ATOLL RESEARCH BULLETIN NO. 579 TERRESTRIAL AND MARINE ECOLOGY OF POIVRE, AMIRANTES, SEYCHELLES BY ANNELISE B. HAGAN, THOMAS SPENCER, JENNIFER ASHWORTH, JUDE BIJOUX, RODNEY QUATRE, MARTIN CALLOW, AND BEN STOBART ISSUED BY NATIONAL MUSEUM OF NATURAL HISTORY SMITHSONIAN INSTITUTION WASHINGTON, D.C., U.S.A. MARCH 2010 □ oPSW2 start oC*/es oPSW< oPSW< oPSW4 start oPSWI start oPSW5 start 0 PSW 6 start end °PSW7 start cPSW3 start r&.Vr ■ 11SP Kilometres Habitat classes Unclassified Saline pond n Rocky fore-reef slope Rock pavement □ Reef-flat sand r .i Other trees and shrubs ^■1 Medium density seagrass □ Mangrove woodland Low density seagrass / macroalgae Littoral hedge Lagoon sand Lagoon patch reef High density seagrass Herbs and grasses ] | Fore-reef slope sand [ | Cleared / bare ground □ Fore-reef slope rubble and sand □ Buildings and other structures □ Fore-reef slope coral spurs with coralline algae n Coral sandstone / raised reef □ Beach sand | | Coral rubble with coralline algae □ Coral boulders Coconut woodland Beach rock Figure 1 . Location of vegetation Line Intercept Transects (PV1 and PV2), beach profiles (PBl, PB2, PB3 and PB4), reef-flat sediment samples (Sed sample), shallow-water transects (PSW1-7) and SCUBA dive surveys at Poivre atoll, 26-27th January 2005. Habitat map from Spencer et al. (2009). TERRESTRIAL AND MARINE ECOLOGY OF POIVRE, AMIRANTES, SEYCHELLES BY ANNELISE B. HAGAN, 1 THOMAS SPENCER, 1 JENNIFER ASHWORTH, 1 JUDE BIJOUX, 2 RODNEY QUATRE, 2 MARTIN CALLOW, 3 AND BEN STOBART 1 INTRODUCTION The Amirantes group, Seychelles, comprises 24 islands and islets lying between 5° and 6° south of the equator on the Amirantes Bank, western Indian Ocean. The islands were discovered by the Portuguese navigator Vasco de Gama on his second voyage to India in 1502, soon after acceding to the rank of Admiral, and the islands were subsequently named Ilhas do Almirante or Admiral’s Islands (Lionnet, 1970). The group extends over a distance of 138 km, from African Banks in the north to Desnoeufs in the south. Poivre was visited by the Chevalier du Roslan and named in 1771 by the Chevalier de la Biolliere, after Pierre Poivre, former governor of lie de France (now La Reunion), an agronomist who was instrumental in setting up the spice industry in the Seychelles in the late 1770s. Poivre Atoll lies close to the eastern margin of the Amirantes Bank, in the southern half of the main bank, 15 km southeast of Sand Cay (which is in the centre of the Bank) and 40 km west southwest of the submerged atoll of Desroches. The platform reef consist of two main islands, Poivre island in the north and lie du Sud in the south, with a large, elongate, shallow lagoon between them and an encircling reef. A third island, Florentin, lies within the western margin of He du Sud. The Percy Sladen Trust Expedition visited Poivre on 9 th October 1905. Having crossed the sand-flat from Poivre island to He du Sud, expedition members became trapped by the tide and thus had time to fully explore the southern island (Gardiner, 1936). The morphology, soils, ecology and agriculture at Poivre island were documented in the 1960s (Baker, 1963; Piggott, 1969) and a plant list was compiled by S.A. Robertson and F.R. Fosberg following a two day visit to the island by the first author on 26 th and 27 th October 1976. Robertson found the island in much the same condition as when described by Piggott (1969) but with greater agricultural activity and with cattle and pigs being kept (Robertson and Fosberg, 1983). When Poivre island was visited in 2005, the only agriculture taking place on the island was for subsistence purposes, undertaken by a small number of workers from the Island Development Company (a government parastatal which took over ownership of the island in 1981). A small settlement exists on the eastern tip of Poivre island. In 2005, a hotel development was underway on the island, but progress has been slow. The centre of the island has been 'Cambridge Coastal Research Unit, Department of Geography, University of Cambridge, Cambridge, CB2 3EN, UK. Seychelles Centre for Marine Research and Technology - Marine Parks Authority, Victoria, Mahe, Seychelles. 3 The Trident Trust, The Smokehouse, Smokehouse Yard, 44-46 St John Street, London, EC1M 4DF, UK. Manuscript received 15 October 2009. 2 bisected by an 1,100 m long airstrip and the existing boat channel is being enlarged to facilitate the entry of larger vessels. The Netherlands Indian Ocean Programme, onboard the R.V. Tyro , called at Poivre on 29 th and 31 st December 1992 and 1 st January 1993. SCUBA and snorkel surveys were conducted on the north and west reef-slopes of Poivre (van der Land, 1994). A single seabed trawl at a depth of 57 m was conducted using a 3.5 m Agassiz trawl west of Poivre atoll and three baited hshtraps were deployed at a depth of 43 m northwest of Poivre island (van der Land, 1994). A collaborative expedition between Khaled bin Sultan Living Oceans Foundation, Cambridge Coastal Research Unit and Seychelles Centre for Marine Research and Technology - Marine Parks Authority to the southern Seychelles was conducted onboard M.Y. Golden Shadow , from 10 th - 28 th January 2005. The primary aim of the expedition was to use a CASI (Compact Airborne Spectrographic Imager) sensor onboard a seaplane to conduct large-scale mapping of the southern Amirantes, Alphonse/St. Francois (Spencer et al., 2009) and Providence Bank. All surveys at Poivre were conducted on 26 th and 27 th January 2005. TOPOGRAPHY AND GEOLOGY Of the seven reef types identified in the Seychelles by Stoddart (1984), three are present in the Amirantes: platform reef, atoll and drowned atoll. The platform reefs vary in their morphology; Spencer et al. (2009) identified three categories of platform reef. Along with the island of D'Arros, they defined Poivre as a Type 3 platform reef, where the infilling of the platform surface has allowed the development of subaerial islands which exceed 2 km 2 in total area. The percentage coverage of the area of the reef platform inside the breaker zone by islands is comparable to the Type 2 reefs but the shallow water habitats seaward of the breaker zone are much less extensive than at the Type 2 sites where the islands sit on extensive, shallow and gently sloping rock platforms. Table 1 provides quantitative information on the Type 3 morphology at Poivre. Table 1. Morphometry of the platform reef at Poivre. Total reef Peripheral Land Land area as platform reef area 2 area 3 proportion of area 1 (km 2 ) (km 2 ) (km 2 ) total reef platform area (km 2 ) 20.24 12.01 2.66 13.14 1 area of terrestrial and shallow marine habitats classified by Spencer et al. (2009) from airborne imagery 2 area between the breaker zone and island marginal sediments 3 area of terrestrial habitats and coarse beach materials (including beachrock) 3 The reef system probably sits on one of a number of topographic highs on the bank margin; local water depths of 15 - 17 m characterize these highs compared to 22 - 39 m on the general bank surface to the west of Poivre. To the east, water depths rapidly increase to in excess of 1,500 m. The reef platform is 5 km in length along a central N - S axis. The eastern margin runs S - N but the western margin is orientated SE - NW in its southerly two-thirds before trending SW - NE. As a result, the atoll is 3.5 km in width towards the north but tapers to less than 2 km in width towards its southern limits. The reef platform supports three reef islands (Fig. 1 ; Plate 1). The most northerly island, Poivre island (110 ha) is a low-lying, oval shaped island measuring approximately 1,700 m from east to west and 750 m north to south (Plate 2). Poivre is connected to the large (135 ha) and topographically complex southern island, lie du Sud (Plate 3), by a 750 m long causeway (Plates 4 and 5) (wrongly positioned by Baker, 1963). The very small (2.4 ha) third island, Florentin, lies behind and within the western margin of lie du Sud. Reef-flat widths, which are modest compared to the widths on other atolls in the Amirantes and the Alphonse Group further south, are greatest on the northwest and northeast coasts, opposite the western and eastern points of Poivre island respectively, where they reach 1 km. Around He du Sud reef-flat widths range between 500-750 m (west-facing coasts) and 750 m (east and south-facing coasts) (Plate 6). Thus, notwithstanding some differentiation between windward and leeward coasts, unlike the majority of islands in the Amirantes, reef flats are not best developed on the highest energy coasts facing the southeast trades. Baker (1963) identified shallow breaks in the reef flat at the southern point of the atoll and at several locations along the western margin but the only substantial channel, which gives access to Poivre island, is on the northern reef. There is no central lagoon, the area between Poivre island and He du Sud is characterized by reef-flat sands and muds which dry at low tide (Plate 7). Poivre island is arcuate in form, with a strongly convex northern margin, measuring approximately 1,700 m from east to west and 750 m north to south. Baker (1963) mapped a large area of phosphatic sandstone at its eastern end, with basins, pipes and crevasses filled with guano and mixed to varying degrees with sand and humus. This phosphate outcrops on the inner reef flat immediately offshore from the settlement and along the eastern margin of the boat channel. Gravel sand bars and small spits characterize Pointe Baleine, the eastern point of the island, and He Corail, at the northeastern end of the airstrip, where the spits enclose a saline pond. A further discrete patch of phosphatic sandstone lies in the centre of the island with two small outliers further west. The largest land area at He du Sud contains a 9 ha area of phosphatic sandstone which presumably represents the old core to the island. This core tapers to the north, to La Pointe; a number of spit structures appear to have developed around it. Firstly, a 2.75 km long eastern arm has probably extended over time northwards, the result of longshore sediment transport driven by waves from the southeast. This arm thins significantly at 1.8 km from the southern tip of He du Sud, suggesting a reduction in long-term sediment supply. Recent (2008) satellite imagery suggests continued northward development, in a small arc of mobile sand between the tip of the vegetated island (Pointe Mozambique) and the causeway to Poivre island. Secondly, a complex of islands - named ‘Grande Terre' by Baker (1963) - extends to the northwest from the central core of He 4 du Sud, formed of a series of closely-spaced low storm-ridges. There are at least three fossil spit units which run SW - NE from the seaward reef-flat margin of this complex. It appears that these spits increase in age to the northwest as the most northwesterly of these units, which has a complex star-like structure with five radiating arms, is clearly currently active, with numerous sand washover deposits and gravel ridges recurving towards the island of Florentin which lies behind this area. This active extension is not shown on Baker’s (1963) map and has therefore developed within the last 40 years. It is likely that the island Florentin is now starved of new sediment and is a fixed, fossil feature. In each case, the positioning of the fossilized and vegetated spits represents the landward extension of features analogous to the Transverse rubble bars’ described from the western reef flat at St. Joseph by Stoddart and Coe (1979). These are tongues of storm debris about 1 m above the reef flat which flare seawards to join with extensive coral rubble banks immediately shoreward of the breaker zone and extend across the whole width of the reef flat. METHODS FOR TERRESTRIAL SURVEYS Terrestrial Flora and Fauna Vegetation surveys were conducted using the Fine Intercept Transect (FIT) technique over a horizontal distance of 30 m. By summing the intercept lengths for each plant species and dividing this value by the total length of the transect, percentage cover for each plant species was calculated: Percentage cover = Total length of plant species x 100 Fength of transect Two transects were laid at Poivre, both on the east side of the island, close to the settlement (PV 1 and PV2; Fig. 1). The short length of these FITs was due to the opportunistic nature in which the work was being undertaken and limited time available. Plants that could be not identified in situ were labelled and photographed with a high resolution (4.1 mega pixels) digital camera for later identification by local botanists Murugaiyan Pugazhendhi and Katherine Beaver. In addition to these vegetation surveys, insects were collected in small plastic jars for identification back on Mahe. Beach Surveys Four beach profiles were conducted on Poivre island on the eastern end (PB1; 5°44.932'S, 53°18.664'E - 5°44.932'S, 53°18.67FE), southwestern (PB2; 5°45.202'S, 53°18.043 f E-5°45.208'S, 53 o 18.042'E), north northwest (PB3; 5°44.67FS, 53°18.123'E - 5°44.66FS, 53°18.117 f E) and northern (PB4; 5°44.606’S, 53°18.335 f E - 5°44.597'S, 53 0 18.333'E) sides of the island (Fig. 1). Profiles were measured by Abney level and tape, 5 in an offshore direction perpendicular to the beach, beginning at the terrestrial vegetation line and continuing to the offshore step (where the waves were breaking, typically marked by a downward step) or as far as safely possible into the water. Eight surface scrape sediment samples of ca. 200 - 350 g by weight were collected from Poivre, typically at the start and end points of each beach profile. Positions were recorded for the start and end of each beach profile and for the sites of the sediment samples using a hand- held GPS unit (horizontal resolution = ±10 m). Sediments were dried, disaggregated and sieved using standard techniques at 0.25 phi intervals. METHODS FOR MARINE SURVEYS Shallow-water Boat Transects Two rigid inflatable boats were used to conduct shallow-water transects at seven sites around Poivre atoll. On the west side of the atoll, two transects were run NW-SE (bearings of 120° and 150°) in towards Poivre island and two transects were run W-E in towards lie du Sud (bearings of 090° and 070°). On the east side of the atoll, transects were run due E-W, one transect was run into Poivre island, one transect was run towards the gap between the two islands and one transect was run in towards lie du Sud. (All shallow water transects are marked on Fig. 1 as PSW1-7). Transects were started at a water depth of approximately 20 m, the limit at which the bottom substrate could be accurately determined from the surface. Each time the boat was stopped a GPS position was taken and the water depth and bottom substrate (viewed through a glass-bottomed bucket) recorded. Between 11 and 18 substrate observations were recorded on each transect line. Benthic Surveys Quantitative benthic surveys (January 2005) were conducted using SCUBA at 20 m, 15 m, 10 m and 5 m depths on the western fore-reef slope (~5°45.334'S; 53°17.348 r E; Dives on Fig. 1). The video transect method was used as this technique enables a large area of reef to be surveyed in a short time period as well as providing a permanent visual record of the reef at a specific time (Carleton and Done, 1995). A Sony digital DCR-SC100 video camera, positioned vertically 30 cm above the substrate, was used to conduct all video transects over a horizontal distance of 20 m following the depth contour of the reef. The video data recorded was a plan view of a rectangular section of benthic reef community measuring 20 m x ~ 0.3 m; by recording both sides of the transect, double the area was covered (or 20 m x ~ 0.6 m). The video transect footage was analysed using the AIMS 5-dot analysis method by pausing the video at regular intervals and recording the substrate captured by each of the 5 dots (Christie et al., 1996; Osborne and Oxley, 1997). Ten major benthic categories were identified: sand, rubble, bare substrate, dead standing coral, pink calcareous algae on bare substrate, pink calcareous algae on dead standing coral, Scleractinia, non- 6 Scleractinia, macroalgae and others (e.g. zooanthids, molluscs, bivalves). Scleractinia, non-Scleractinia and macroalgae were identified to genus level. Percentage cover for the 10 benthic categories was calculated as follows: Percentage cover = Total number of dot captures for single benthic category x 100 Total number of dot captures for entire transect length In addition to the benthic video surveys, fish species observed at Poivre were recorded during the dives. All fish species seen during a 35 minute period at depths of between 20 m and 5 m were recorded. A random search pattern was followed and both pelagic and demersal species noted. Reef-Flat Surveys The video transect technique was also used to conduct quantitative surveys on the reef flats at Poivre. The reef-flat transects ran perpendicular to the beach and covered a horizontal distance of 25 m. Only one side of the tape line was videoed, thus giving a rectangular section of reef measuring 20 m x ~ 0.3 m. Five reef-flat video transects were conducted at Poivre - three on the east side of the island and two on the west side. The video was analysed in the same way as the underwater SCUBA video transects, using standard AIMS methods (Christie et al., 1996; Osborne and Oxley, 1997). Surface scrape sediment samples were collected from the reef flats, off the south-west coast and eastern tip of Poivre island. Seagrass on the reef flats of Poivre was sampled along two 30 m long transects. A pair of 0.5 m x 0.5 m quadrats was placed at each 10 m interval (one either side of the transect), so in total, 8 quadrats were placed along each transect (at 0 m, 10 m, 20 m and 30 m). At each site a visual estimation of percentage cover by different seagrass/ macroalgal species within each quadrat was made. The above ground biomass was then collected, put into labelled plastic bags and frozen. After the expedition, the frozen samples were dried out using a furnace at 225°C. The samples were frequently weighted, until a constant dry weight was obtained. RESULTS OF TERRESTRIAL SURVEYS Flora and Fauna Surveys Poivre island is largely covered in coconut palms and was once an important site of copra production (Plate 8). The first vegetation Line Intercept Transect (PV1; Fig. 1) was conducted approximately 80 m west of the settlement. The area was shady with a high canopy formed by coconut palms ( Cocos nucifera). The dominant vegetation at this site was the sedge Cyperus dabius which occupied 35% of the total cover (Fig. 2). The short grass Stenotaphrum dimidiatum was the second most dominant species (24% cover), followed by the creeper Passiflora suberosa (12% cover) and Stachytarpheta 7 jamaicensis (12% cover). The herb Vernonia cinerea , the roots of which are often used locally for treating abdominal pains occupied 7% of the total cover at this site. Other plant species found along the transect were Crotalaria sp. (4% cover), Cyperus aromaticus (2% cover), Turnera ulmifolia (2% cover), Euphorbia hirta (2% cover) and Alocasia macrorrhiza (0.6% cover). The second transect was laid about 200 metres from the settlement perpendicular to the path going to the southeast of the island (PV2; Fig. 1). Once again this was an area shaded by the high canopy of coconut palms. The vegetation was very low, seldom higher than 40 cm. The dominant species at this site was the mat forming grass Stenotaphrum dimidiatum which covered 39% of the total transect, followed by 2 creepers; Passiflora foetida (27% cover) and Tridax procumbens (15% cover) (Fig. 2). Stachytarpheta jamaicensis occupied 9% of the total cover. Other plant species found along the transect were Nephrolepis biserrata (5% cover), Passiflora suberosa (3% cover), Vernonia cinerea (1.5% cover), Synedrella nodiflora (1.3% cover) and Crotalaria sp. (0.2% cover). Areas of other trees and shrubs, Casuarina forest and groves of broadleaf forest are characteristic near the settlement, near the eastern point, immediately to the east of the northeastern end of the airstrip and along the north coast to the west of the airstrip. lie du Sud and Florentin are also dominated by coconuts, although there are other trees and shrubs in pockets along the western margins of both islands, particularly on Grande Terre. A littoral hedge, dominated by Scaevola taccada is particularly prevalent on the eastern tip of Poivre (Plate 9) and characterizes the eastern and southwestern coast of lie du Sud, being particularly well developed on the southeastern coast (Fig. 1). Historical records, and collections and sight records from 1976, have established a plant list of 87 species at Poivre (Robertson and Fosberg, 1983). Fifty-four plant species were observed at Poivre during a 3.5 hour period in January 2005; these are listed in Table 2. 100% 90% 80% 70% i_ CD o 60% O (D 3 50 % c o fe 40% 30% 20% 10 % 0% FV1 PV2 ■ A. macrorrhiza ■ S. nodiflora □ T. procumbens □ V. cinerea □ C. aromaticus ■ C. dubius □ E hirta B Crotalaria sp. □ N. biserrata B P. foetida □ P. suberosa □ S. dimidiatum B T. ulmifolia □ S. jamaicensis Figure 2. Percentage cover of plant species along two 30 m long LITs at Poivre, January 2005 (see Figure 1 for locations of PV1 and PV2). 8 Table 2. Scientific and Creole / common names of plants observed at Poivre, January 2005. Total number of plants observed = 54. Number of new records compared to Robertson and Fosberg (1983) = 13 (new records marked with *). Family and Species Creole / Common Name Amaranthaceae *Amaranthus viridis Apocvnaceae Catharanthus roseus Roz enmer / Madagascar periwinkle *Ochrosia oppostifolia Araceae Bois sousouri Aloe as ia macrorrhiza Arecaceae (Palmae) Vya Cocos nucifera Asteracea (Compositae) Syne dr ell a nodiflora Tridax procumbens Vernonia cinerea Caricaceae Cocotier / Coconut palm Carica papaya Casuarinaceae Papayer / Papaya Casuarina equisetifolia Clusiaceae (Guttiferae) Cedre / Pin / Casuarina Calophyllum inophyllum Combretaceae Takamaka / Indian laurel Terminalia catappa Convolvulaceae Badamier / Bodanmyen / Indian almond Ipomoea pes-caprae Crassulaceae Kalanchoe pinnata Cucurbitaceae Batatran / Goats foot creeper / Beach morning-glory Cucurbita moschata Cvperaceae Pumpkin *Cyperus aromaticus *Cyperus compressus Sedge Cyperus dubius Euphorbiaceae Sedge Acalypha indica Euphorbia hirta Euphorbia prostrae Phyllanthus amarus Fabaceae (Leguminosae) * Crotalaria sp. Desm odium incanum *Desmodium triflorun Herbe chatte / kerb sat / Cat grass Table 2 (Con’td) Family and Species Creole / Common Name Goodeniaceae Scaevola taccada Scaevola Hernandiaceae Hernandia nmyphaeifolia Bois Blanc Lilaceae Zephyranthes rosea Lomariopsidaceae Nephrolepis biserrata Fern Lythraceae Pemphis acidula Malvaceae Gossypium hirsutum Cotton Sida acuta *Thespesia populnea Bois de rose / Porcher Moraceae Artocarpus altilis Friapen / Arbre a pain / Breadfruit Musaceae Musa sp. Bananier / Banana Nyctaginaceae Boerhavia sp. Pata covin / Patate cauvin Bougainvillea spectabilis Vilea / Bougainvillea Passifloraceae *Passiflora foetida Passiflora suberosa Poaceae (Gramineae) Dactyloctenium ctenoides Grass Eleusine indica Grass Stenotaphrum dimidiatum Grass * Stenotaphrum micranthum Grass Unidentified grass sp. Grass Rhamnaceae *Colubrina asiatica Rhizonhoraceae Rhizophora mucronata Mangliye rouge / Rhizophora mangrove Rubiaceae Morinda citrifolia Bois tortue / Indian mulberry Scrophulariaceae Striga asiatica Lerb du riz Solanaceae *Solanum nigrum Turneraceae * Turner a angustifolia Koket / Coquette Turner a ulmifolia Koket / Coquette Urticaceae Laportea aestuans Verbenaceae *Lantana camara Ville fille Stachytarpheta jamaicensis Epi bleu 10 Beach Surveys Narrow sandy beaches surround Poivre island, widening on the central southern shore (Plate 10) and at the eastern point, where they are associated with coral rubble deposits. Sandy beaches also characterize the seaward margin of lie du Sud, being widest on the southeast coast. Florentin has no sandy beaches. Beach sediments are generally composed of poorly sorted to moderately well sorted coarse to very coarse sands (Table 3, Fig. 3). Textural groups are slightly gravelly sand to sandy gravel. Beach profile PB1, to the west of the eastern point, Pointe Baleine, of Poivre island, is a narrow (12 m), low angle beach, ending in a 1 m high cliff landward margin (Fig. 4a). At the southwestern point (PB2), the beach is a similar width and characterized by slope angles of 4 - 8° (Fig. 4b). On the northern, seaward side of Poivre island (PB3 and PB4), beach widths are wider (18-22 m), with beach angles of 4 - 7° (Figs. 4c and 4d) and some scarping of the upper beach. RESULTS OF MARINE SURVEYS Reef-flat Surveys The reef fl at at the eastern end of Poivre island is composed of poorly sorted, very coarse sands in the textural group sandy gravel. In the southwest, the reef flat is composed of poorly sorted, coarse sand in the textural group gravelly sand (Table 3, Fig. 5). The reef-flat margin on the windward coast is characterized by medium density seagrass, with patches and lineations showing either a WNW - ESE orientation or, near the south point, a NW - SE alignment; away from the seaward margin, medium density seagrass is replaced by a quasi-continuous area of low density seagrass with macroalgae. High density seagrass is rare on the southeastern reef-flat but there is an extensive area at the northeastern corner of the atoll. On leeward coasts west of lie du Sud there is a mosaic of high, medium and low density seagrass with macroalgae. The reef flat to the west of Poivre island is characterized by more extensive coverage of medium density seagrass and a large band of high density seagrass (Plate 11). The centre of the atoll, between Poivre island and He du Sud, and the interior embayments of He du Sud and Florentin, are characterized by bare reef-flat sands and, in places, carbonate muds which dry at low tide. The mangrove Rhizophora mucronata is present between the islands of Poivre and He du Sud, forming a band along the southern coast of Poivre and near to, and along the causeway (Plates 5 and 12) and infilling areas within the southern embayments of He du Sud. 11 dD d co +-> • i-H g b) co d Ch dD dD S-H § co dD b O Hh Sb _G CD d dD -O ,S-H dD a • i-H dD co S-H £ CO cd • i-H H— > CO • i-H 03 h—> CO C G -O • i-H Hh to *3 dD N • i-H CO dD Id • i-H I Oh >n os C3 Pd "d ^H CD CO £ $ ”d d d • -H -G K> Oh dD O hG Ph G rn O dj G dD _o > d -G H bo p o o CO O Q a 4 4 go O P GO TD O G O C/3 13 TD a P GO P • i— H ~o d TD o o t: JO CD O b a. s J !3 & CD CD S-H (D ~o O !2So> co 'C i n o' on Cl o G p b - o T3 G , U S GO no CO P" O'N oo o o d % g 3 P O o +-> GO Oh CD 13 b ^ d d ^ p P3 d § D -b D GO b GO D D D 53 co S-H d co Vh P O p -sur W k ^ > ^ u u > on b in co O co co -p -p -P hP CD a D P D o p D D P D o p D p _Q -P -P X> o S-H S-H S-H S-H S-H ■ ^H D D D D > Dh £ Oh P Oh O b O w b b b b C- >r> p- O Cl b CO CO p- in Cl Cl CO oo oo o m oo p- o D b b b ,—l b -P o p CD hP Vh CD Oh Qh b CD CO -- O P ^ o o > Ph U CO) r- OO o m r- On -P o p CD _o S-H CD o b -H Cl -H Cl -H Cl b b ci ci CO CO PQ PQ PQ PQ pp PQ Ph Ph Ph Ph Ph Ph O GO O u >,u o d p b b s D o > Ph OO in b Ph b b CD CD > > O Cl m cl oo o VO o C| i — i Cl o o\ Cl p- oo o oo CO ON b b b b b b oo P- m NO OO Cl m Cl NO Cl CM — NO T“H o On p- in m b b b b b p- o b b o p CD hO S-H CD Qh Qh b p- PQ Ph PB4.2 Lower beach 0.256 -0.455 1.744 -0.520 0.773 Very Coarse Sand, Poorly Sorted, Very Coarse Skewed, Platykurtic 12 cn C D Q o3 § CO CD a kO TP o £ o p^ t/3 f\ 03 § CO , o PLh CD H CD H H— > C/3 H— > H-» ^ +- P C3 'op c3 -op c3 > S £ 2 £ cq i i ci Ch— 1 J, CD to c +- i j, CD to 4h TP CD p ) Figure 5. Cumulative frequency curves for sediment samples from reef-flats at Poivre island, Poivre. The second transect on the west of lie du Sud (PSW4) observed the substrate between 25 m (5°46.016'S, 53°17.347'E) and 1.9 m water depths (5°45.934'S, 53°17.687'E). Between 25 m and 22 m, the seabed was covered in dense Thalassodendron ciliatum on sand and between 21m and 10 m the substrate was unvegetated sand. Between 10 m and 2.6 m reef was observed, comprising branching corals on coral rock interspersed with Halimeda spp. macroalgae. At 1.9 m a large patch of Thalassodendron ciliatum was observed. Shallow- water Transects: East Side Three shallow-water transects were undertaken on the east side of the islands. The transect that ran due west towards Poivre (PSW5) recorded substrate from 16 m (5°45.09ES, 53°19.43EE) to 5 m (5°45.98ES, 53°19.325'E). Throughout this transect the substrate was observed as coral on coral rock, with sand patches and sand channels shallower than 15 m. The amount of live coral present was estimated to range between 2% and 15% and the dominant genera were Acropora and Pocillopora. No macroalgae or seagrass was observed along this transect. The transect on the east side of the atoll which ran in towards the gap between the two islands (PSW6) observed substrate between 21 m (5°45.657'S, 53°19.383'E) and 3.5 m (5°45.57ES, 53°19.225 r E). Typically the substrate was composed of coral on coral rock, interspersed with sand and rubble. Live coral cover was estimated to range between 5% and 10% and recognisable genera wer q Acropora and Turbinaria. Halimeda spp. macroalgae were observed at one sampling site along the transect. 16 The transect (PSW7) which ran in towards the eastern side of lie du Sud observed the benthos between 20 m (5°46.637'S, 53°19.259'E) and 3.5 m (5°46.689'S, 53°19.065 r E). Between 21 m and 12 m, large amounts of rubble were observed, interspersed with branching corals ( Acropora spp. and Pocillopora spp.), which accounted for approximately 5% of the benthos. At depths shallower than 12 m, coral on coral rock was observed, with live coral cover constituting between 5% and 10% cover. Benthic Surveys Benthic surveys (January 2005) were conducted at 20 m, 15 m, 10 m and 5 m depths on the western fore-reef slope (Dives on Fig. 1). Coral rubble accounted for 53% of the benthos at 10 m, and high levels of rubble were recorded at all depths except for at 5 m depth (Table 4). The amount of bare substrate was greatest at the shallowest (5 m) and deepest (20 m) sites and the amount of calcareous algae was also highest at the deepest site. Live scleractinian cover was greatest at 15 m, followed by that at 20 m, where it constituted 25% and 20% cover respectively (Table 4). Live scleractinian cover constituted only half this amount (approximately 10%) at the shallower two sites. Non- Scleractinia were only observed at the shallower two sites and at 10 m depth, contributed to 5% of the benthos. Macroalgal cover was low at all depths, but displayed the highest cover (4%) at the shallowest site. The coral community was identified to genus level and Table 5 shows the proportions of each coral genus at each of the survey depths. Table 4. Percentage benthic cover on Poivre fore-reef slope at 20 m, 15 m, 10 m and 5 m water depths, from video data analysis. Benthic Category Percentage Cover at 20 m Depth Percentage Cover at 1 5 m Depth Percentage Cover at 10 m Depth Percentage Cover at 5 m Depth Sand 3.1 12.9 8.2 33.3 Rubble 23.6 36.9 53.4 8.8 Bare Substrate 21.9 9.9 12.2 33.4 Dead Standing Coral 0.0 0.1 0.2 0.0 Pink Calcareous Algae on Bare Substrate 30.1 15.6 10.1 8.7 Pink Calcareous 0.0 0.0 0.2 0.1 Algae on Dead Standing Coral Scleractinia 19.5 24.6 8.6 11.0 Non-Scleractinia 0.0 0.0 4.8 0.5 Macroalgae 1.8 0.0 2.1 4.0 Others 0.1 0.0 0.2 0.3 17 Table 5. Coral genera as a proportion of the overall coral community at 20 m, 15 m, 10 m and 5 m water depth, from video transect analysis. Coral Genus Percentage of Percentage of Percentage of Percentage of Coral Coral Coral Coral Community at Community at Community at Community at 20m Depth 15m Depth 10m Depth 5m Depth Scleractinia Porites 39.0 47.8 5.7 23.7 Hydnophora 0.0 0.0 1.7 1.4 Favia 0.0 3.8 1.7 0.7 Favites 1.5 2.3 2.3 7.4 Physogyra 0.4 0.0 0.0 0.0 Galaxea 0.0 0.6 0.0 0.7 Goniastrea 3.4 2.6 4.0 15.5 Leptastrea 1.1 1.5 0.0 0.0 Astreopora 0.0 0.0 2.8 8.1 Cyphastrea 0.8 0.0 0.0 0.0 Goniopora 0.0 0.0 0.0 2.7 Pavona 4.2 12.2 4.0 1.4 Platygyra 0.0 0.6 0.0 1.4 Mycedium 0.8 0.0 0.0 0.0 Acanthastrea 0.0 0.6 0.0 0.7 Acropora 5.3 3.8 6.8 9.5 Pocillopora 21.2 19.1 27.8 18.9 Stylophora 0.0 0.0 1.1 4.1 Montipora 13.6 0.0 0.0 0.0 Leptoseris 1.9 0.3 0.0 0.0 Echinopora 0.0 0.0 0.6 0.0 Fungia 6.4 4.9 2.8 0.0 Herpolitha 0.4 0.0 2.8 0.0 Non- Scleractinia Sinularia 0.0 0.0 0.6 4.1 Dendrophyllia 0.0 0.0 0.6 0.0 Dendronephthya 0.0 0.0 0.0 0.0 Heliopora 0.0 0.0 34.7 0.0 Figures 6-9 graphically display the proportions of the most prevalent coral genera at each survey depth. 18 Figure 6. Percentage cover by different coral genera in order of dominance at 20 m depth, from video transect analysis. POR = Porites, POC = Pocillopora , MON = Montipora , FUN = Fungi a, ACR = Acropora, PAY = Pavona, GON = Goniastrea, LET = Leptoseris, FV = Favites, LEP = Leptastrea. At 20 m, the dominant coral genus was Porites , which represented 39% of the coral community. The second ranking genus, Pocillopora , constitution only half this amount (21%). Montipora was the third most dominant genus and together these three genera constituted three quarters of the total coral community. Coral Genus Figure 7. Percentage cover by different coral genera in order of dominance at 15 m depth, from video transect analysis. POR = Porites , POC = Pocillopora , PAV = Pavona , FUN = Fungia, ACR = Acropora , FAY = Favia , GON = Goniastrea , FV = Favites , LEP = Leptastrea. 19 At 15 m, again the two most dominant genera were Porites and Pocillopora , but at this shallower depth, the dominance of Porites was even greater, representing nearly half of the total coral community. Pocillopora represented 19% of the coral community, but at this depth, the third most dominant genus was Pavona. Figure 8. Percentage cover by different coral genera in order of dominance at 10 m depth, from video transect analysis. HEL = Heliopora, POC = Pocillopora , ACR = Acropora, POR = Porites , GON = Goniastrea, PAV = Pavona , AST = Astreopora, FUN = Fungi a, HER = Herpolitha , FV = Favites. At 10 m, the blue coral Heliopora coerulea was dominant, representing 35% of the coral community, closely followed by Pocillopora spp. which represented 28% of the community. Interestingly, H. coerulea was not found at any other survey depth. The third and forth most dominant genera were Acropora (7% of the community) and Porites (6% of the community). Figure 9. Percentage cover by different coral genera in order of dominance at 5 m depth, from video transect analysis. POC = Pocillopora , POR = Porites , GON = Goniastrea , ACR = Acropora, AST = Astreopora , FV = Favites , STY = Stylophora, SIN = Similar ia, GOP = Goniopora, HYD = Ffydnophora , PAV = Pavona , PFY = Platygyra. 20 At 5 m, coral dominance followed that found at 20 m and 15 m, with Porites (24%) being the dominant genus, followed by Pocillopora (19%). Goniopora and Acropora were found to be the next most dominant genera. Non-scleractinian corals were represented by Similar ia spp. which represented 4% of the coral community. Fish Surveys Seventy-three fish species from 19 families were recorded. These varied in trophic group and size from large lethrinids such as Lethrinus olivaceus to small pomacentrids such as Dascyllus carneus. Labrids were the most speciose family recorded (18 species) followed by pomacentrids (11 species) and acanthurids (8 species). All fish species observed at Poivre are listed, by family group, below: Acanthuridae Labridae Pomacanthidae Acanthurus leucosternon Anampses meleagrides Apolemichthys trimaculatus Acanthurus nigricauda Bodianus anthiodes C entropy ge multispinis Acanthurus nigrofuscus Bodianus axillaris Pomacanthus imperator Acanthurus tennenti Cheilinus fasciatus Abudefduf sexfasciatus Acanthurus triostegus Cheilinus triolobatus Abudefduf vagiensis Ctenochaetus striatus Coris cuvieri Chromis dimidiate Ctenochaetus strigosus Coris frerei Dascyllus carneus Naso lituratus Epibulus insidiator Plectroglyphidodon Gomphosus caeruleus lacrymatus Balistidae Halichoeres scapularis Pomacentrus caeruleus Balistapus undulates Hemigymnus fasciatus Pomacentrus sulfurous Melichthys indicus Hologymnosus annulatus Labrichthys unilineatus Pomacentrus weberi Caesionidae Labroides bicolor Scaridae Caesio xanthonota Labroides dimidi atus Scarus frenatus Macropharyngodon Scarus rubroviolaceus Caransidae bipartitus Scarus sordidus Caranx melampygus Stethojulis albovittata Thalassoma hebraicum Serranidae Chaetodontidae Aetholoperca rogaa Chaetodon auriga Lethrinidae Cephalopholis argus Chaetodon guttatissimus Lethrinus olivaceous Cephalopholis nigripinnis Chaetodon lunula Lethrinus xanthochilus Nemanthias carberryi Chaetodon meyeri Monotaxis grandoculis Plectropomus punctatus Chaetodon trifasciatus Lutianidae Variola louti Cirrhitidae Aprion virescens Siganidae Cirrhitichthys oxycephalus Paracirrhites forsteri Lutjanus bohar Siganus argenteus Monacanthidae Tetraodontidae Gobiidae Ptereleotris evides Cantherhines pardalis Mullidae Canthigaster valentine Holocentridae Parupeneus barberinus Myripristis sp. Neoniphon samara Parupeneus cyclostomus Sargocentron Nemipteridae caudimaculatus Sargocentron spiniferum Scolopsis frenatus 21 The number of fish species recorded at Poivre was lower than in previous surveys in the Seychelles (Jennings et ah, 1995; Spalding and Jarvis, 2002). However the lower sampling effort in this study is presumed to be the major cause of this. Twenty-two species overlapped with the most common and abundant species found by Spalding and Jarvis (2002) in the southern Seychelles and 43 overlapped with Jennings et al.’s (1995) survey of the granitic Seychelles islands. Reef-Flat Surveys The reef-flat video transects identified only two species of seagrass; Cymodocea serrulata and Thalassodendron ciliatum. C. serrulata was the more dominant species, with T. ciliatum only being recorded at one site on the eastern reef-flat (Table 6). However, the video transect results do not concur with the results of the seagrass surveys (Table 7). Table 6. Percentage benthic cover on Poivre reef flat from video data analysis. Benthic Category Reef-flat T1 (East) Reef-flat T2 (East) Reef-flat T3 (East) Reef- flat T4 (West) Reef-flat T5 (West) Sand 11.1 6.3 24.2 62.7 79.4 Cymodocea serrulata 88.9 24.2 75.8 37.3 20.6 Thalassodendron ciliatum 0.0 69.5 0.0 0.0 0.0 Table 7. Percentage composition of seagrass and macroalgae within sample quadrats on Poivre reef flat. * indicates that the species was not recorded within the quadrats but this does not mean that these species were absent at this site. Percentage Composition Seagrass Thalassodendron ciliatum 39.3 Cymodocea serrulata 1.8 Halophila ovalis * Halophila uninervis * Macroalgae Halimeda opuntia 35.3 Halimeda discoidea 7.0 Sargasum sp. * Other compositions * Sand 10.9 Average dry weight (g/m2) 1279.6 22 The seagrass survey results show that the dominant species sampled on the Poivre reef flat was Thalassodendron ciliatum, although this value was equalled by macroalgae of the genus Halimeda. The samples showed a low reef-flat diversity, with just two species (T. ciliatum and H. opuntia ) contributing to three quarters of the total composition. Fauna on the reef-flat included numerous species of sea cucumbers and brittle stars, and between the islands of Poivre and lie du Sud, mudskippers and barnacles were observed. Other Observations A pod of short-finned pilot whales ( Globicephala macrorhynchus) were sighted on the northeast coast of Poivre (at 5°45.317'S, 53°16.594'E). DISCUSSION Terrestrial Surveys Fifty-four plant species were observed during the short visit (3.5 hours) to Poivre in January 2005. This is approximately two-thirds the number described by Robertson and Fosberg (1983) having spent two days on the island in October 1976 (Robertson and Fosberg (1983) identified 87 species in total). Thirteen of the 54 plants observed in 2005 are new records. Forty-one species were identified by both Robertson and Fosberg (1983) and the 2005 expedition but Robertson and Fosberg (1983) identified a further 46 species that were not observed by the 2005 expedition. The vegetation at Poivre was seen to closely resemble that at Marie-Touise, but was less similar to that found at Desnoeufs. This could be attributed to the larger sizes of Marie-Touise and Poivre, and the presence of their extensive coconut groves which create microhabitats that enable a more diverse plant community to be maintained. Robertson and Fosberg (1983) did not indicate which were the most dominant species on Poivre, hence direct comparisons cannot be made. Marine Surveys The 1997-98 coral bleaching event as a result of increased sea surface temperature had a very severe impact on reefs of the Indian Ocean. As coral rubble is a typical sign of recent coral mortality (Rasser and Riegl, 2002), the high proportion of rubble and bare substrate on the reefs of Poivre suggests that this recent bleaching event may have led to a benthic community with reduced scleractinian cover. However, although the granitic Seychelles islands in the north suffered over 90% coral mortality during the 1997-98 ocean warming (Linden and Sporrong, 1999), the southern islands were less severely affected, with average mortality of around 60% (Spencer et al., 2000). Although the extent of the 1997-98 coral bleaching in the Amirantes is unknown, due to the relatively shallow nature of the Amirantes Bank (maximum depth ~70 m but typically 11-27 m) 23 it is likely that the bleaching impact was more similar to in the grantics compared to Seychelles reefs further south which are surrounded by much deeper water. Live coral cover on the western reefs of Poivre was seen to represent between 9% (at 10 m) and 25% (at 15 m) of the benthos in 2005. The two most dominant genera were typically Porites followed by Pocillopora , except at 10 m where Heliopora coerulea dominated. Pocillopora damicornis has been described as an opportunistic species, due to its rapid reproductive cycle, widespread larval dispersal and fast growth rate on settling, enabling it to quickly occupy any newly available space (Endean and Cameron, 1990) such as that available following the 1997-98 coral bleaching event in the Amirantes group. However, the presence of Porites as the most dominant scleractinian genus at Poivre may suggest that these slow-growing, massive colonies survived the 1997-98 bleaching event. The Netherlands Indian Ocean Programme expedition conducted three SCUBA surveys at Poivre, two on the northern reef-slope (5°45'S, 53°18'E) on 29 th and 31 st December 1992 and one on the western reef slope (5°46’S, 53°18’E) on 1 st January 1993. On the northern reef-slope, the first survey reported 41% live coral cover, comprising 67% massive Porites spp., 40% acroporids and 6% pocilloporids and the second survey reported 50% live coral cover comprising 12% Porites rus , 76% acroporids, 10% pocilloporids and 3% Pavona various (van der Land, 1994). The survey on the western reef slope reported only 10% live coral cover, and this comprised 63% massive Porites spp., 14% acroporids and 18% pocilloporids (van der Land, 1994). These previous surveys suggest that the 1997-98 bleaching event removed a number of corals, especially branching corals from the community, which may explain the large amounts of coral rubble recorded on the reefs here. Porites appears to have always been the dominant coral genus found on the reefs of Poivre. Shallow water transects at Poivre indicated that Thalassodendron ciliatum seagrass beds were abundant on the western side of the atoll but less prevalent on the eastern side. On the reef flats of Poivre, Thalassodendron ciliatum was recorded to be the most dominant seagrass species, although it constituted an average of only 39% cover. The sites surveyed at Poivre also exhibited very high coverage by Halimeda spp. macroalgae, specifically Halimeda opuntia and Halimeda discoidea. At Poivre, the substrate on which the seagrass was growing consisted of a greater percentage of mud and silt compared to sand, whereas at other Amirantes islands there was greater percentage of sand and coral rubble in the substrate. This might help explain the greater percentage of Halimeda spp. macroalgae compared to at other Amirantes sites. ACKNOWLEDGEMENTS Observations in the Republic of Seychelles were supported through a collaborative expedition between Khaled bin Sultan Living Oceans Foundation, Cambridge Coastal Research Unit, University of Cambridge and Seychelles Centre for Marine Research and Technology - Marine Parks Authority (SCMRT-MPA). The authors would like to acknowledge Prince Khaled bin Sultan for his generous financial support 24 of the expedition and use of the M.Y. Golden Shadow and Capt. P. Renaud, Executive Director, Khaled bin Sultan Living Oceans Foundation for extensive logistical support. We also graciously acknowledge the encouragement, collaboration, and logistical support provided by the Seychelles Government and thank the Island Development Company, Seychelles, for permission to visit islands in the southern Seychelles. Laboratory analyses for the particle size distributions of beach sediments were undertaken by Chris Rolfe, Senior Laboratory Technician, Department of Geography, University of Cambridge; statistics were obtained through the ‘Gradistat’ package (© S. Blott). We are extremely grateful to Murugaiyan Pugazhendhi and Katherine Beaver for help with vegetation identifications, David Stoddart and Lindsay Chong-Seng for help with vegetation listings, and Sarah Hamylton for providing Figure 1. REFERENCES Baker, B.H. 1963. Geology and mineral resources of the Seychelles archipelago. Geological Survey of Kenya Memoir 3 : 1 - 1 40. Carleton, J.H. and T.J. Done 1995. Quantitative video sampling of coral reef benthos: large-scale applications. Coral Reefs 14:35-46. Christie, C.A., D.K. Bass, S.L. Neal, K. Osborne and W.K. Oxley 1996. Surveys of Sessile Benthic Communities Using the Video Technique. Long-term monitoring of the Great Barrier Reef Standard Operational Procedure Number 2 Australian Institute of Marine Science, Townsville, Australia. 42 pp. Online: http://www.aims.gov.au/pages/research/reef-monitoring/ltm/mon-sop2/sop2-00. html. Endean, R. and A.M. Cameron 1990. Trends and new perspectives in coral-reef ecology. In: Ecosystems of the World 25: Coral Reefs Dubinsky, Z. (ed.) Elsevier, Oxford, UK. pp. 469-492. Folk, R.L. and W.C. Ward 1957. Brazos River bar: a study in the significance of grain size parameters. Journal of Sedimentary Petrology 27:3-26. Gardiner J.S. 1936. The reefs of the western Indian Ocean. I. Chagos Archipelago. II. The Mascarene Region. Transactions of the Linnean Society of London, series 2, Zoology 19:393-436. Jennings, S., E.M. Grandcourt and N.V.C. Polunin 1995. The effects of fishing on the diversity, biomass and trophic structure of Seychelles’ reef fish communities. Coral Reefs 14:225-235. Land, J. van der, (ed.) 1994. Oceanic reefs of the Seychelles, volume 2. Report on a cruise of R.V. Tyro to the Seychelles in 1992 and 1993. Netherlands Indian Ocean Programme National Museum of Natural History, Leiden, The Netherlands. 192p. 25 Linden, O. and N. Sporrong 1999. Coral Reef Degradation in the Indian Ocean. CORDIO Project Report. FRN/ MISTRA/SID A/World BanlCWWF Publication. 108p. Lionnet J.F.G. 1970. Appendix: Names of the islands. Atoll Research Bulletin 136:221-224. Osborne, K. and W.G. Oxley 1997. Sampling benthic communities using video transects. In: Survey Manual for Tropical Marine Resources (2 nd edition) English S, Wilkinson C, Baker V (eds.) Australian Institute of Marine Science Townsville, Australia, pp. 363-376. Piggott, C.J. 1969. A report on a visit to the Outer Islands of Seychelles between October and November 1960. Tolworth: Directorate of Overseas Surveys, Land Resources Division, vi, 122 pp. Rasser, M.W. and Riegl, B. 2002. Holocene coral reef rubble and its binding agents. Coral Reefs 21:57-72. Robertson, S.A. and F.R. Fosberg 1983. List of Plants of Poivre Island, Amirantes. Atoll Research Bulletin 273:165-176. Spalding, M.D. and G.E. Jarvis 2002. The impact of the 1998 coral mortality on reef fish communities in the Seychelles. Marine Pollution Bulletin 44:309-321. Spencer, T., K.A. Teleki, C. Bradshaw and M.D. Spalding 2000. Coral bleaching in the Southern Seychelles during the 1997-1998 Indian Ocean warming event. Marine Pollution Bulletin 40:569-586. Spencer, T., A.B. Hagan, S.M. Hamylton and P. Renaud 2009. The Atlas of the Amirantes Cambridge Coastal Research Unit, University of Cambridge, Cambridge, UK. vi + 66p. Stoddart, D.R. and M.J. Coe 1979. Geography and ecology of D ? Arros Island. Atoll Research Bulletin 223:3-18. Stoddart, D.R. 1984. Coral reefs of the Seychelles and adjacent regions. In: Biogeography and Ecology of the Seychelles Islands Stoddart, D.R. (ed.). Monographiae Biologicae, vol. 55, Dr W. Junk Publishers, The Hague, The Netherlands, pp. 63-81. PLATES 27 Plate 1. Aerial view of Poivre Atoll in 1999, showing Poivre island in the north, lie du Sud in the south and the man-made causeway connecting the two islands (source: Maps Geosystems; reproduced with kind permission of the Government of the Seychelles). 28 Plate 2. Poivre island looking west, January 2005. Note emergent reef-flat sands, extensive reef flats (with dark areas of high density seagrass), channel to Poivre island and range of fore-reef slope habitats (photograph: Herb Ripley, January 2005). Plate 3. lie du Sud and Florentin, looking northwest, January 2005 (photograph: Herb Ripley, January 2005). 29 Plate 4. Poivre island looking northwest, showing causeway connecting Poivre island to lie du Sud. Note lineations in reef-flat low density seagrass and macroalgae (photograph: Herb Ripley, January 2005). Plate 5. View of causeway from Poivre island, looking south to lie du Sud. Note presence of Rhizophora mucronata mangroves on the causeway (photograph: Martin Callow, January 2005). 30 Plate 6. Southwestern reef flat and lie du Sud, looking southeast. Note broad embayments in the reef flat and the series of transverse rubble bars on the reef flat which stretch from coral rubble at the reef-flat margin to underpinning the spit structures on northwestern lie du Sud. Plate 7. Emergent reef-flat sands and muds between Poivre island and lie du Sud with single Rhizophora mucronata mangrove (photograph: Jen Ashworth, January 2005). 31 Plate 8. One of the many tracks on Poivre island showing high canopy of Cocos nucifera (photograph: Martin Callow, January 2005). Plate 9. Littoral hedge showing Scaevola taccada and Casuarina equisetifolia (photograph: Martin Callow, January 2005). 32 Plate 10. Beach on southern margin of Poivre island (photograph: Jen Ashworth, January 2005). Plate 11. Medium density seagrass exposed at low tide on the reef flat on the western side of Poivre island (photograph: Martin Callow, January 2005). 33 Plate 12. Rhizophora mucronata mangroves on the southern side of Poivre Island (photograph: Jen Ashworth, January 2005).