[email protected] The Palms NDONIAINA IJORO … · Madagascar, and yet much of the region is inaccessible and remains unexplored. Here, we report the findings of an expedition in November

It is well established that palm diversity inMadagascar peaks in the humid forests of theisland’s north-east (Dransfield & Beentje 1995,Rakotoarinivo et al. 2013, 2014). Within thenortheast, the Masoala Peninsula is arguablythe most important palm hotspot. Palm speciesrichness is exceptionally high there with as

many as 60 species having been recorded(Rakotoarinivo et al. 2009). While other areas,such as the adjacent Makira Protected Area,may rival Masoala in species diversity(Rakotoarinivo et al. 2009), Masoala isparticularly diverse at the genus level – all the“big game” rarities of the Madagascar palm

PALMS Baker et al.:Masoala Palms Vol. 60(4) 2016

169

The Palmsof theMasoalaPeninsula

WILLIAM J. BAKER ANDWOLF L. EISERHARDTRoyal Botanic Gardens, Kew,Richmond, Surrey, TW9 3AB,[email protected]

MIJORO RAKOTOARINIVO AND

ANDONIAINA Z. ANDRIAMANANTENADépartement de Biologie et EcologieVégétales, Faculté des Sciences,Université d’Antananarivo, BP 906,Antananarivo 101, Madagascar

ROMER N. RABARIJAONAKew Madagascar ConservationCentre, Lot II J 131 B AmbodivoanjoIvandry, Antananarivo 101,Madagascar

SOLO H.J.V. RAPANARIVOParc Botanique et Zoologique deTsimbazaza, Rue Kasanga Fernand,Antananarivo 101, Madagascar

The Masoala Peninsula is arguably the most celebrated destination for palms in

Madagascar, and yet much of the region is inaccessible and remains unexplored.

Here, we report the findings of an expedition in November 2015, during which

we visited both the west side of the peninsula and the scarcely known east,

encountering extraordinary palm diversity and several species new to science.

PALMS 60(4): 169–193

flora, such as Lemurophoenix, Satranala andVoanioala, are found there.

The richness of the Masoala Peninsula mayarise from a combination of factors (Rakoto-arinivo et al. 2013). Firstly, rainfall in this areais extremely high and has been so even duringthe last glacial maximum (>20,000 years ago),when climates globally were cooler and drier.High rainfall is an established driver of palmspecies richness worldwide (Kissling et al.2012). Secondly, despite its modest size (80km long, 60 km wide at widest point), thepeninsula is topographically varied, with highmountains to more than 1100 m rising sharplyon the west side in contrast to the more gentlelandscapes of the east side. Masoala is ratherhomogeneous geologically, consisting largely

of granitic basement rocks. However, small-scale variations in substrate that do not appearon geological maps may contribute to thepeninsula’s palm richness.

A significant proportion of Madagascar’sremaining humid forest is found in theMasoala Peninsula and around 240,000 ha ofthe peninsula is now protected within theMasoala National Park. Important areas of thepeninsula remain threatened, however,especially on the east side where deforestationis greatest and significant areas of forest falloutside the park boundary. Even within thepark, risks to biodiversity remain high due toillegal logging, for example for ebony androsewood, and hunting for bushmeat. In thecase of palms, the larger species are targeted for


170

1. Map of the study area. Centre: location of the Masoala peninsula in Madagascar. Upper left: overview ofthe Masoala peninsula with fieldwork areas indicated in red, and GPS track logs indicated in light blue. Tracklogs and collections (triangles) in Iketra (upper right), Tampolo (lower left) and Antalavia (lower right) areas.

palm heart extraction, a common practicethroughout Madagascar, which is of particularconcern for critically endangered species withlow population sizes, such as Voanioala.

The goal of our expedition in November 2015was to explore for the widest diversity ofMasoala palms, collecting herbariumspecimens of all species encountered and leaftissue for DNA-based research on the evolutionof the Madagascar palm flora. As most previouspalm research effort has focused on sites inWest Masoala, we gave most of our attentionto inland sites in East Masoala, reached via theAnaovandrano River valley (Fig. 1). This areawas explored during a short reconnaissanceexpedition in 1996 by Dr. John Dransfield andcolleagues (Dransfield 1996), resulting in thediscovery of remarkable novelties, such asDypsis metallica, D. reflexa and D.vonitrandambo (Rakotoarinivo & Dransfield2010), and sightings of several more that couldnot be collected at the time. Our expeditionconcluded with a short stay in better knownsites in West Masoala in the vicinity ofTampolo and Antalavia.

Voyage to Masoala

On 11 November, the expedition team arrivedin Maroantsetra, the gateway to the MasoalaPeninsula. Our departure from Antananarivohad been fraught with anxiety as AirMadagascar had threatened not to carry all ofour copious baggage until the very last minute.But the sedate pace of life in Maroantsetra isirresistible, and having set up base (with allour luggage) at the sleepy Coco Beach Hotel,we soon regained our composure. Over thefollowing day and a half, we made finalarrangements for our fieldwork, visiting theoffices of Madagascar National Parks (to hireour MNP guide, Donné, and cook, Angèle),chartering a boat and sourcing provisions inthe market and shops around town.

At 8 a.m. on 13 November, our boat departedfor the 120 km journey around to the east sideof the peninsula. The weather was perfect,giving us glorious views of the island of NosyMangabe across the oily calm waters at thehead of the Bay of Antongil. We sped out ofthe estuary into a sublime scene, with the highmountains of Masoala rearing up to one side,


171

2 (left). Dypsis acaulis. 3 (right). Dypsis pachyramea. Photos: W.J. Baker.

fish leaping from the water with fishermen inpirogues in hot pursuit and the bay openingup in front of us. The experience would havebeen very different, no doubt, had the weatherbeen less kind to us or our boat less sturdy.

After rounding the tip of the peninsula at CapMasoala into the choppier waters of the IndianOcean, we stopped briefly on a spectacularwhite sand beach at Ambodilaitry to seekadvice on local guides. We then continuednorthwards along the east coast for the final20 km of the voyage before turning inlandinto the mouth of the Anaovandrano River.Near the coast, the river runs through patchesof littoral forest and large clumps of Dypsislutescens sway precariously over the banks.Finally, at around 2 p.m., we reached ourdestination, Sahamalaza, a village that flanksan important track running north-south alongthe east of Masoala that is accessible at leastto motorbikes. It was far too late in the day tostart trekking into the interior, and so we madecamp in the village and went on a search forlocal knowledge of the route into the interior.By nightfall, we had been joined by two guides,Emmanuel and Donat, and a team of excellentpirogue men who offered to punt our luggageup river by boat, which relieved us of theinevitable challenges that come withmanaging a large group of porters. So far, so

good, we thought. Our careful planning haddelivered us safely to a place that, for the pasttwo years, had existed only in our imagi-nations. But the following day, the real workwould begin in earnest.

Hike from Sahamalaza to Iketra

We broke camp early next morning and, fueledwith a typical breakfast of rice and beans,hauled our luggage to the river where ourpirogues were loaded to the gunnels. At 8 a.m.,the pirogue men cast off, while the rest of uswere ferried across the river to begin our hikeinto the forest on foot to an area known asIketra, some 14 km inland to the east whereDransfield and colleagues had camped in 1996.We followed a path that lay to the north of theAnaovandrano. The first two hours of the hikepassed through a landscape of fields anddegraded vegetation, crossing many streamsand swampy hollows. The sky was perfectlyblue, but this turned out to be a mixed blessingas the heat bore down on us. Finally the pathentered fragmented forest, bringing welcomeshade and, almost immediately, interestingpalms.

One of the first palms we encountered, Dypsisacaulis (Fig. 2), was seen at no other timeduring the expedition. Dransfield too recordedthis species in the same area in 1996, the first


172

4. View of East Masoala from the trail to Iketra. Photo: W.J. Baker.

record since the type collection by Perrier dela Bâthie in 1912 (Dransfield 1997). This dwarf,stemless palm occurred in small colonies,sometimes mixed with another understorypalmlet, Dypsis pachyramea (Fig. 3), in a wetvalley bottom and on stream banks. Its bifidleaves are highly distinctive being leathery andalmost eophyll-like, with dense gray-whiteindumentum on the undersurface. Incon-spicuous, spicate inflorescences emerge frombetween the leaves. Other more widespreadpalms encountered along the path includedDypsis dransfieldii, D. forficifolia, D. fibrosa andD. pinnatifrons.

The forest now became contiguous, thoughthe path connected several open areas clearedfor farming, often with a small settlement ineach site accommodating a small, extendedfamily (Fig. 4). Near lunchtime, we stopped atone such settlement, a cluster of small houseswith a dramatic forest backdrop, where wegorged on coconuts and jackfruit. The pathnow threaded along the forest edge, making iteasy to spot the bigger tree palms, such asDypsis lastelliana, Orania longisquama, O.trispatha, Ravenea julietiae and R. dransfieldii.We saw also one scruffy individual of anundescribed Dypsis with the local name sira,discussed in more detail below. As we nearedIketra, we were excited to see a hut part-thatched with fan leaves, and shortly aftercame upon the source, Satranala decussilvae,the first of the “big game” palms of the trip,albeit an underwhelming specimen of it.

Finally, after fording the chest-deep waters ofthe Anaovandrano River, we reached Iketraaround 4 p.m., where we set up camp on theedge of the village football pitch, the site ofspirited, mixed soccer matches every evening.Iketra consists of an area of cleared farmlandflanking a 2.5 km section of the river, muchof it uncultivated and only lightly grazed byzebu cattle. A large area of forest had beenfreshly cleared across the river, opposite ourcamp. Iketra village itself comprises just a fewclusters of tiny houses sparsely inhabited by asmall population subsisting on slash-and-burnagriculture. Life for the people of Iketra isundoubtedly isolated and tough.

The core palm flora of Iketra

Despite new clearances, spectacular primaryforest surrounds Iketra. Having walked fromthe east side of Iketra, we decided to use thesix fieldwork days that followed to explore inthe remaining directions to the north, south

and west. The landscape throughout the areais characterized by low rolling hills and ridgesunder 100 m elevation, with many smallstream valleys draining towards theAnaovandrano River. The palm flora was rich,with at least 40 species observed in total, butlocal species richness was patchy and tendedto be most concentrated in wetter valley sidesand floors.

A core palm flora comprising eleven species,mostly understory palms, was encountered inalmost every site around Iketra. The smallestof these, Dypsis pachyramea, is the commonestunderstory palmlet in Masoala. This specieswas extremely abundant especially in valleybottoms, growing gregariously in largenumbers. The remaining species of understorypalm were taller in stature. On slopes, D.fasciculata and D. forficifolia were frequentlyobserved, the former displaying a bewilderingdiversity of leaf forms. Less common in thesame habitat but still widespread were D.confusa and D. pinnatifrons. Perhaps the mostspectacular of the taller undergrowth palmswas Dypsis procera, which formed large clumpsin and around valley bottoms. Although formswith divided leaves were present, the


173

5. Dypsis vonitrandambo. Photo: W.J. Baker.


174

6. Orania longisquama. Photo: W.J. Baker.


175

7. Orania trispatha. Photo: W.J. Baker.

spectacular entire-leaved form was mostprevalent.

Arguably the most remarkable of thewidespread understory palms was D.vonitrandambo (Fig. 5), a Vonitra-type Dypsisdiscovered by John Dransfield on the 1996expedition to Iketra and described in 2010(Rakotoarinivo & Dransfield 2010). This squatpalm has erect or leaning stems typically up to1.5 m tall that are covered in persistent leafsheath fiber and produce abundant, thin stiltroots at the base. The leaves are unique amongVonitra-type Dypsis in bearing few, broad,hooded leaflets (around 10 each side of therachis), rather than numerous, narrow, linearleaflets. The emerging young leaves are deeppurple-red and the inflorescences are erect andbrush-like, with rather few rachillae. Morestriking is the fact that D. vonitrandambo waslocally very common in wet valley sides andbottoms throughout the Iketra area and yethas never been recorded from any other sitein Masoala. It appears to replace a similarspecies, D. pusilla, which is widespread in WestMasoala, but is unrecorded in the East.

Four larger palms featured among the elevencore species. Another Vonitra-type species,Dypsis dransfieldii, was common throughoutthe area. This moderate, clustering, under- tomidstory species has distinctive inflorescencesthat project from the crown. The handsomeemergent D. lastellianawas also scattered acrossevery site though never in large numbers.Finally, two species of Orania, O. trispatha andO. longisquama (Figs. 6 & 7), were abundant,though not in the same habitats. Thespectacular, distichous O. trispatha occurredmainly in wetter areas, especially aroundstreams and on the banks of theAnaovandrano River. In contrast, O.longisquama was found only on drier slopes.The Iketra form had a compact crown of erect,shuttlecock leaves, in contrast to other formselsewhere on the east coast with looser crownsof recurved leaves.

South of Iketra

We spent two days exploring the ridges andvalleys south of Iketra (Fig. 8). Several settledclearings lie around this area and as a resultpeople are actively exploiting the forest, fortimber extraction and some clearing forfarmland. Nevertheless, the forest was largelyin good condition, supporting primarily thetypical palm flora of the area. Our first dayfocused on a search for another of the “biggame” species of Masoala, Marojejya darianiior velatra as it is known locally. John Dransfieldhad recorded it in two sites around Iketra in1996, and we were anxious to see it ourselves.Our guide Emmanuel assured us that he knewa site in a southwesterly direction from ourcamp, but after some hours walk, wading inand out of streams, and looping back onourselves, we began to doubt that we wouldever find it. Our luck held, however, when wecame into a flat and swampy valley bottomwith a rather broken canopy, through whicharound 30 massive, stemless individuals of M.darianii were scattered with their largelyundivided leaves reaching to 6 m in length(Fig. 9). Encountering this almost mythicalpalm in so wild and remote a place wasincredibly exciting. Curiously, none of theindividuals formed visible stems or showedsigns of flowering. Perhaps this was anestablishing colony, or alternatively stemmedadults may have been targeted for palm heartextraction. Evidence of human activity in thesite suggests that felling for palm heart or othermaterials could be a significant risk to thepopulation.


176

8. Hills south of Iketra. Photo: W.J. Baker.

Our second foray into this general area tookus due south of camp in search of more “biggame.” John Dransfield had reported sightingthree juveniles of Voanioala gerardii, butEmmanuel was hopeful that he could locateadult palms for us. Sadly, he did not succeed,although we did find three juveniles widelyscattered across the landscape. It seems entirelylikely that adults must exist in this area, giventhe rather young age of the juveniles that weobserved, but they may occur at extremely lowdensities. Emmanuel reported that he knew ofone that grew close to Iketra, but that it hadbeen felled for palm heart. On our quest forVoanioala, we made other interesting finds,including a few juveniles of Satranaladecussilvae and occasional individuals ofRavenea julietiae, which is distinctive in thefemale inflorescence protruding from thecrown. We also found a few individuals ofDypsis metallica, discussed further below.

West of Iketra

In principle, the Anaovandrano River providedready access to the forests west of Iketra bypirogue. However, the river is shallow and fast-flowing, with numerous rapids, makingpaddling or punting the pirogue a slow andtiring business. On our first attempt, we tookalmost two hours to travel only 3 km,

repeatedly wading upriver while the pirogueswere dragged over the rapids. The slow journeygave us plenty of time to admire the forest, its


177

9. Marojejya darianii. Photo: W.J. Baker.

10. Dypsis crinita. Photo: W.J. Baker.


178

11. Dypsis reflexa. Photo: W.J. Baker.

wildlife and palms, especially Orania trispatha,Dypsis crinita (Fig. 10) and D. lastelliana.

Finally, we abandoned the pirogue to explorethe hills above the river. Among the ever-present Dypsis pachyramea in the undergrowth,occasional individuals of the superficiallysimilar D. mocquerysiana were concealed. In aswampy stream valley, we found the elegant,slender understory Dypsis reflexa (Fig. 11),discovered by John Dransfield in 1996, withinflorescences arching from the center of thecrown with reflexed inflorescence branchesthat account for the species name. Non-descript D. hiarakae grew nearby. As the pathbecame steeper, palm diversity declined, butbrought new finds, such as D. perrieri and anunknown species with affinities to D. baronii.

We returned to the west of Iketra later in theexpedition to follow up reports of a giantRavenea known locally as kona. Returning upriver by pirogue again, we turned up a narrowtributary some 1.5 km west of our camp. Wethen walked in a westerly direction for 2.5 km,mostly wading up streams through exquisitelybeautiful valley bottom forest (Fig. 12). Weturned repeatedly as one stream met another,but our guides knew the area well, having

spent time there recently hunting for crabs.Finally, we came upon a small colony ofmonumental palms. This immense species ofRavenea (Fig. 13) is known from several sitesalong the east coast of Madagascar and hasbeen confused with R. robustior whereas itappears to be an undescribed species (J.Dransfield, pers. comm.). The specimens thatwe observed towered 30 m above the forestfloor, with 80 cm wide stems at the base thattapered to about half the diameter towards theapex. As we marveled at these forest giants, atroop of endemic red-ruffed lemurs passed by,barking noisily to one another but remainingfrustratingly out-of-sight.

North of Iketra

John Dransfield’s 1996 expedition yieldedespecially rich pickings in a site around 2 kmnorth of Iketra known as Ankorabe or TananyRabe Pierre, meaning Rabe Pierre’s land(Dransfield 1996). We too found this area soexciting for palms that we spent twoconsecutive days collecting there, in thecompany of Monsieur Rabe Pierre himself, thepatriarch of the area who, though nowadvanced in years, clearly rememberedcollecting palms with Dransfield and co. in


179

12 (left). Fieldwork west of Iketra. 13 (right). Ravenea sp. “kona.” Photos: W.J. Baker.

1996. Tanany Rabe Pierre is an intensivelyfarmed forest clearing of perhaps a third of asquare kilometer, sustaining an extendedfamily living in a handful of small houses.

The trail from Iketra to Tanany Rabe Pierrepassed through hilly forest, somewhatdisturbed in places, typical of the area adjacentto Iketra itself, yielding nothing that we hadnot seen elsewhere. However, M. Rabe Pierreguided us to a quite different forest area northof his home. The path led first down into aswampy hollow where we found clumps ofDypsis metallica (Figs. 14 & 15), another speciesdiscovered by John Dransfield during his 1996visit. This is an intensely beautiful, slender,clustering species, somewhat Chamaedorea-likewith its short entire leaves and erect orange-yellow inflorescences. The leaves are very dark,glossy green with a yellow midrib and are veryleathery. Nearby grew D. lantzeana and largeclumps of D. procera.

The path then led up on to a level terrace withwhite sand soils so rich in interesting palmsthat we had no time to explore beyond it. Wefirst came upon a colony of juvenile Marojejyaplants with the local name sira siribe (Fig. 16),all sporting entire, somewhat hooded juvenileleaves, but regularly pinnate leaves at laterstages. This was clearly not M. darianii, but thisleaf character does not correspond with theother species, M. insignis, either. We saw noadults, but Dransfield reported one whichpossessed a stepped trunk, which is alsoinconsistent with M. insignis. This palmunderscores the need for a reappraisal of thetaxonomy of Marojejya now that more isknown about its morphological diversity.


180

15. Dypsis metallica. Photo: W.J. Baker.

14. Dypsis metallica. Photo: W.J. Baker.


181

16. Marojejya sp. “sira siribe.” Photo: W.J. Baker.


182

17. Dypsis sp. “lafaza maitso.” Photo: W.J. Baker.

Perhaps as a result of the nutrient-poorsubstrates on which it grew, the forest herewas made up of slender, even spindly treeswith a rather light canopy. It was, however,rich in tree palms, including Dypsishovomantsina, D. lastelliana, D. perrieri, Oranialongisquama, Ravenea sambiranensis andRavenea sp. “kona.” We also found moreindividuals of the Dypsis species with affinitiesto D. baronii that we had seen west of Iketra(Figs. 17 & 18). This palm, known locally aslafaza maitso, appears to be an undescribedspecies, a suspicion that needs to be confirmedby carefully comparing its characters withother Chrysalidocarpus-type Dypsis species.

Then came arguably the most exciting palmhotspot of the entire expedition. We walkedinto a pock-marked landscape, characterized bypot-holes in the ground, perhaps 2 m wideand half a meter deep. M. Rabe Pierreattributed these to fallen Ravenala trees,though this seemed hard to believe as weobserved few live Ravenala individuals. Wefound huge shuttlecock juveniles namedhovodrakidraky, which we later confirmed as

Masoala madagascariensis (Fig. 19), when welocated a majestic flowering adult. Satranaladecussilvae was scattered across the site,including some impressive groups of adults.

Most exciting however, was the presence ofthree undescribed solitary, canopy Dypsisspecies. The first of these, locally named sira(Fig. 20), was a robust, plumose-leaved species,with pronounced rusty red-brown and grayindumentum on its leaf sheaths, which formedan ill-defined crownshaft. The inflorescencesare erect, highly branched and borne betweenthe leaves. This species was not restricted tothis site – we had seen occasional individualselsewhere around Iketra – but was rathercommon at this site.

The second new species, known as ovojavavy(Fig. 21), grew among Dypsis sp. “sira” and wasof similar stature but was characterized by itsneat, recurved leaves with ascending leaflets,gray-green, rather open leaf-sheaths and itshighly branched, pendulous inflorescence thatemerged between the leaves. The fourthspecies, known as angolafa, remains somethingof a mystery. John Dransfield reported seeingonly juveniles on his 1996 trip and added onlya sighting of a short-trunked, non-floweringindividual. We encountered the exact samesituation: thanks to the large, red scales on thepetiole of this palm, we readily identified somejuveniles as angolafa, but like Dransfield, wefailed to find fertile individuals that wouldallow us to describe this exciting palm as anew species.

Iketra to Tampolo

After six concentrated field days, we hadexhausted the palms of Iketra. We broke campand made the 12 km trek back to Sahamalaza,where we were met by our boat the followingmorning. The boat took us back around CapMasoala, stopping half way up the west side ofthe peninsula at Tampolo, where we wouldhave a further three days’ fieldwork. Tampolois a rather well-known location for palms. WestMasoala is much more mountainous than theeast, with steep slopes rising directly from theshore. The undulating coastal flatlands thatsurround Tampolo are unusual in this respect(Fig. 22). We stayed at the rustic, but verycomfortable Tampolodge, which sits at theback of a painfully beautiful beach, a steep arcof golden sand, backed by glorious littoralforest and with granite crags at either end. Itwas an ideal location to recuperate from ourexertions in East Masoala while having furtherproductive experiences in the field.


183

18. Dypsis sp. “lafaza maitso.” Photo: W.J. Baker.


184

19. Masoala madagascariensis. Photo: W.J. Baker.


185

20. Dypsis sp. “sira.” Photo: W.J. Baker.

From our base in Tampolo, we took theopportunity to explore for more of the “biggame” palms of Madagascar. We had heard ofa site for Lemurophoenix halleuxii above the

nearby village of Ambodiforaha and made thatthe target of our first foray from Tampolo.Ambodiforaha lies about 2 km walk along thecoastal path running north from Tampolo.There we met a local guide, Marco, on whoseland the Lemurophoenix palms grow. Ittranspired that he had guided many otherpalm enthusiasts to this locality. The path tothe site wound inland through local farmland,eventually climbing steeply into themountains behind the low coastal plain. Theforest was rather disturbed but yielded manypalm species, including Dypsis confusa, D.dransfieldii, D. faneva, D. fibrosa, D. forficifolia,D. hovomantsina, D. lastelliana, D. mocquery-siana, D. pachyramea, D. pinnatifrons, D. proceraand Ravenea dransfieldii.

About 1 km inland from Ambodiforaha, onthe edge of a recently opened forest clearingwe came to the Lemurophoenix site (Fig. 23).The population comprised five extremelyimpressive adult palms with stems to around20 m high and 50 cm in diameter. We foundmany yellow-brown ripe fruit on the forestfloor, displaying the typical corky warts andheart-shaped button on the spherical seed.Seedlings were numerous, but survival seemedto be low, as we observed no individuals of anage class between seedling and adult. Theadjacent forest clearing had been opened onlyrecently, which clearly concerned Marco, whorecognized the value of the palm stand as anattraction for tourists. We could not resent the


186

21. Dypsis sp. “ovojavavy.” Photo: W.J. Baker.

22. West Masoala view from Tampolodge. Photo: W.J. Baker.


187

23. Lemurophoenix halleuxii. Photo: W.J. Baker.

rather high guide fee that Marco charged if ithelped to protect this spectacular stand of thisendangered palm.

Hiking beyond the Lemurophoenix site, wecame into high, pristine rain forest. With noclear path to follow, we bashed our waythrough the vegetation hoping to find moreLemurophoenix but without success. The forestwas full of other large tree palms, however,elegant, tristichous Dypsis tsaravoasira,plumose-leaved D. hovomantsinawith its white,waxy crownshaft and the immense D.tokoravina, with its massive crown of open leafsheaths and feathery leaves.

Antalavia

The temptation of the legendary palm locationof Antalavia, some 8 km south of Tampoloproved too much to bear. Antalavia is the typelocality of the forest coconut, Voanioalagerardii, one of the most exciting finds in theentire exploration of the Madagascar palmflora, and perhaps the ultimate “big game”palm on the island. The discovery of Voanioalaby Gerard Jean, John Dransfield and DavidCooke in 1986 is an adventure story more thanworthy of the annals of palm exploration(Dransfield 1989), and all the more impressivefor the challenges of getting aroundMadagascar at that time. In contrast, wehopped on our chartered speed boat atTampolo and were disembarking just half anhour later.

In the Antalavia area, Voanioala occurs ataround 500 m elevation in the mountains thatrise steeply from the shore. To reach this site,Dransfield and co. followed a treacherous,boulder-strewn river, whereas our routeavoided this by starting at Marofototra, asizeable village a little to the north. At first, ourpath passed through low-lying coastal forest,swampy in places and thick with understoryspecies such as Dypsis confusa and D. procera.The route then climbed gently before crossingan idyllic, rocky stream, lined with D. crinita,tree ferns and dracaenas. Beyond the streamcrossing, the ascent began in earnest,becoming increasingly extreme. The palmsightings on the climb are a bit of a blur, as weconcentrated more on staying upright andcatching our breath, but we still managed tospot and impressive diversity of species,including D. dransfieldii, D. faneva, D. fanjana,D. fibrosa, D. hovomantsina, D. lastelliana, D.minuta, D. mocquerysiana, D. pachyramea, D.pinnatifrons, D. pusilla, D. tsaravoasira, Oranialongisquama, Ravenea dransfieldii and R. julietiae.

Eventually the route leveled out on to broad,flat ridge covered in tall forest. Our guide fromMarofototra assured us that we were nearly atthe Voanioala site, but as we slogged on for afurther hour, we all began to doubt that wewould ever make it. We passed impressivespecimens of Dypsis tokoravina and Ravenealakatra, which kept our spirits up. Finally, theguide announced that we had arrived, butVoanioala was nowhere to be seen. A largeMasoala madagascariensis stood nearby and webriefly wondered if we had been taken on awild goose chase to see this, rather thanVoanioala, but then we found mini-coconutendocarps on the forest floor. Scattered in thedense vegetation around us stood three adultVoanioala individuals with their classic steppedtrunks and open-sheathed crowns with Attalea-like inflorescences protruding from betweenthe leaves. A number of juveniles were spottedin the undergrowth, but the species was farfrom abundant. Our GPS indicated that wewere perhaps less than 1 km from the typelocality itself, which suggests that Voanioalamay be widely but sparsely distributed in thelocal area. The inadequate understanding ofthe distribution of this critically endangeredpalm is a significant obstacle to itsconservation. A systematic survey of theMasoala Peninsula would be extremely usefulin this respect but would be very difficult toachieve.

The descent from Voanioalawas as tough as theascent, but by the late afternoon, we reachedthe shore, exhausted, but extremely satisfiedafter a formative palm hunting experience.

Tampolo

Our final day in the Masoala Peninsula wasspent exploring the littoral forest directly southof Tampolodge. We entered the forest via adisused railway line, which had once been usedto facilitate the harvest of timber duringcolonial times. All that remains of the linetoday is a straight and level path, and a fewtimber railway sleepers. Littoral forest of thequality found at Tampolo is very rare now inMadagascar. Many of Madagascar’s botanicalwonders could be seen there, such as Cycasthouarsii growing near the shore and largeplants of Darwin’s Comet Orchid, Angraecumsesquipedale, clinging to rocks. The goldenflowers of Sarcolaena multiflora, a member ofthe woody plant family Sarcolaenaceae, whichis entirely endemic to Madagascar, were found,and the curious leafless Iridaceae saprophyte,Geosiris aphylla lurked on the forest floor.


188

Though the forest was beautiful, it wasrelatively poor in palms. The commonlyobserved species included Dypsis confusa, D.

dransfieldii, D. fibrosa, D. forficifolia, D. procera,D. pusilla, D. tsaravoasira, Ravenea dransfieldiiand R. sambiranensis. Stands of Satranala


189

24. Satranala decussilvae. Photo: W.J. Baker.

decussilvae (Figs. 24 & 25), including somereproductive adults, were distributed along theloop trail that we followed, although they hadbeen badly damaged by local people collectingleaves for thatch. A curious unidentifiedVonitra-type Dypsis was also common, whichproduced long, slender entire leaves at youngerstages, becoming pinnate at later stages, butmature individuals of this species could notbe found. This palm was first seen byDransfield in 1997 and despite many visits tothe area has never been seen fertile. Scatteredindividuals of D. carlsmithii, a species describedfrom cultivation as recently as 2002, toweredin the canopy, looking like giant D. lutescens.

We regretted having insufficient time toexplore the flatlands behind Tampolo in moredetail. This area is still densely forested andmust be relatively accessibly from Tampolodge,and yet, to date, most formal botanical recordsin this area come from the coastal areasbetween Ambanizana and Tampolo, andAntalavia. Who knows what lies undiscoveredin the stunning forests of West Masoala?

Conclusions

Our expedition to eastern and western Masoalawas incredibly productive and easily the mostfruitful in a series of four field trips to different

parts of eastern Madagascar recently organizedby the Royal Botanic Gardens, Kew. In total,we found 46 palm species, including perhapsas many as seven undescribed species (fiveDypsis, one Ravenea, and one Marojejya). Wewere able to obtain sufficient material todescribe four of those new species, includingthree massive canopy palms that are eitherlocally abundant (Dypsis sp. “sira” and Dypsissp. “ovojavavy”) or widespread in north-easternMadagascar (Ravenea sp. “kona“). These findsillustrate that the palm flora of Madagascar,in spite of decades of intensive explorationand taxonomic treatment (Dransfield &Beentje 1995), is still far from being completelyknown. In fact, it is this foundational researchon Madagascar palm diversity that makes itpossible to explore this extraordinary palmflora still further and paves the way for broaderstudies in other disciplines.

Despite our considerable efforts, ourexpedition only scratched the surface of thisvast, underexplored tract of humid forest.Though we penetrated some way into thepeninsula from the East, we were unable topush deeper due to the tough terrain,difficulties of moving supplies and equipment,and unpredictable local taboos (fady). Even theWest, which is better known, has only been


190

25. The endocarps of Satranala decussilvae are ridged and lobed. Photo: W.J. Baker.

explored for palms near to the coast. A morethorough exploration of Masoala, e.g. bymeans of transects crossing the peninsula east-west, would be highly worthwhile, but wouldrequire a much larger effort than theexpedition described here. More than 80% ofMadagascan palm species are threatened withextinction (Rakotoarinivo et al. 2014), and anyyet-unknown species is highly likely to be atrisk also. The chance that species coulddisappear before we have had a chance toexplore them scientifically and horticulturallyis high. The systematic exploration ofMasoala’s magnificent forests and its palmsmust be treated as an urgent priority.

Acknowledgments

We are extremely grateful to our colleagues atthe Kew Madagascar Conservation Centre,especially Franck Rakotonasolo and HeleneRalimanana, without whose support thisexpedition would have been entirelyimpossible. John Dransfield and Stuart Cableprovided guidance tirelessly, ensuring that thetrip was as productive as possible. MadagascarNational Parks and numerous local guides alsoprovided essential logistical support.Madagascar National Parks and theMadagascan Ministry of the Environmentkindly granted permission to collect (permitNo. 259/14/MEEF/SG/DGF/DCB.SAP/SCB) andexport plant specimens. This fieldwork wasmade possible thanks to generous fundingfrom the International Palm Society, the RoyalHorticultural Society (RHS Coke Trust BursaryFund), the National Geographic Society(Global Exploration Fund – Northern Europe),the Bentham-Moxon Trust, Pam Le Couteurand the Royal Botanic Gardens, Kew. WLE wassupported financially by the European UnionFP7-PEOPLE program (grant # 327259).

LITERATURE CITED

DRANSFIELD, J. 1989. Searching for a forestcoconut in Madagascar. In HEPPER, F.N. (ed.)

Plant Hunting for Kew. Royal BotanicGardens, Kew. Pp. 51–60.

DRANSFIELD, J. 1996. Palms in Masoala: reportof fieldwork in November 1996. Un-published report. DOI: 10.13140/RG.2.1.1087.6401

DRANSFIELD, J. 1997. Dypsis acaulis. Principes41: 138 –139.

DRANSFIELD, J. AND H. BEENTJE. 1995. The Palmsof Madagascar. Royal Botanic Gardens, Kewand the International Palm Society.

KISSLING, W.D., W.J. BAKER, H. BALSLEV, A.S.BARFOD, F. BORCHSENIUS, J. DRANSFIELD, R.GOVAERTS AND J.-C. SVENNING. 2012.Quaternary and pre-Quaternary historicallegacies in the global distribution of a majortropical plant lineage. Global Ecology andBiogeography 21: 909–921.

RAKOTOARINIVO, M., A. BLACH-OVERGAARD, W.J.BAKER, J. DRANSFIELD, J. MOAT AND J.-C.SVENNING. 2013. Palaeo-precipitation is amajor determinant of palm species richnesspatterns across Madagascar: a tropicalbiodiversity hotspot. Proceedings of theRoyal Society B-Biological Sciences 280:1757.

RAKOTOARINIVO, M. AND J. DRANSFIELD. 2010. Newspecies of Dypsis and Ravenea (Arecaceae)from Madagascar. Kew Bulletin 65: 279–303.

RAKOTOARINIVO, M., J. DRANSFIELD, S.P. BACHMAN,J. MOAT AND W.J. BAKER. 2014. Compre-hensive Red List assessment revealsexceptionally high extinction risk toMadagascar palms. PLoS ONE 9: e103684.

RAKOTOARINIVO, M., M.S. TRUDGEN AND W.J.BAKER. 2009. The palms of the MakiraProtected Area, Madagascar. Palms 53:125–146.


191


192

Appendix 1: Palm species seen and collected, by date in November, 2015. X = species seen. Collection numbers (e.g. WLE124) are given

where collections have been made.

Iketra

Trek in/out West

South

North

South

West

Ambodi-

Antalavia

Tampolo

foraha

Day

14&21.11

15.11

16.11

17&18.11.

19.11

20.11

23.11

24.11

25.11

Dypsis acau

lisWLE124

Dypsis carlsm

ithii

XDypsis confusa

WB1416

XX

XX

XDypsis crinita

XX

XDypsis dran

sfieldii

XX

WLE131

XX

XX

XX

Dypsis faneva

WB1426

XDypsis fanjan

aWLE149

Dypsis fasciculata

WB1411

XX

XWB1412

Dypsis fibrosa

XX

XX

XDypsis forficifolia

XWLE125

XX

XX

XX

Dypsis hiarakae

WLE127

XDypsis hovoman

tsina

X?

WLE145

XDypsis lantzean

aWB1418A

WLE142

Dypsis lastellia

naX

XX

XWLE143

XX

XDypsis metallica

XWB1420,

WB1424

WLE132

Dypsis minuta

WLE147,

WLE148,

WLE152

Dypsis mocquerysiana

WLE129,

XX

WLE130

Dypsis nodifera

XDypsis pachyram

eaX

WB1413

XX

XX

XX

Dypsis perrieri

WB1414

XDypsis pinn

atifrons

XX

XX

XX

XDypsis procera

XX

WLE133

XX

XX

WLE134


193

Dypsis pusilla

WB1428

XDypsis reflexa

WLE126

Dypsis tokoravina

WLE146

XDypsis tsaravoasira

XX

XX

Dypsis vonitran

dambo

WB1415

XX

Dypsis sp. entire-

WLE153

leaved Vonitra

Dypsis sp. “angolafa”

XDypsis sp. “ovojavavy”

WB1419

Dypsis sp.

XWLE128

WLE137,

aff. baronii

WLE138,

“lafaza maitso”

WLE139

Dypsis sp. aff.

XWB1418B,

lastellia

na“sira”

WLE135

Lemurophoenix

WB1425

halleuxii

Marojejya darianii

WB1417

Marojejya sp.

WLE136

“sira siribe”

Masoala

WLE140

Xmadagascariensis

Orania longisquam

aX

XX

XX

XOrania trispatha

XX

XX

XX

Ravenea dransfieldii

XX

XX

Ravenea julietiae

XWB1423

XRavenea lakatra

XX

Ravenea

XX

sambiranensis

Ravenea sp. “kona”

XWLE144

Satran

ala

WB1427

WB1422

Xdecussilvae

Voanioala gerardii

WLE141

WLE150,

WLE151

Documents

[email protected] The Palms NDONIAINA IJORO … · Madagascar, and yet much of the region is inaccessible and remains unexplored. Here, we report the findings of an expedition in November