ATOLL RESEARCH BULLETIN NO. 350 PISONIA ISLANDS OF THE GREAT BARRIER REEF PART I. THE DISTRIBUTION, ABUNDANCE AND DISPERSAL BY SEABIRDS OF PISONIA GRANDIS BY T. A. WALKER PISONIA ISLANDS OF THE GREAT BARRIER REEF PARTII. THE VASCULAR FLORAS OF BUSHY AND REDBILL ISLANDS BY T. A. WALKER, M.Y. CHALOUPKA, AND B. R KING. PISONIA ISLANDS OF THE GREAT BARRIER REEF PART 111. CHANGES IN THE VASCULAR FLORA OF LADY MUSGRAVE ISLAND BY T. A. WALKER ISSUED BY NATIONAL MUSEUM OF NATURAL HISTORY SMITHSONIAN INSTITUTION WASHINGTON D.C., U.S.A. JULY 1991 SEA 'J . (60 m m e gauge) (104 m w e peak) Figure 1-1. The Great Barrier Reef showing localities referred to in the text. Mean monthly rainfall data is illustrated for the four cays and the four rocky islands where records are available. Sizes of the ten largest cays on the Great Barrier Reef are shown below - three at the southern end (23 -24s) and seven at the northern end (9-1 1s). L i d Island 14 years (1973-1986) armual mean 15% mm annual median 1459 mm (10 metre gauge) (341 m m e peak) 4m 1 m Low Islet 97 yeam (1887-1984) annualmeana080mm annual median 2038 mm $> .:+ .:.:. + 8 n - Pine Islet 52 yeus (1934-1986) "A O ' O N D M J J A S 100 . m 100 O ONDJFIVlnJJAS MO O ONDMJJAS MO O ONDMJJAS &al mean 878 mm. m a l m e d m 814 mm (58 m w e hgh puge. 68 m e m i d d peak) Nonh Reef Island l6years (1961-1977) m u a l mean 1067 mm. m m l median 1013 mm Haon Island 26 years (19561982) annual mean 1039 mm, m a l median 1026 mm Lady Elliot Island 47 yeus (1539-1986) annual mean 1177 mm, m a 1 median 1149 mm PISONIA ISLANDS OF THE GREAT BARRIER REEF PART 111. CHANGES IN THE VASCULAR FLORA OF LADY MUSGRAVE ISLAND BY T. A. WALKER ABSTRACT The flora of Lady Musgrave Island has progressed from a belt of small strand trees surrounding a central scrub less than a metre high in 1843 to a mature Pisonia grandis forest in the 1980s. This succession was interrupted by phosphate mining and severely retarded over seventy years by grazing goat herds. Seven plant species lists recorded between 1927 and 1989 illustrate the changes occumng from natural and anthropogenic processes. 45% of the flora are naturalised alien species. There has been an eight-fold increase in human visitation in recent years but this has not resulted in a significant increase in colonisation by weeds. INTRODUCTION Lady Musgrave Island is the second most southerly island (23O54'S. 152O23.E) on the second most southerly reef of the Great Barrier Reef. It is the first island in the Bunker Group and consists of 13 ha (above high tide) of coral shingle, sand and phosphate rock. Lady Musgrave Island is the best example of a Great Bamer Reef island where advancing Pisonia grandis forest is displacing other vegetation ( Figure 1-1 ). Substantial areas of vegetation and soil were cleared for phosphate mining in the 1890s and the vegetation was stripped bare by goats released by the miners in 1898 (Ellis 1936). Over twenty thousand seabirds, primarily Wedge-tailed Shearwater Puffinus pacificus and Black Noddy Anous minutus, nest on the cay during summer. Less abundant species of ground-nesting terns have been partially displaced by tourists visiting the cay. The island experienced a progres- sive eight-fold increase in human visitation between 1984 and 1989 elevating it to the position of sixth most heavily visited cay on the Great Bamer Reef. The potential impact of tourists on the vegetation is primarily that of introduction of weeds and other alien species. Summer and winter floristic surveys were carried out on 12 January (dry conditions) and 3 June 1989 (following extensive rain) in order to evaluate vegetation changes that have occurred since previous surveys. FLORISTIC HISTORY The earliest description of the island and its flora is that of Jukes (1 847) who landed on 7 January 1843: 'The beach was composed of coarse fragments of worn corals and shells, bleached by the weather. At the back of it a ridge of the same materials, four or five feet high, and as many yards Department of Environment and Heritage, PO Box 5391, Townsville 4810, Australia. 0 metres 200 Fig. 3-1. Lady Musgrave Island in 1967 and in 1987 showing the increase in cover by Pisonia grandis over twenty years. A few isolated trees were not identifiable from the aerial photography. across, completely encircled the island, which was not a quarter of a mile in diameter. Inside this regular ridge were some scattered heaps of the same stuff, the whole encircling a small sandy plain. The encircling ridge was occupied by a belt of small trees, while on the plain grew only a short scrubby vegetation, a foot or two in height. The materials of the encircling ridge were quite low and thinly covered with vegetable soil among the trees; but the sand of the central plain, which was dark brown, was sufficiently compact to be taken up in lumps, and a l i ae underneath the surface it formed a kind of soft stone, with imbedded fragments of coral. Some vegetable soil also was found, a few inches in thickness in some places, the result of the decomposition of vegetable matter and birds' dung." Jukes also noted that the trees were "loaded" with the nests of Black Noddies and that shearwaters were abundant. More recently, floristic surveys have been carried out on seven occasions commencing in 1927 (Tbls. 3-1 3-2; Fig. 3-2), The first survey in November 1927 reported only eleven species growing (h4acGillivray and Rodway 1931). At that time up to 300 goats (23 per ha) had removed all small trees, shrubs, herbs and grasses with the exception of a few sessile plants growing too close to the ground to be grazed (Nebe 1928, Napier 1928, MacGillivray 1928, MacGillivray and Rodway 193 1). The four species of trees were browsed clear to a height of 120 cm and the goats had started to eat bark from Pisonia grandis. Only two "miserable specimens" of Tournefortia argentea were Notes to Table 3-1: 1 - not on the incomplete 1966 list but possibly present; 2 - seeds of Caesalpinia bonduc but no plants; 3 - inferred tobestillpresent by Cribb butnot seen by Belmont& Lentfer with the exceptionof anunidentified Asteraceae (possibly Conyza bowiensis); 4 - species not identified withcertainty; 5 -seen by Belmont & Lentfer but not by Cribb; 6 - species uncertain, variously referred to B. a lb i 'om, B. d i f f i a and B. tetrandra; 7 -seen by Cribb but not by Belmont & Lentfer; 8 - not seen by Belmont & Lentfer but noted as Panicum sp. by Cribb; 9 - not recorded by Chaloupka and Domm but almost certainly present. 33 - -- Table 3-1. Vascular plant species recorded at Lady Musgrave Island in different years. NATIVE SPECIES Casuarina equisetifolia Ficus opposita Pandanus tectorius Pisonia grandis Portulaca oleracea Caesalpinia bonduc Twrnefortiaia argentea Hydrocotyle acutiloha Abutilon asiaticum Achyranthes aspera Canavalia rosea lpomea pescaprae Lepturus repens Sesuvium portulacastrum Solanum arnericanum Sporobolus virginicus Tetragonia tetragonioides Thuarea involuta Tribulus cistoides Boerhavia repens6 Malvastrum coromandelianum Ophioglossum vulgatum Ruppia mariCma Oxalis perennans lpomea macrantha Euphortia tannensis Spinifex sericeus lpomea sp. NATURALBED ALIEN SPECIES Oxalis corniculata Coronopus integrifolius Coronopus didymus Cyperus rotundus Amaranthus viridis Bidens pilosa Cakile edentula Cenchrus echinatus Sonchus oleraceus Conyza bonariensis Cynodon dactylon Eleusine ind i i Euphortia prostrata Argemone ochroleuca Digitaria ciliaris EragrosCs minor Panicum maximum Trachymene cussonii lpomea indica Conyza wmatrensis Lepidium bonariense Lepidium virginicum Raphanus raphanistrum CULTIVATED SPECIES Lycopersicon esculentum Cucurbita p e p Carica papaya Allium cepa Solanum hrberosum Cocos nucifera Status in 1989 common, south-western margin common, throughout cay centre common, north and east strand dominant forest over cay widespread. not abundant large thickets, south-eastern side common, north and east strand common in forest clearings widespread. not abundant common, south half periphery common, south-eastern herb fields co-dominant ground cover pond meadow, rare elsewhere widespread, not abundant uncommon, south-eastern clearings rare, south-eastern strand only co-dominant ground cover widespread, not abundant ullcommon dense patch in southeast clearing uncommon, mainly ground cover widespread, not abundant rare one patch on herbs, camp area rare rare widespread, not abundant patchy abundance. southern half common, north and eastern strand common but patchy uncommon, patchy, south half of cay few, south-eastern clearings uncommon, patchy uncommon, patchy rare one patch, northern strand few, northern strand large patch near light tower few rare, in clearings rare small vines, camping area six vines, camping area seedling 34 pp Table 3-2. Non-cultivated plant species at Lady Musgrave Island in different years. Total species number 11 3 1 35 40 39 41 5 1 Native species number 7 20 22 23 23 24 28 Alien species number 4 11 13 17 16 17 23 Alien species % 36 3 5 37 42 41 42 45 Sea-dispersed species 3 14 15 17 17 17 17 Sea-dispersed species % 27 45 43 43 44 41 3 3 alive and MacGillivray and Rodway (193 1) observed that the "animals seem now to be dependent upon the dead leaves that fall from the trees and the seaweed on the beach". There was competition between goats and nest-building Black Noddies for every leaf that feu to the ground (MacGiUivray and Rodway 193 1, Nebe 1932). Nebe (1928) described the Pisonia trees as stunted whereas Napier (1928) referred to 40 foot (12 m) high trees. MacGillivray and Rodway (1931) reported that most of the centre of the cay was occupied by Ficus opposita from 15 to 20 feet (4.5-6 m) high and abundance of this species was previously inferred in the 1890s by Ellis (1936) who noted flocks of doves feeding on the "wild figs". Four years after the 1927 survey the goats had been greatly reduced by hunting and "much undergrowth, mostly Abutilon and coarse grasses" was present (Nebe 1932). Five years later the cay was "thickly wooded and the tangled undergrowth made direct levelling or cross-traverses impossible" (Steers 1938). A small holiday resort was built and operated at the southern side of the cay during the 1930's and the last remnants and debris were removed in 1984. The extent of island habitat modification and garden importation associated with the resort is unknown but no cultivated vegetables or ornamentals survived at the time of the next floristic surveys with the possible but unlikely exceptions of Lycopersicon, Carica or Cucurbi~ . Over the following forty years the goat herd fluctuated in size from hunting (surviving at least one eradication attempt in 1948) until they were eliminated in 1971 with the exception a single animal that escaped until 1974 (P. Ogilvie, pers. comm.). H. S. Curtis sketched a vegetation map in October 1965 and made an incomplete species collection in November 1966 (Queensland Forestry Department files). Floristic surveys undertaken in July 1967 and in April 1969 were combined into a vegetation list by A. B. Cribb (letter to the Forestry Department). The lists of Curtis and Cribb are shown under "1966" and "1969" respectively inTable 3-land demonstrate a large increase in native and alien plant species since the grazing-impoverished list of 1927. In August-September 1975 twenty-eight plant species were recorded by Belmont and Lentfer (Heatwole 1984). Cribb also visited in July 1975 and listed seven new species in addition to species reported in 1969 (letter to the Forestry Department). Cribb noted that goat damage to Casuarina and Pandanus was repairing and that "removal of goats has led to a marked increase in ground cover of plants. Bare shingle ridges of conglomerate exposed by phosphate mining had been fairly conspicuous features of the cay during the 1969 visit but are now mostly obscured by ground cover plants." Three new species established at the cay in the early 1970s. Four more new species were present in January 1982 when Elsol (1986) mapped the vegetation and analysed distribution patterns. xx x Total Fig. 3-2. Increase in species diversity - X at Lady Musgrave 1&nd from 192-7 to 1989. Strong linear correlations X between species diversity and time (r = 0.996 total, 0.993 native, 0.975 0 00 0 Native alien) are misleading. Between 0 surveys in 1927 and 1%5 the plant mm Alien density fluctuated with changing hunting intensity on the goats and 1 diversity would also have fluctu- ated widely. 0 Year I I I 1920 1940 1960 1980 2000 The 1982 floristic survey marks a turning point for the island vegetation. The last effects of destructive or selective grazing by goats had finally disappeared and subsequent floristic surveys in 1984 (Chaloupka and Domm 1986) and 1989(Tables 3-1, 3-2)indicate that species diversity has stabilised. Neglecting the introduced food plants, four species disappeared between 1982 and 1989 and five new species appeared. The changes may be less if, as seems likely, there was confusion with identification or collection of the two species each of Coronopus, Conyza or Lepidium reported. Ipomea indica was almost certainly overlooked or dormant in the 1984 survey as it dominated the same site in 1982 and in 1989. The vegetation is presently experiencing strong competition for space. Establishment of new species has become difficult and some established species are being displaced by expansion of others. Dense ground cover of Thuarea, Lepturus and other species inhibit establishment of Cusuarina seedlings and the population is decreasing slowly as the old trees die (but seedlings have been planted in recent years by park staff). Pisonia stands have expanded vegetatively shading out all other species to cover an area in 1989 more than twice that covered in 1967(Fig.3-1). Displacement of previously dominant Ficus opposita has been accompanied by extinction of the frugiverous Bar-shouldered Dove Geopelia humeralis population since 1927. Pisonia forest permits no undergrowth and in the absence of destructive cyclones or other events will form the climax vegetation. Pisonia was presumably present at the time of phosphate mining although the description of Jukes (1 847) indicates that Pisonia cover was not significant in 1842. Impenetrable thickets of Caesalpinia bonduc are presently expanding and displacing other species (Elsol 1982). In 1989 several Pisonia and Pandanus trees were severely overgrown by this woody vine which could pose a challenge to the dominance of Pisonia. Ipomea indica also blanketed a large area of ground vegetation and trees near the navigation light in June 1989. This vine spread rapidly from a small area in January as a result of unusually high rainfall throughout the first half of the year. PLANT DISPERSAL AND COLONISATION The method of a species' arrival at a cay can be partially inferred from diaspore adaptations for anemochory, hydrochory, epizoochory or endozoochory and from previous knowledge of dispersal events. Arrival mechanisms are unclear when species have more than one natural dispersal mode or when there has been introduction of garden soil, building materials or other materials likely to contain diaspores. Building materials were imported to Lady Musgrave Island for a shelter hut in - - 36 about 1930, for resort buildings in the late 1930s, for an automatic navigation light toweFiTi974 and for camper toilets in 1987. Food plants including tomato, pumpkin, pawpaw, onion and coconut have been taken to Lady Musgrave Island by visitors from time to time. These were intentionally planted or have propagated from discarded seeds and are not considered in the following examination of plant colonisation. About 15% of species recorded at Lady Musgrave Island have diaspores that are dispersed by wind but with one exception this is primarily a short-distance mechanism. Offshore winds are weak, infrequent and not often likely to transport seeds of grasses, weeds or Casuarina across 60 km of sea from the mainland. One exception is Ophioglossum vulgatum (presumably the same species reported as 0. lusiranicum on adjacent Fairfax Island by Cribb 1986) which has the tiny fern spores capable of extended aerial buoyancy. The grasses and weeds with wind dispersal capability have alternative dispersal modes by birds or people which are the more probable routes of amval at Lady Musgrave Island. Casuarina equiserifolia is assumed to have colonised by sea dispersal. Plants with diaspores dispersed by the sea make up 33% of the total species list and 54% of the native species(Tab1e 3-2) .There are two sea-dispersed alien species, Trachymene cussonii and Cakile edentula. About 65% of all species appear to have dispersed to Lady Musgrave Island via birds or people. The corresponding proportions ofnative and alien zoochorous species are 43% and 9 1% respectively. It is often difficult to differentiate between dispersal of small seeds carried internally by birds and those carried externally by birds or by people. One can speculate that the bulk of the alien species were accidentally camed to the cay by campers and tourists but there is little direct evidence to support this. Weeds initially introduced to Australia by human agency have subsequently spread by natural means. Chaloupka and Domm (1986) have argued that anthropochory is the primary determinant of colonisation of southern Great Bamer Reef cays by alien plants. Using data from the Capricorn- Bunker Islands they reported that the percentage of alien species on each cay was strongly correlated with the amount of human visitation. This correlation was attributed to inadvertent dispersal of diaspores attached to the clothing and footwear of the visitors (anthropochory). The conclusions of Chaloupka and Domm (1986) have been reviewed by Heatwole and Walker (1989) who showed that while anthropochory may be an important process, other factors including introduction of gardens and soil, habitat modification by human activities and avian zoochory could equally well account for the observed patterns of alien plants on the Capricorn-Bunker Islands. The eight-fold increase in tourist visitation at Lady Musgrave Island between the 1984 and 1989 floristic surveys makes it an ideal location to observe the effects of anthropochory. Annual numbers of campers at Lady Musgrave Island increased from 212 in 1984 to 992 in 1985 and 1,475 in 1988 (Department of Environment and Heritage permit records). Prior to 1985 the island was closed to campers for six months each year but from 1985 camping was permitted year-round. Accessibility of the cay was greatly boosted in July 1985 by commencement of a large tourist catamaran service from the mainland. The vessel and its associated seaplane camed more than 11,000 day-visitors in 1988. In addition to this the mean numbers of cruising yachts and motor boats present at the reef increased from 2.1 and 2.3 respectively in 1984 to 3.8 and 3.8 respectively in 1988 (data from twice-weekly aerial surveillance flights by the Department of Environment and Heritage). Simple estimates from the camper, tourist and vessel records (assuming an average of two people go ashore from each private vessel which remains at the reef for an average of two days) - Fig. 3-3. The relationship between arrival and survival of a new plant species. C 0- Low Low High Availability of Suitable Habitat indicate that numbers of people visiting the cay increased from about 1,800 in 1984 to about 15,000 in 1988 (the most heavily visited cay on the Great Barrier Reef is Green Island which received about 240,000 tourists in 1988). The number of visitors in the four years from 1985 to 1988 is probably greater than the total number of visitors to Lady Musgrave Island during the preceding thirty year period. Floristic differences between 1984 and 1989 involve only four species and are not consistent with a flora primarily influenced by anthropochory. Argemone ochroleuca disappeared (probably a seed- bank temporal) while Raphanus raphanistrum, Conyza sumatrensis and a new species of Ipomea appeared (the Ipomea resembles I . indicu but is glabrous and may be an undescribed taxon). Conyza sumatrensis could have been present in 1984 but overlooked amongst the Conyza bonariensis (Raphanus raphanistrum was rare and might also have been overlooked previously). The floristic changes at Lady Musgrave Island from 1927 to 1989 indicate that habitat disturbance is more important to establishment of native or alien plants than anthropochory. This is a long-term evaluation because the relative importance of diaspore amval undoubtedly fluctuates with respect to diaspore survival. There may at times be a surplus of immigrant diaspores but no suitable available habitat, or an excess of available habitat with an absence of immigrant diaspores. The situation most likely to occur is between these extremes with the probability of species establish- ment being directly dependent on the product of diaspore arrival frequency and habitat availability (Fig.3-3). CONCLUSION The floristic composition of Lady Musgrave Island is typical of the Capricorn-Bunker islands. The only species not recorded from the other islands of the group is the unidentified newly amved Ipomea species. Scaevola sericea and Wollastonia biflora are notably absent from Lady Musgrave Island and from adjacent East Fairfax Island which was also defoliated by goat herds until recent years (Cribb 1986). These two sea-dispersed species may have difficulty colonising the strand which is excavated by nesting turtles during summer. Only 55% of species recorded at the island are native to Australia. The large increase in numbers of visitors to the cay between 1984 and 1989 has not been accompanied by significant change in the botanical species composition. If these visitors are transporting numerous diaspores to the cay then there can be little suitable habitat available for colonisation and trampling effects must be insignificant in creating such habitat. -- Human interference with the vegetation has been severe particularly with respect to phosphate mining and release of goats but these effects should not obscure the fact that the flora of Lady Musgrave Island has progressed from scrubby vegetation less than a metre high in 1843 to a predominant cover of mature Pisonia grandis forest. Mining and grazing retarded this progression but, as with some other Capricorn-Bunker cays, the vegetation was notably less advanced last century prior to known European impact. Whether this is indicative of the youth of the cays or of a history interrupted by cyclonic destruction is unknown. The abundance of phosphate cay rock could be interpreted as evidence of a previous Pisonia forest (Fosberg 1957) but such rock might also form in the absence of Pisonia ( Part I ). ACKNOWLEDGEMENTS I thank George Batianoff and the Botany Branch, Department of Primary Industries, for assistance with confirmation and identification of species, and Peter Ogilvie for assistance with historical and archival informa tion. REFERENCES Chaloupka, M. Y. and Domm, S. B. (1986). Role of anthropochory in the invasion of coral cays by alien flora. Ecology 67: 1536-1 547. Cribb, A. B. (1986). The terrestrial vegetation of Fairfax Islands, Great Barrier Reef. Qd. Nat. 26: 119-126. Ellis, A. F. (1936). Adventuring in coral seas. Angus and Robertson, Sydney. Elsol, J. A. (1986). Vegetation of an eastern Australian coral cay - Lady Musgrave Island, Great Barrier Reef. Proc. Roy. Soc. Qd. 96: 33-47. Fosberg, F. R. (1957). Description and occurrence of atoll phosphate rock in Micronesia. Am. J. Sci. 255: 584-592. Heatwole, H. (1984). Terrestrial vegetation of the coral cays, Capricornia Section, Great Barrier Reef Marine Park. In: The Capricornia Section of the Great Barrier Reef: Past, present and future. (Ward, W. T. and Saenger, P. eds.) Roy. Soc. Qd. and Aust. Coral Reef Soc., Brisbane. pp 87-139. Heatwole, H. and Walker, T. A. (1989). Invasion of coral cays by alien plants. Ecology 70: 787-790. Jukes, J. B. (1847). Narrative of h e surveying voyage of HMS Fly. T. and W. Boone, London. MacGillivray, W. D. K. (1928). Bird-life of h e Capricorn and Bunker Islands. Emu 28: 230-250. MacGillivray. W. D. K. and Rodway, F. A. (193 1). Plants on islands of the Bunker and Capricorn Groups. Reports of the Great Barrier Reef Committee 7: 58-63. Napier, S. E. (1928). On the Great Barrier Reef. Angus and Robertson, Sydney. Nebe, J. (1928). A naturalist's holiday on the Great Barrier Reef. Qd. Nat. 6: 102-108. Nebe, J. (1932). A second holiday on the Great Barrier Reef. Qd. Nat. 8: 54-59. Steers, J. A. (1938). Detailed notes on the islands surveyed and examined by the geographical expedition to the Great Barrier Reef in 1936. Reports of the Great Barrier Reef Commiuee 4: 51-104. Plate 1. Pisonia grandis with large double trunk (Cairncross Island, October 1988). Plate 2. Shallow Pisonia root system uncovered by Megapodius reinwardt digging at Bushy Island (August 1989). Plate 3. Soil profile in Pisonia forest at Heron Island (August 1988). Less than a metre of sand with Pisonia humus and mots overlying white coral sand. Plate 4. Coppice shooting fmm a fallen Pisonia branch (Heron Island, August 1988). Plate 5. Pisonia forest at Lady Musgrave Island (September 1986 photo from light tower). Plate 6. Deciduous Pisonia forest (Douglas Island, October 1988).