in the Llanos deMoxos, and is classified as blackwater from itsheadwaters to its middle reaches. Blackwaters arecharacteristically low in nutrients and suspendedsediments, and are stained dark from the tannicacids of decaying vegetation. As the TijamuchiRiver flows, it receives input from small channelsand rivers that originate in whitewaters (white-waters are turbid waters of Andean origins that arehigh in nutrients and suspended sediments). TheTijamuchi River is eventually transformed intomixed black-and-whitewaters, until it empties intothe whitewater Mamoré River. The width of theTijamuchi River is between 35 and 200 m. Theannual water temperature varies between 25 and27%C, dissolved oxygen is 20–23% and depth variesfrom 0 to 5 m (Loubens et al., 1992).

During the rainy season, the lagoons that are lowfor part of the year are converted into enormousbodies of water. Lagoons are places of high fishabundance, as nutrients that were incorporated inthe soil are released into the water column, allowingfor the accelerated development of phyto- andzooplankton (Lara, 1996). The diversity of wildlife(including fish) in the Mamoré River system is high.More than 320 species of fish have been reported inthe region (Lara, 1996), of which humans consumeat least 40 species, and many have commercial valuein the ornamental market. The principal humanactivity in the region is cattle ranching.

FieldworkFieldwork took place between January 1998 andSeptember 1999. Four surveys were conductedduring different hydro-climatic seasons: lowwater (between August and September); rising

313Distribution and abundance of the river dolphin

Page 3: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

water (the end of October and December); highwater (January); and falling water (May andthe beginning of June). In each season anupriver/downriver transect of approx. 3 weeks was

conducted. Approximately 185 km of river weresurveyed during each season.

Because the Tijamuchi River has an averagewidth of 70 m, and an observer stationed mid-river

Figure 1. Map of the study area, Tijamuchi River, Department of Beni, Bolivia.

314 E. Aliaga-Rossel

Page 4: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

can see clearly to both banks, 100% strip-widthtransects were used to survey for dolphins. For allof the transects, a wooden boat with a 15-HPoutboard motor was used, with two observers plusa boat driver. Standing eye-height of observers wasapprox. 2.5 m above the water. The boat travelledat a constant velocity between 7 and 9 km/hr. Thetransects were conducted between 0800 and 1715 h,which gave enough daylight for good visibility. Arest period of 1–2 h was taken during this time,usually around noon. The transect was suspended ifclimatic conditions were unfavourable, since factorssuch as rain or strong winds (>13 km/hr) impairedvisibility. Two transects (i.e., upriver and down-river) were repeated in the same area each season.The boat navigated in the middle of the river,except when hazardous obstacles necessitated atemporary course alteration. The observerssearched for dolphins in front of the boat and alongeither side of the boat. Observations were onlymade from the rear of the boat to confirm numbersand size classes of dolphins previously detected infront of or alongside the boat. This was done toreduce the probability of recounting the samedolphin, as dolphins often dove in front of the boatand resurfaced behind it. The presence of dolphinswas determined visually, or by listening to the noiseof the exhalation and then by visual confirmation.To avoid duplication in the count, the two observ-ers communicated and did not record the samedolphin. Each time one or more dolphins weredetected, it was considered to be a sighting. When asighting was made, the geographic coordinatesand the speed of the boat were recorded using aGPS. River width was estimated using telemetricbinoculars. A wind meter was used to measure thevelocity of the wind.

For every sighting the number of dolphins, groupsize, and age classes were recorded on a datasheet.When observers were unable to determine the exactnumber of individuals present, the lowest reliablecount was registered (McGuire & Winemiller, 1998;Vidal et al., 1997). The term ‘group’ was used torefer to the total number of animals observed,singles or in or an apparent aggregation. This wasdefined as the total number of dolphins observedduring a period of observation within a 25-m radiusof the boat. This is different than the conventionaldefinition of a group, and does not imply anythingabout the social cohesion or interactions of theobserved dolphins (McGuire & Winemiller, 1998).

Age class was categorized in the following way:Adults had a total length (snout to tail) between 1.5to 2.5 m, and were whitish, pink, or grey withpatterns of shades of pink. Adults often had scarson the body (principally on the fins), and pinkcoloration on the melon, Juveniles were 1–1.5 m inlength, had narrow beaks, and were dark grey with

few scars and patches of colour. Calves were lessthan 1 m total length, were uniform gray in colour,had very short beaks with visible vibrissae on thelower jaw, and were almost always accompanied byan adult. The criteria employed to define age class issubjective due to the difficulty in estimating the sizeof a dolphin in the water. All of the lengths werevisually estimated, and the designation of adult orjuvenile does not necessarily indicate sexual ma-turity or immaturity (Hurtado, 1996; McGuire &Winemiller, 1998).

The aquatic habitat was characterized by meas-uring the following environmental parameters every100 m of a transect: pH (with pH paper); transpar-ency of the water (with Secchi disk); temperature ofthe water surface and temperature of the water at1 m depth (with electronic thermometer); and waterdepth (with weighted line).

In each hydro-climatic season, ecosystem andhabitat type were noted. Ecosystem type wasdefined as the type of water, such as blackwater,mixed white- and blackwater, and lagoons. Habitattypes were categorized as main river, confluences(places on the river where a tributary discharges itswater into the Tijamuchi), curves or bends (placeswhere the river bed is wide and has a higher thanaverage current), and lagoons (attached to the river,but with unique physical and chemical charac-teristics). Lagoons were unique in their water char-acteristics (e.g., water colour, pH, transparency)and their geomorphology, and were therefore clas-sified as distinct habitat types and ecosystems types.Zigzag transects were used to survey lagoons, asthese bodies of water were too wide for effective useof a mid-line transect. Lagoon transects did notinclude an estimate of group size.

To learn more about the causes of dolphinmortality, informal interviews were conducted withpeople who live along the river, and with fishermenof the region. When a dead dolphin was encoun-tered, a necropsy was conducted to determine poss-ible cause of death, based on the protocol of Norris(1961).

Data analysisThe associations between group size, season, andecosystem, and age structure and season wereexamined with a chi-square test. ANOVA analysiswas used to examine the effect of habitat (i.e.,lagoons, confluences, bends, or the river proper)on dolphin distribution, and using the followingfactorial model:

$ijk=' "i #j %k ("#)ij ("%)jk (#%)ijK &ijk

where, i=season (low, falling, high, rising);j=ecosystem (blackwaters, mixed white- and black-waters, lagoons); and k=habitat (river channel,

315Distribution and abundance of the river dolphin

Page 5: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

confluences, bends/curves, lagoons). When theANOVA indicated a significant difference, a Tukeytest was used to identify the differences. Because thedata did not have normal distribution they werefirst transformed to ranks. Statistical results wereconsidered significant at the P"0.05 level.

Results

Two-hundred and twelve h were spent surveyingdolphins. This was only the time spent searching forand following dolphins, and does not take intoaccount the time spent on photo-identification orreconnaissance of the study area.

AbundanceIn total, there were 2040 sightings of bufeos (I.geoffrensis) in the Tijamuchi River, includingresightings from the lagoons. There was a densityindex of 1.12 individual dolphins per km transect.The total distance surveyed was approximately 185linear river km. The average total estimate ofdolphins for the four seasons combined was 208(SE$28) individuals.

The population estimated varied among seasons,as well as within each season, depending if thedolphins were seen in the main river or in thelagoons (Fig. 2). Dolphins were observed mostoften during the seasons of low and falling water.The maximum number of bufeos corresponds to thelow water season, with 295 individuals (30.4% ofthe total), and the minimum number was obtainedduring high water, with 209 (21.5% of total).Sampling effort did not vary by season. Differences

in dolphin abundance among all seasons were stat-istically significant (F=80.55, df=3,4, P=0.0005).Pair-wise differences between seasons were alsosignificant, with the exception of rising and highwaters (P=0.66).

In the rivers, dolphins were most likely to beobserved during falling and low waters. In thelagoons, dolphins were most frequently seen duringrising and high waters. The number of dolphins inthe lagoons declined during low water, until theywere often entirely absent from the lagoons, as thechannels that connected some lagoons to the riverswere reduced to depths of 50 cm, or became totallylow, and the bufeos left the lagoons for the mainriver.

Group SizeIn total, there were 920 groups of bufeos seen(including lonely animals). The predominate groupsize was of solitary individuals, which were 41%observations. Next were groups of two individuals,which were 32% observed (3% of all groups weremothers and calves), followed by groups of threedolphins (15%). The largest group observed was 19dolphins, and was encountered during fallingwater at the confluence of the river and a smalllagoon channel where bufeos feeding onnumerous, medium-sized (100–300 mm) fish thatwere travelling from the lagoon to the river.

The frequency of group size varied significantlyby season. The chi-square analysis indicated thatthere was a significant association (Fig. 3) betweengroup size and season with 31% ((2=73.67; df=15;P=0.00). Solitary individuals were seen most

Figure 2. Seasonal averages of dolphins by season.

316 E. Aliaga-Rossel

Page 6: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

frequently in all seasons, with the exception of lowwater, during which time pairs were seen more oftenthan other group sizes. Maximum group size was 14individuals during low and rising waters, 13 indi-viduals during high water, and 19 individuals dur-ing falling water. The greatest concentrations ofbufeos were observed in the lagoons, where morethan 30 individuals were encountered in a singlelagoon, but these were not considered in the calcu-lation of group size and were seen instead astemporary groups due to the restricted area (exceptfor mother/calf pairs).

Group size was associated with ecosystem (i.e.,whitewater, blackwater, and lagoons). Chi-squareanalysis indicated a significant differences in groupsize between white water and mixed water((2=12.26; df=5; P=0.03). Black waters and mixedwaters were predominated by solitary animals.Groups of five or more individuals were seen inlagoons, and smaller groups were not seen in theseecosystems.

Age structureOf the total observations, the majority of dolphinswere adults (93.9%), then calves (3.27%), and juve-niles (2.81%). There were individuals of unknownage, due to ambiguity of size (especially betweenjuveniles and adults), and these were not analyzed.

The frequency with which age classes (Fig. 4)were observed varied according to season((2=19.74; df=6; P=0.003). Calves were seen leastfrequently during high water (1.6% of the total

seasonal population), and most frequentlyduring falling water (5.37% of the total seasonalpopulation).

Ecosystem and habitat preferenceWith respect to ecosystems, the Tukey test indicatedthat differences in dolphin abundance betweenblackwaters and mixed waters were not significant(Table 1). However, it did appear that dolphinsexhibited a statistically significant preference forlagoons. Dolphins were encountered indiscrimi-nately between blackwaters and whitewaters, whilethere was a greater tendency to encounter them inlagoons (except when the lagoons are dry).

The variation in mean dolphin abundancebetween the different habitats suggests that there isa preference for lagoons, then the confluences, thencurves/bends, and finally the main channel of theriver. Significant differences in dolphin abundanceamong all habitats, as indicated by the Tukey test,existed for all comparisons, except for the differencebetween rivers and curve/bends. (Table 2).

Dolphin preferences according to the factorialmodelThe results of the analysis using a factorial model toexamine the association of dolphin abundance withseason, ecosystem, and habitat (Table 3) indicatedthat season and habitat were significant factors,while ecosystem was not. The interaction of seasonwith habitat was significant, as was the interactionof ecosystem and habitat, while the interaction of

Figure 3. Group size by season.

317Distribution and abundance of the river dolphin

Page 7: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

season and ecosystem was not. The three-wayinteraction of season, ecosystem, and habitat wasnot statistically significant.

Dolphin preferences according to thephysical-chemical characteristics of theTijamuchi RiverThe physical-chemical characteristics of theTijamuchi River were measured in each season(Table 4). The distance of the dolphin to the nearestshore was not taken into account, owing to the factthat the width of the river in some places was lessthan 30 m, and a group was defined as thoseindividuals that were within a 25-m radius. Withrespect to depth of the river, 94% all of the sightingswithin the Tijamuchi were between 1 and 18 mdepth, with an average depth of 7 m. There waslittle variation in pH or water temperature and it

does not appear that these factors or transparencyaffect the distribution of the dolphins (none of thePearson correlation coefficients were statisticallysignificant at the P"0.05 level). Water depth andwidth are determined by season and habitat, whichdo affect dolphin distribution.

Causes of mortalityTwo dead Inia were encountered. The first wasencountered during the low season at the conflu-ence of the Tijamuchi River and the Mamoré River.The second dead dolphin was encountered duringfalling water, near a bend in the Tijamuchi River.Both events were photographed.

The first death may have been due to sexualaggression. One day before the carcass wasdiscovered, reproductive activity among variousindividuals (approximately 12) was observed along

Figure 4. Seasonal age distribution.

Table 1. Tukey test indicates the differences in dolphins abundance between blackwaters andmixedwaters.

Ecosystem(I)

Ecosystem(J)

Meandifference

(I–J)Std.

error. Sig.

95% Confidence interval

Lowerbound

Upperbound

Whitewaters Mixedwaters 9.80E+01 0.94 "0.59 0.79Lagoons "10.80* 0.90 0.00 "129.24 "86.92

Mixedwaters Whitewaters "9.80E"02 0.29 0.94 "0.79 0.59Lagoons "10.90* 0.92 0.00 "130.78 "87.34

Lagoons Whitewaters 10.81* 0.90 0.00 86.92 129.24Mixed-black 10.91* 0.93 0.00 87.34 130.78

*The mean difference is significant at the 0.05 level.

318 E. Aliaga-Rossel

Page 8: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

a sandy beach at the confluence. There was a greatdeal of splashing, chasing, playing, and interactivecontact between the dolphins. This activity con-tinued throughout the night, accompanied byabnormally loud exhalations and labored breathingby at least one dolphin. The following morning thedead body of a male Inia, 2.25 m in total length,was encountered on the beach. Its pectoral andcaudal fins displayed numerous deep tooth scarsfrom other Inia (determined by scar depth,shape, and spacing), bleeding from the anal slit(from internal hemorrhaging), and heavy bruisesand swelling around the blowhole, the injuries tothe blow hole may have resulted in the laboredbreathing of the previous night.

The second dead dolphin was a 2.15-m long male,and possibly struck by an outboard motor, as the

left side of the beak was heavily cut and the lowerleft jaw was broken. This animal was discoveredin the river, snagged in debris. (Note: the debrisdid not cause the death of the dolphin, as it hadbeen in the water long before the dolphin died. Itsimply impeded the downstream movement of thecarcass).

Interviews with fishermen indicated that theseason of greatest fish abundance is between Mayand September (period of falling and low water).During this time, the fishermen place their nets atthe mouths of the lagoons and affluents and fishleaving these areas for the main river are caught.In these months, fishermen frequently encounterdolphins trapped in their nets. They are more likelyto trap the younger dolphins, as these animalsare the least experienced with nets. Generally,

Table 2. Tukey test indicates the differences in dolphin abundance among all habitats.

Habitat(I)

Habitat(J)

Meandifference

(I–J)Std.error Sig.

95% Confidence interval

Lower bound Upper bound

River Confluences "2.09* 0.411 0.000 "31.50 "10.36Curves/bends "0.49 0.380 0.565 "14.7 0.4835Lagoons "11.12* 0.874 0.000 "133.6 "88.72

Confluences River 2.09* 0.411 0.000 10.36 31.50Curves/bends 1.60* 0.522 0.012 0.259 29.41Lagoons "9.02* 0.945 0.000 "114.52 "65.98

Curves/bends River 0.49 0.380 0.565 "0.48 14.7Confluences "1.60* 0.522 0.012 "29.41 "0.259Lagoons "10.62* 0.931 0.000 "130.18 "82.32

Lagoons River 11.12* 0.874 0.000 88.72 133.63Confluences 9.02* 0.945 0.000 65.98 114.51Curves/bends 10.62* 0.931 0.000 82.32 130.18

*The mean difference is significant at the 0.05 level.

Table 3. Factorial model test to examine the association of dolphin abundance with season,ecosystem and habitat.

SourceType III

Sum of squares dfMeanSquare F Sig.

Corrected model 1698.43a 27 62.91 13.66 0.00Intercept 2065.03 1 2065.03 448.51 0.00Season 150.14 3 50.05 10.87 0.00Ecosystem 0.80 1 0.804 0.175 0.68Habitat 159.99 2 79.99 17.37 0.00Season* ecosystem 19.23 3 6.41 1.39 0.25Season* habitat 69.29 6 11.55 2.51 0.02Ecosystem* habitat 49.52 2 24.76 5.38 0.005Season* ecosystem* habitat 28.56 6 4.76 1.03 0.40Error 1643.71 357 4.60Corrected total 3342.14 384

*R2=0.508 (adjusted R2=0.47).

319Distribution and abundance of the river dolphin

Page 9: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

fishermen release the dolphins, although in somecases they leave the dolphins to die or kill them withmachetes to prevent the destruction of their nets.The fishermen may use the dolphin meat as bait toattract fish, or use the oil to cure maladies of thelungs. At times, the dolphins are fired on with gunsfor amusement, or by sports hunters practising theiraim.

Local people report that during one year in thestudy area, six dolphins (three calves, two juveniles,and one adult) died in fishing nets. Two calves, onejuvenile, and the adult were captured in the monthsof May to September, which corresponds to theseasons of falling and low water. The nets werelocated across the mouth of a lagoon. One of thefishermen indicated that the calf died because therewas not time to free it, but that an adult dolphin(believed to be the mother) was seen hoveringaround it and was heard loudly exhaling for about20 m after the calf died. Eventually, the adult leftthe scene.

Photo identificationTwenty-seven photographs were of sufficientquality to use for photo identification of individuals(Trujillo, 1994). From these photographs, two indi-viduals were identified. One animal was resightedonce, The day after it was first identified. The otheranimal was resighted four times, with 239-day rangebetween the initial and final sighting. Photo identi-fication effort was limited due to time and economicconsiderations.

Discussion

The density of dolphins in the Tijamuchi River ishigh (1.12 dolphins/km), especially compared withthe rivers of Colombia and Venezuela, where thereported densities are below 1 (Trujillo, 1992;McGuire & Winemiller, 1998). The greater abun-dance of dolphins in the Tijamuchi River could bedue to many factors, including the fact that the

region has very little motorized boat traffic and orhuman fishing activity, since human activity in thestudy area is primarily limited to cattle ranching.

The average number of dolphins in the TijamuchiRiver varied according to season, although thesedifferences were not statistically significant. The lowseason had the majority of sightings. During thelow season, the river width decreased to an averagewidth of 50 m and the reduced area could facilitatethe increase in dolphin sightings. Researchers in theBrazilian Amazon reported more dolphins duringhigh waters than any other season (Best & Da Silva,1989a). Although dolphin abundance peaks atopposite water levels in the Bolivian and Brazilianstudy areas, the calendar months (May to July) arethe same. The number of dolphins in the study areadecreased during the high water season, perhapsdue to the dispersion of individuals in the inundatedforest, or migrations of the dolphins to the MamoréRiver and its numerous lagoons. This decrease inbufeo abundance during high water was alsoobserved by McGuire & Winemiller (1998) in theVenezuelan Orinoco. As the waters began to rise inthe Tijamuchi River, heavy rainstorms occurredwhich greatly increased the river current andwashed out many logs and fallen trees. This couldhave caused the dolphins to temporarily abandonthe Tijamuchi River and move to the MamoréRiver, where the deluge is attenuated by the size ofthe river and the dolphins could find refuge fromthe current in wide river bends or adjacent lagoons.During high and rising waters, dolphin aerialactivity is very low, and thus it is more difficult toencounter them.

GroupsThroughout the year, dolphins were generallyencountered as solitary animals or in pairs. Duringthe high water seasons the dolphins were dispersedand were most frequently seen as small aggrega-tions of 1 or 2 individuals. During falling and lowwater seasons, larger groups were encountered,

Table 4. Physical–chemical characteristics of the Tijamuchi River in the different hydroclimatic seasons.

T%C(sup)

T%(depth)

Depth(m)

Transp(cm)

Width(m) pH

Low 28.54 27.045 4.4 11.125 54 6.8Rising 28.45 28.3 7.5 15.351 58.88 6.4High 28 27.8 11.2 24.256 84.62 6.5Falling 23 23 5.25 14.46 73 6.5Average 27 26.5 7.08 16.298 67.625 6.55

Where: T%C (sup): Temperature at water’s surface. $15 cmT%C (depth): Temperature at depth. $1.5 m.

320 E. Aliaga-Rossel

Page 10: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

principally in the mouths of the small tributaries orthe mouths of the lagoons. This coincides with thefindings of Best & Da Silva (1989a) and Magnussonet al. (1980). The largest group was of 19 individ-uals, and was encountered in a confluence, near themouth of the Mamoré River. Large dolphin aggre-gations were found in areas that presented optimalconditions for fishing or swimming, such as conflu-ences with their greater abundance of fish, or themore tranquil waters of wide river bends. Thegroups were more numerous in the confluence areasthan in the main river channel. Confluences have agreater density of fish than do the river channels(Lara, 1996).

Kendall (1994) reported that in Colombia riverdolphins that appeared to be solitary actuallybelonged to a dispersed group, and these animalsmay regroup for fishing or socializing and thendisperse again. When large aggregations in theTijamuchi River were observed the dolphins weregenerally in subgroups of one to three dolphins.

A statistical correlation between the type of water(whitewater or blackwater) and the size of thegroup was not detected, and it appears that thedolphins did not exhibit a preference for water type.

In the seasons of low and falling water, there wasa higher percentage of dolphins in pairs. Thiscoincided with the period of reproduction and thecalving period, indicated by the increased sightingsof mother/calf pairs.

Age structureMore that 90% of total sightings were adults, whilethe rest were of calves and juveniles. There wereslight variations in the age-class rations accordingto season. The proportion of calves increased inthe months between May and September. This isthe same period reported by Da Silva (1993) as thecalving season of the bufeo in central Brazil,although they attribute the calving season to themaximum height of the water level of the Brazilianrivers. In the Tijamuchi River, these same monthscorrespond to falling and low waters, when fish areconcentrated in the channels of the rivers andtributaries. According to Louzanne (personal com-munication), this is when the commercial fisherieshave their maximum catch. The greater abundanceand ease of capture of fish in these months mayrepresent favourable conditions for the periodof mating and pregnancy of the river dolphins.Gestation is believed to be as long as 10 to11 months (Kendall, 1994), which would indicatethat the mating season coincides with the calvingseason of the following year. In the TijamuchiRiver, this is corroborated by an increase in obser-vations of both calves and mating activity betweenadults during these seasons. This was also observedin Venezuela by McGuire & Winemiller (1998).

Sadly, the increase in the proportion of calvesduring falling and low water periods also coincideswith the greatest number of calf deaths in fishingnets.

In the seasons of rising and high water, calvesand juveniles were observed. It could not be deter-mined if the same young dolphins had also beenobserved in the previous seasons, or if they wererecent births. In the encounters with bufeo calves,the mother/calf bond was always observed. Ingeneral, these pairs were observed in areas, such aswide bend and curves, that were protected from thecurrent. The proportion of pairs of mothers andcalves was lower in the lagoons than in the otherbodies of water. Kendall (1994) reported the exist-ence of ‘nurseries’, where groups of calves andyoung dolphins were cared for by one or two adults,while the other adults travelled to the mouths of therivers. These nurseries were never observed in theTijamuchi River or in the vicinity of its confluencewith the Mamoré River.

Juveniles were encountered more often in themeanders and lagoons, especially in periods offalling and rising water. These places have deeperand slower water and are therefore easier places forthe juveniles to swim and fish.

DistributionThe distribution of the bufeo in the TijamuchiRiver includes all habitats, but dolphins were mostlikely to be found in lagoons, confluences, andmouths of lagoons; all places of greater preyavailability. The same was observed by Trujillo(personal communication) for the bufeos of theupper Amazon River and the Arauca River in theColombian Orinoco.

Dolphins were always observed in the conflu-ences of the rivers, where fish are concentrated asthey leave the channels for the principal rivers andalso because masses of water meet at confluencesand create whirlpools and eddies that cause the fishto become disoriented and thus more easily trappedby the dolphins (Best & Da Silva, 1989b). It isinteresting to note that few dolphins were seen inthe middle reach of study area, which is a stretch ofriver without confluences for many kilometres.

Dolphins appeared to prefer lagoons to the mainriver channel. This may be due to the greaterbiomass and diversity of fish found in the lagoons.Fishery yields are maximum in the lagoons duringlow water; when the average catch of fish is between16.5 to 38.6 kg/100 m2/24 h (Lara, 1996).

There was little variation in water temperaturealong the river itself, although the range of tem-peratures among the water bodies was broad(between 24 and 32%C). It was not surprising thatwater temperature did not appear to affect dolphindistribution. As mammals, dolphins have internal

321Distribution and abundance of the river dolphin

Page 11: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

homeostasis which keeps their general bodytemperature independent of the environment inwhich they find themselves. The pH and the trans-parency of the water did not appear to affect thedistribution and abundance of the dolphins.Although it was not directly examined in this study,it would seem that the primary productivity of thewater, which is low in zones with few fish andhighest in the zones where the fish are abun-dant, might have an important effect on dolphinabundance.

The dolphins were seen in a range of river widths,for example from a channel 25-m wide to a lagoonmore than 200 m. The largest dolphin aggregationsoccurred at the widest confluences.

MortalityHumans are the principal ‘predators’ of riverdolphins and of many of the species of the studyarea. Human activities in the area directly andindirectly affect river dolphin populations. Thepeople of the zone use river dolphin oil as atraditional treatment for respiratory problems (suchas tuberculosis) or weakness of the lungs. There isno evidence that people of the Bolivian Amazon killriver dolphins as a source of protein.

According to interviews with fishermen, themajor cause of dolphin mortality is accidental deathcaused by the use of fishing nets. The same wasreported for Colombia (Trujillo, personal com-munication). Although fishermen usually releasethe dolphins that are trapped in their nets, in somecases the same fishermen will leave the dolphins todie in the net, or will kill them to prevent thedestruction of their nets. Fishermen use the meat ofdolphins they have killed as bait to attract fish. Thefishermen of the Tijamuchi River say that they donot consider themselves to be in competition withthe dolphins. In Peru’s Pacaya-Samiria Reserve,fishermen do consider the dolphins as competition,and some fishermen poison dolphins to protect theirnets and their catch (McGuire & Zúniga 1997,Würsig & Reeves 1998). The death rate in nets isgreatest for calves and juveniles, which may bedue to their lack of experience with nets and totheir curiosity. The seasons of falling and lowwaters coincide with an increase in human fishingactivity in the zone, as well as the bufeo calvingseason.

The increase in boat traffic in rivers, such as thoseinvestigated by Pilleri & Gihr (1977), is verydangerous for the dolphins. This was demonstratedby the second dead Inia that was encountered, inwhich the lower jaw was destroyed by the propellerof a boat. This danger was observed to increaseduring low water, when the space available to thedolphins was limited to only a few centimetres tospare in a channel that is used by fishing boats.

Published studies regarding aggression amongbufeos do not exist and thus, it is not known howusual the aggressive behaviour is in association withthe death of the first necropsied Inia.

This study provides important baseline dataregarding river dolphins in Bolivia and underscoresthe importance of continuing and expandingthe study to learn more about the ecological andbehavioural aspects of this charismatic species.

Acknowledgments

Thanks to Healy Hamilton and Mario Baudoinwho encouraged and assisted me in the undertakingof this study. And thanks to the Institute ofEcology of Bolivia for all of their help in this study,to LIDEMA, Andrew Taber and WCS, FernandoTrujillo and the Fundación Omacha, to MauricioPenarrieta, Marcelo Lara, Nicolás Molina, to myboat driver Ramiro Cuellar and his family. To myreviewers and all of my family. Tamara McGuireassisted with initial field work, statistics, manuscriptcomments, and translations. Elizabeth Zuniga fortheir comments on the manuscript.

Literature Cited

Anderson, S. (1997) Mammals of Bolivia, Taxonomyand Distribution. Bulletin of the American Museum ofNatural History. Number 231, 652.

Banguera-Hinestrozoa, E., Cardenas-Henao, H.,Hamilton, H., Ruiz-Garcia, M., Vasquez, R. &Garcia-Vallejo, Y. F. (2000) Variabilidad genética de laregión d-loop del cyt-b (600) y D-loop 570(bp)mitocondrial en poblaciones del delfín del río de Iniassp (Cetacea:Inidae) en las cuencas del los ríos Orinoco,Amazonas y Beni-Mamoré. Pg. 7. Abstracts, 9threunión de trabajo de especialistas en mamíferosacuáticos de América del Sur. Oct/Nov. 2000, BuenosAires, Argentina.

Best, R. C. & Da Silva, V. M. F. (1989a) Amazon RiverDolphin Boto Inia Geoffrensis (De Blainville. 1817). In:S. H. Y. Ridgway & R. Harrison (eds.) Handbook OfMarine Mammals. Vol. 4, River Dolphins And LargerToothed Whales, pp. 1–23. Academic Press. London.

Best, R. C. & Da Silva, V. M. F. (1989b) Biology, Statusand Conservation of Inia geoffrensis in the Amazon &Orinoco River Basins. In: W. F. Perrin, R. L. Brownell,Jr., Z. Kaiya & Y. L. Jiankang (eds.) Biology andConservation of the River Dolphins, pp. 24–34. IUCN/Species Survival Commission (SSC) Occasional PapersNo. 3, Gland, Switzerland.

Da Silva, V. M. F. (1993) Inia geoffrensis. MammalianSpecies. 426, 1–8.

Emmons, L., Feer, Y. F. (1997) Neotropical RainforestMammals, A Field Guide, 2nd edn. The University ofChicago Press, U.S.A. 170–171.

Guyot, J. L. (1993) Hydrogéochimie des Fleuves del’amazonie Bolivinne. ORSTOM, Paris.

Hanagarth, W. (1993) Acerca De La Geoecología de lasSábanas del Beni en el Noroeste Bolivia. Instituto deEcología, La Paz, Bolivia.

322 E. Aliaga-Rossel

Page 12: Distribution and abundance of the river dolphin ( Inia ... · boto, abundance, distribution, mortality, photo-identification, Tijamuchi river, Bolivia. Introduction There are few

Hurtado, L. A. (1996) Distribución, uso de hábitat,movimientos y organización social del bufeo coloradoInia geoffrensis (cetacea: iniidae) en el Alto RíoAmazonas. Instituto tecnológico y estudios SuperioresDe Monterrey. Campus Guaymas. Tesis De Maestría.México.

Kendall, S. (1994) Los Delfines de la Amazonia y laOrinoquia. Fundación Omacha, Bogotá- Colombia.

Klinowska, M. (1991) Dolphins, Porpoises and Whales ofthe World. The IUCN Red Data Book. Published byIUCN, Gland, Suize, Cambridge, UK. Pp. 52–60.

Lara, M. (1996) Observación del estado gonadal deHoplosternum Thoracatum y de Hoplosternum. Littorale(Callyctyidae). En el Área de Trinidad—Beni—Bolivia,de Agosto A Octubre. Universidad Del Beni ‘MariscalJosé Ballivian’. Tesis De Licenciatura.

Loubens, G., Lauzanne, L. & Le Guennen, B. (1992) LesMilleux Aquatiques de la Region de Trinidad (Béni,Bolivie). Rev Hydrobiol. Trop. 25, 3:21.

Magnusson, W., Best, R. C. &. Da Silva, V. M. F. (1980)Numbers and behaviour of Amazonian dolphins, Iniageoffrensis and Sotalia fluviatilis, in the río Solimoes,Brasil. Aquatic Mammals 8, 27–32.

McGuire, T. L & Winemiller, K. O. (1998) Ocurrencepatterns, habitat associations, and potencial prey of theriver dolphin, Inia geoffrensis, in the Cianuco river,Venezuela. Biotropica 30, 625–638.

McGuire, T. L. & Zúniga, E. (1997) River dolphin andfishery interactions in South America. 1997. Abstract,annual meeting of the society for conservation biology.Fort Collins Colorado.

Norris, K. (1961) Standardized methods for measuringand recording data on the smaller cetaceans. ByCommittee on Marine Mammals—American Society ofMammalogists. Journal of Mammalogy 42, 471–476.

Pilleri, G. (1969) On the behaviour of the AmazonDolphin, Inia geoffrensis in Beni Bolivia. Revue Suissede Zoologie 76, 57–74.

Pilleri, G. & Gihr, M. (1977) Observations on the Bolivian(Inia geoffrensis d’Orbigny, 1834) and the amazonianbufeo (Inia geoffrensis de Blainville, 1817) with descrip-tion of a new subspecies (Inia geoffrensis humboldtiana).Invest. Cetacea 8, 11–76.

Trujillo, F. (1992) Estimacion poblacional de las especiesdulceacuícolas de delfines Inia geoffrensis y Sotaliafluviatilis en el sistema lacustre de Tarapoto y ElCorreo, Amazonia Colombiana. Special report. Vol.49. Centro de Investigaciones Científicas, UniversidadJorge Tadeo Lozano, Bogotá-Colombia 199 pp.

Vidal, O., Barlow, J., Hurtado, L., Torre, J., Cendon, P. &Ojeda, Z. (1997) Distribution and abundance of theAmazon River Dolphin (Inia geoffrensis) and the tucuxi(Sotalia fluviatilis) in the Upper Amazon River. MarineMammal Science 13 (3), 427–445

Würsig, B. & Reeves, R. (1998) Dolphin ecology in thePeruvian Amazon. Whale. Oceanic society expedition.

Yanes, M. (1999). Etología, ecología y conservación deldelfín Inia geoffrensis en los ríos Itenez y Paragua delParque Nacional Noel Kempff Mercado. UniversidadMayor de San Andres. Facultad de Ciencias Puras yNaturales. Tesis de Maestría.

323Distribution and abundance of the river dolphin