Spatial and Temporal Variation of Coral Recruitment in Taiwan

Transcription

1 Spatial and Temporal Variation of Coral Recruitment in Taiwan Keryea Soong 1, Ming-hui Chen 1, 2, Chao-lun Chen 3, Chang-feng Dai 4, Tung-yung Fan 2, Jan-jung Li 2, Honmin Fan 1, Kun-ming Kuo 5, Hernyi Hsieh 6 1: Institute of Marine Biology, National Sun Yat-sen University, Kaohsiung, Taiwan 2: National Museum of Marine Biology & Aquarium, Pingtung, Taiwan 3: Institute of Zoology, Academic Sinica, Taipei, Taiwan 4: Institute of Oceanography, National Taiwan University, Taipei, Taiwan 5: Kenting National Park Authority, Pingtung, Taiwan 6: Penghu Aquarium, Taiwanese Fisheries Research Institute, Penghu, Taiwan Corresponding author: Keryea Soong Fax no.: , keryea@mail.nsysu.edu.tw Keywords: coral, recruitment, spatial scale 1

2 Abstract Coral settlement off Taiwan s southern coast between 1997 and 2000 was investigated using artificial plates. A hierarchical experimental design was used, exploiting a total of 723 plastic plates. Scales of spatial variation were analyzed for the above nest-design for 4 consecutive years. In 1999 and 2000 a similar design was expanded to 5 other isolated coral reefs as well as to 2 non-reef regions around the whole of Taiwan, with a total area of about 400 km x 200 km, and exploiting an additional 902 plates. Most coral recruits (~90%) belonged to brooding species, i.e., pocilloporids, with some acroporids and a few unidentified ones. Densities of coral recruits varied by more than 3 orders of magnitude between years at certain sites. Some reef sites had none or very few recruits for the 4 consecutive years of this research. Significant spatial variation occurred more frequently at small scales than at large scales during the 4 years of analysis. For example, variations at the Region level, with a scale of km, can all be explained by that occurring at smaller scales. The two non-reef regions recorded no coral recruits in both 1999 and No combinations of significant levels repeated in any of the years of this study. The pattern of recruitment variation found in this study suggests: 1. Self-seeding at the scale of 10 m is a likely explanation for aggregation of pocilloporid recruits. Moreover, most coral recruitment of a 10 km reef off Southern Taiwan for example, may be limited to, or concentrated at, a particular site for several years. 2. Great spatial variation in recruitment may persist for several years for brooding pocilloporids in Taiwan. The dominating factors, which influence coral recruitment in different spatial scales, may change from year to year. 2

3 Introduction The effect of recruitment on the structure and dynamics of marine populations and communities has received much attention in recent years (Caley et al. 1996, Gains and Roughgarden 1985, Grigg 1988, Roughgarden et al. 1988, Sutherland 1990). Even for coral reefs, which are often dominated by long-lived corals, recruitment plays a critical role in the distribution and abundance of the species (Bak and Engel 1979, Connell et al. 1997, Harriott and Banks 1995, Miller et al. 2000, Rylaarsdam 1983). The history of coral recruitment studies can be traced back a century (Duerden 1904, Harriott and Banks 1995, Stephenson 1931). However, most research on the subject has been conducted in the last 3 decades (e.g., Birkeland 1977, Birkeland et al. 1981, Goreau et al. 1981, Grigg 1988, Hughes et al. 1999). Research has now shown that coral larvae are sensitive to microhabitats, e.g., substrate composition (Benayahu and Loya 1984, Harriott and Fisk 1987), neighbors (Maida et al. 1995), light conditions and orientation of substrates (Babcock and Mundy 1996, Rogers et al. 1984). Densities and composition of coral recruits also differ among seasons and between years, often by several orders of magnitude (e.g., Harriott 1985, Hughes 1985, Rogers et al. 1984, Wallace 1985, Yoshioda 1996). Different reefs may have different patterns of recruitment, dominated by either brooding or broadcasting species (Fitzhardinge 1986, Harriott 1992, Harriott and Fisk 1988, Kojis and Quinn 2001). The patterns of recruitment are known to change with different spatial scales (Dunstan and Johnson 1998, Fisk and Harriott 1990, Fitzhardinge 1985, Sammarco 1985, Sammarco and Andrews 1988). Despite the numbers of studies, 3

4 the results so far seem to suggest that coral recruitment pattern is a characteristic particular to a reef, and there is still a lot to learn before a reliable prediction can be made. The scale of spatial distribution of coral recruitment has been explored mostly on the GBR and its neighboring reefs (e.g., Hughes et al. 1999). It has been suggested that the pattern found in this 2000-Km stretch differs from reefs that are relatively isolated from one another (Dunstan and Johnson 1998, Harriott and Banks 1995). For example, a larva in the GBR has a good chance, after drifting for 2 weeks and for 200 Km, to settle on the GBR (Harriott and Fisk 1988). This scale of self-seeding is unlikely for many other reefs that are relatively isolated and small (e.g., see Kojis and Quinn 2001). Thus, a different recruitment pattern may exist, and the underlying mechanisms may be revealed by first investigating the spatial patterns (Levin 1992, Underwood and Chapman 1996) Information of coral recruitment is important when the reef is degrading due to natural or anthropogenic factors. A damaged reef may or may not have the potential to recover depending on the recruitment rate (Edwards and Clark 1998, Hughes and Tanner 2000). Thus recruitment patterns can potentially determine the best approach to conserve or restore a natural coral reef, once the damaging agents have been eliminated In this study, we first surveyed the coral recruitment pattern in Southern Taiwan for two years, then the study sites were expanded to the whole of Taiwan for an additional 2 years. The purpose was to find out the scale of spatial variation that could reveal the source of recruits, and especially if the above pattern was consistent throughout the years. Materials and Methods Preliminary experiments 4

5 Since the way plates are set up influences the densities of coral settlement (see references in Mundy 2000), several preliminary data sets were obtained before the experiment was expanded to encompass all of Taiwan. 1. Plate orientation: Stacks of 11 plates with 1-cm space between them were fixed on either horizontal or vertical orientations. This comparison was made at Hobihu, at 5 m depth. Coral spats that settled on the upper/lower surfaces of horizontal plates, and those that had settled on surfaces of vertical plates, were compared 1.5 months later. 2. Plate material and density: We compared ceramic tiles and PVC plates in 1997, as well as 2-plates versus 11-plates per rack in horizontal orientation in 1997 and in vertical orientation in Depths: We compared 4 depths, i.e., 1, 5, 10 and 15 m at 2 sites, i.e., Shiju (site 1) and Tiaoshipi (site 7, Figure 1) in Southern Taiwan, April All the above comparisons were made with other factors fixed. Scale of spatial variation The data used to analyze scales of spatial patterns were based on coral recruits on horizontal ceramic tiles in Then vertical PVC plates were adopted in 1998 and 2000 for higher densities of coral recruits (see Results). The vertical carbonic plates used in 1999 were no less suitable for coral spats, but they were less resistant to abrasion than PVC plates. Although there were modifications in procedures between the years (Table 1), the orientation, number of plates per rack and the material of plates were kept constant between sites in each year. The plates were placed at 5-10 m depth in the Hengchun Penninsula area, on welldeveloped fringing reefs in Southern Taiwan, in 1997 and A total of 4 "Areas" 5

6 (several km apart) were designated, and in each "Area" there were two "Sites", about 1- km apart from each other. Two racks, about 10 m apart, were placed in each of the 8 sites, and this level (racks) was designated Locations. In addition to Southern Taiwan, the experiment was expanded to 5 other "Regions" covering all the major reefs, i.e., Hsiaoliuchiu, Penghu, Northeast, Green Island and Orchid Island around Taiwan, and 2 non-reef Regions, i.e., Shitzwan and Yungan, in 1999 and 2000 (See Figure 1). In this Taiwan wide study, the whole of the Hengchun Peninsula in Southern Taiwan was treated as a region with 8 sites, i.e., combining Area and Site into the Area/Site level, with a scale of 1-5 Km between each other. All sites in the reef regions were fringing reefs (Randall and Cheng 1977, Randall and Cheng 1979). The two non-reef sites, Shitzwan and Yungan were located along the sandy West coast of Taiwan. A few species of corals were present on hard or artificial substrates, but there were no reefs at these two localities. Plates were set at a depth of about 5 m at all the sites. Spawning of many broadcasting species occurs in late spring (around May) in Southern Taiwan (Dai et al. 1992), so the plates were placed 1-2 weeks before the spawning dates and retrieved 4-8 weeks later. In 1998, most late-spring plates were lost; although data of another period in January-February were available. Very few coral recruits occurred in 2000, and retrieval of plates was delayed until September for Orchid Island. After collection, the plates were air-dried, and brought back to the laboratory for analysis using a magnifying glass and a dissecting microscope. Identification of coral spat was based on backtracking corals 1-2 years old in a separate study (Kuo 2001), as well as those published in Australia (Baird and Babcock 2000, English et al. 1997). 6

7 Results Preliminary experiments A panel has an area of 225 cm 2, including the holes (1.5 cm diameter) we drilled in the plates. Thus 1 recruit per panel is equivalent to a recruitment rate of 44/m 2. Coral recruitment densities were almost twice as great on vertical as on horizontal plates (38.0 vs. 20.5/panel, p<0.01, Mann-Whitney U test). The above results were obtained after ignoring the 2 outward facing surfaces of the 11-plate assemblages, since they are obviously under different environmental conditions. These ignored surfaces often had high amount of sediments and/or algae, and relatively low number of coral recruits. Among vertical plates, for example, the surfaces facing each other had an average of 43.9 recruits/panel, higher than the external surface (2.7/panel, p<0.01, Kruskal-Wallis test). Among horizontal plates, lower surfaces had higher densities of coral recruits than upper surfaces (20.5 vs. 2.7/panel, p<0.01, Wilcoxon signed rank test). The difference in settlement on ceramic tiles (58.3±9.7/panel) and PVC plates (39.8±7.4/panel) was not significant (p=0.08, Mann-Whitney U test). The number of plates per rack made a difference in coral recruit densities among vertically orientated racks. Higher densities were recorded in 2-plate racks than in 11-plate racks (54.7 vs. 38.0/panel, p<0.01, Mann-Whitney U test). However, no such significant difference was found among horizontal plates (9.6 vs. 5.4/panel, p=0.36, Mann-Whitney U test, Table 2). At both of the sites used to test the effect of depth on coral recruitment, 15 m was found to have the lowest densities. The difference, however, was significant at only one of the two sites. The other 3 depths, i.e., 1, 5 and 10 m did not differ in coral recruit densities (Table 3). Temporal and spatial comparisons 7

8 In 1997, pocilloporids were the dominant recruits (95%), while acroporids constituted the rest (5%). In 1998, the only year when data outside the broadcasting season was used, only pocilloporids were found. In 1999 and 2000, pocilloporids were still the dominant group (90 and 95%, respectively); in addition acroporids were found in Penghu, a region in the Taiwan Straits. The total number of coral recruits ranged from 217 in 1999 to 468 in In 4 years, only 2 recruits belonging to groups other than pocilloporids or acroporids were discovered. Densities of recruits were low except at the Hobihu site (Site 5 in Hengchun Region, Figure 1) in 1997 and 1998, where more than 10 coral spat were recorded per panel. Otherwise, the highest averages at a site were all below 3/panel in 1999 and The Hobihu site recorded the highest recruitment in the Southern region of Taiwan (with at least 8 sites each year) for 3 consecutive years. This is especially pronounced in 1997 and 1998 (Figure 1). In Southern Taiwan, significant variation occurred at the "Site" level (1 km apart) in both 1997 and 1998, and at the "Location" level (10 m apart) from 1998 to 2000 (Table 4). Area level spatial variation was significant only in As to a comparison for the whole of Taiwan, significant spatial variation occurred at the "Location" level in both 1999 and 2000, and at the Area/Site level (1-5 km) only in No significant difference was found at the Region (100 km) level in either year (Table 5). No coral recruits occurred at Yungan and Shitzwan, the 2 non-reef sites of the study in 1999 and Discussion 8

9 Our preliminary results confirmed that microhabitats make a great difference in the coral recruitment densities on artificial substrates (Table 2). Most of the factors have been discussed in great details by earlier investigators (see citations in Introduction). Little attention has been paid to the number of plates per rack, although it also made a difference in some situations (when they were oriented vertically in our experiment). We believe the mechanism may be related to an intensification effect (Pineda and Caswell 1997) in which the abundance of suitable substrate inversely affected the recruit densities. To exclude the factors of microhabitats, our comparison of spatial patterns has to be based on plates with same materials, sizes, orientation, depth as well as plate numbers in the rack. This is accomplished within each year, but variation in set-up existed among years (Table 1). The inconsistency among years is due to our effort to increase overall recruit densities on the plates, so that possible patterns were more likely to be detected. Densities of coral recruits were low in Taiwan from 1999 to This was especially pronounced in the broadcasting species and when compared to results in the GBR (up to 700/panel, see quotes in Harriott 1992). Whereas mass spawnings were observed by both scientists and amateur divers in the past 10 year period, in Southern Taiwan (Fan et al. 2001), successful recruitment of broadcasting species seems extremely limited. Similar phenomena was also observed around the high latitude (30 o S) islands south of the Great Barrier Reefs (e.g., Dunstan and Johnson 1998, Harriott and Banks 1995). High rates of recruitment may occur several months after mass spawning for broadcasting species in Bowden Reef, GBR (Babcock 1988). Most planktonic larvae might have traveled hundreds to thousands of Km away from the reefs producing them (Scheltema 1986, Williams et al. 1984). Thus inter-reef dispersal may be the main source 9

10 of recruitment for these species (Harriott and Fisk 1988). The results of an independent study with new plates placed every two months in Southern Taiwan did reveal that new settlement of spawning corals might occur several months after the mass spawning time in April-May, but the densities were nevertheless low (Kuo 2001). The rates of coral recruitment is known to vary greatly among years (e.g., Fitzhardinge 1985, Hughes 1985, Hughes et al. 2000, Wallace 1985), thus one cannot rule out the possibility that a low rate of recruitment is natural in most years for isolated reefs like those around Taiwan. Occasional recruitment success of broadcasting species may be enough to maintain the long-lived populations (Veron and Don 1979, Yoshioda 1996); although an extended period of recruitment failure may adversely affect the coral communities under threats from various sources (Hughes et al. 2000). Coral reefs at the southern portion as well as at other regions of Taiwan are under threats from both natural and anthropogenic sources (Dai et al. 1998). Eutrophication and high sediment load are known to adversely affect coral recruitment (Babcock and Davies 1991, Harrison and Ward 2001, Hodgson 1990, Nzali et al. 1998, Wittenberg and Hunte 1992). These factors, however, cannot explain the low rates of coral recruitment in all our study sites, since offshore islands, e.g., Green Island and Orchid Island, are apparently neither affected by eutrophication nor by excessive sediment loads (personal observation). Taiwanese coral reefs are, on average, 150 Km from one another (our estimation), thus isolation may be invoked to explain the domination of recruitment by brooding corals (see Dunstan and Johnson 1998, Harriott 1992, Harriott and Banks 1995, Kojis and Quinn 2001). Despite their very low recruitment densities, broadcasting corals dominated the coral communities in all our study sites (Dai 1991). 10

11 The consistently higher recruits at Hobihu among 8 sites in Southern Taiwan can be explained 2 ways. The first hypothesis involves physical, or special current conditions at Hobihu. A model, using surface current data and topographical features of the area, predicts that there is a cyclonic (anticlockwise) eddy in the west side of the bay (where Hobihu is) during ebb (Lee et al. 1999). Thus, if coral recruits settle during ebb tide, we expect to find more recruits on the West than on the east side of the Southern Taiwan Region. However, the resolution of the model is not fine enough to predict a difference between the Western and the Northern sites (site 4 and 5 in Figure 1) in this study. A similar mechanism, i.e., invoking local currents, has been used to explain aggregation patterns of coral recruits in other studies (Fisk and Harriott 1990, Sammarco and Andrews 1988). If the above mechanism indeed applies to Southern Taiwan, we would expect the same aggregation pattern to apply to species of other groups, e.g., barnacles, which remains to be investigated. The second hypothesis assumes that most larvae of the brooding pocilloporids do not travel far after planulation; thus high local recruitment may simply reflects high local abundance of fecund colonies (Harrison and Wallace 1990). This hypothesis is supported by the results of a survey in 1998 of existing colonies of Seriatopora hystrix, the dominant pocilloporid in the region. Hobihu site had the highest number of colonies, largest mean colony diameters as well as the highest estimated coverage (3 times that of the second highest site, and more than 10 times that of the rest) of S. hystrix along belt transects of 20 m 2, among the 8 sites investigated in the region (unpublished data). Under this hypothesis, the dispersal of S. histrix may be limited even within the scale of a Region in this study. 11

12 The high variation in coral recruit densities between the years, e.g., that at Hobihu (Figure 1), was also evident in other studies (Fitzhardinge 1985, Hughes 1985). Besides the explanation of the yearly variation hypothesis mentioned previously, severe bleaching occurred in all coral reefs around Taiwan in the summer of 1998 (personal observations). This may be one of the causes of the reduction in coral recruitment in the following years (see Michalek-Wagner and Willis 2001). Significant spatial variation occurred at the Site level in 1997 and 1998, and at the smallest, Location level in 3 out of 4 years of the surveys in the Southern Taiwan region. It seems to suggest that the source of coral recruits is very local. This is compatible with the hypothesis that most larvae of brooding pocilloporids do not disperse far upon settlement. Transplantation of fecund colonies to sites with low abundance of corals may effectively increase local densities of brooding species Significant variations in various spatial scales are inconsistent over the years in this study. This differs from that of the GBR in which both brooders and spawners had a consistent pattern over 2 years (Hughes et al. 1999). In a 4-year study at Heron Island (GBR), however, spatial pattern of coral recruits were not consistent between years, and this prompted Dustan and Johnson (1998) to suggest that coral recruitment patterns are determined by mechanisms that manifest over a large range of spatial scales. Wallace s (1985) discovery of reversed rank order of recruitment at compared sites from one year to the next is compatible with the hypothesis that the dominating factors affecting recruitment may shift over the years. Since coral recruits in Taiwan are dominated by a few brooding taxa, spatial variation may be affected by very local events. For example, dispersal of brooding larva may be influenced by wind velocity that tends to be variable in the Hengchun Region (Leu 2001). When wind velocity is combined with tidal current, 12

13 dispersal of short-ranged brooding larvae may be highly variable from year to year. The same mechanism may make little difference on larvae of broadcasting species, since they have a long planktonic stage. Any small-scale variation tends to even out after a long period. Acknowledgements We would like to thank the numerous volunteer divers who participated in this research. Two reviewers, who wish to remain anonymous, contributed significantly to the improvement of the manuscript. The project was sponsored by grants to K. Soong from the National Science Council of Taiwan, ROC (NSC B ) and from the Taiwan Power Company References Babcock R, Davies P (1991) Effects of sedimentation on settlement of Acropora millepora. Coral Reefs 9: Babcock R, Mundy C (1996) Coral recruitment: consequences of settlement choice for early growth and survivorship in two scleractinians. J. Exp. Mar. Biol. Ecol. 206: Babcock RC (1988) Fine-scale spatial and temporal patterns in coral settlement. Proc. 6th Int. Coral Reef Symp., Australia 2: Baird AH, Babcock RC (2000) Morphological differences among three species of newly settled pocilloporid coral recruits. Coral Reefs 19: Bak RPM, Engel MS (1979) Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Mar. Biol. 54:

14 Benayahu Y, Loya Y (1984) Substratum preferences and planulae settling of two Red Sea alcyonaceans: Xenia macrospiculata Gohar and Parerythropodium fulvum fulvum (Forskal). J. Exp. Mar. Biol. Ecol. 83: Birkeland C (1977) The importance of rate of biomass accumulation in early successonal stages of benthic communities to the survival of coral recruits. Proc. 3rd Int. Coral Reef Symp. 1: Birkeland C, Rowley D, Randall RH (1981) Coral recruitment patterns at Guam. Proc. 4th Int. Coral Reef Symp., Manila 2: Caley MJ, Carr MH, Hixon MA, Hughes TP, Jones GP, Menge BA (1996) Recruitment and the local dynamics of open marine populations. Annu. Rev. Ecol. Syst. 27: Connell JH, Hughes TP, Wallace CC (1997) A 30-year study of coral abundance, recruitment, and disturbance at several scales in space and time. Ecol. Mono. 67: Dai CF (1991) Reef environment and coral fauna of Southern Taiwan. Atoll Res. Bull. 436: 1-21 Dai CF, Chen YT, Kuo KM, Chuang CH (1998) Changes of coral communities in Nanwan Bay, Kenting National Park: J. Nat. Park (Taiwan, ROC) 8: (in Chinese with English abstract) Dai CF, Soong K, Fan TY (1992) Sexual reproduction of corals in northern and southern Taiwan. Proc. 7th Coral Reef Symp., Guam 1: Duerden JE (1904) The coral Siderastrea radians and its postlarval development. Carnegie Institute, Washington (1-130) Dunstan PK, Johnson CR (1998) Spatio-temporal variation in coral recruitment at different scales on Heron Reef, southern Great Barrier Reef. Coral Reefs 17: Edwards AJ, Clark S (1998) Coral transplantation: a useful management tool or misguided meddling? Mar. Poll. Bull. 37: English S, Wilkinson C, Baker V (1997) Survey manual for tropical marine resources. Australian Institute of Marine Science, Townsville 14

15 Fan TY, Guo KM, Guan CS, Tsai YC, Dai CF (2001) Coral spawning in Kenting, Southern Taiwan, a 10-year review. Newsletter of Taiwanese Coral Reef Society (December, 2001), Taipei (1-6 (in Chinese)) Fisk DA, Harriott VJ (1990) Spatial and temporal variation in coral recruitment on the Great Barrier Reef: implications for dispersal hypotheses. Mar. Biol. 107: Fitzhardinge R (1985) Spatial and temporal variability in coral recruitment in Kaneohe Bay (Oahu, Hawaii). Proc. 5th Int. Coral Reef Cong. 4: Fitzhardinge R (1986) Spatial variability in the recruitment of corals and other organisms in Kaneohe Bay, Oahu. In: Jokiel PJ, Richmond RH, Rogers CL (eds.) Coral reef population biology. Hawaiian Institute of Marine Science, Report #37 (pp ) Gains SD, Roughgarden J (1985) Larval settlement rates: a leading determinant of structure in an ecological community of the marine intertidal zone. Proc. Natl. Acad. Sci., USA 82: Goreau NI, Goreau TJ, Hayes RL (1981) Settling, survivorship and spatial aggregation in planulae and juveniles of the coral Porites porites (Pallas). Bull. Mar. Sci. 31: Grigg RW (1988) Recruitment limitation of a deep benthic hard-bottom octocoral population in the Hawaiian Islands. Mar. Ecol. Prog. Ser. 45: Harriott VJ (1985) Recruitment patterns of scleractinian corals at Lizard Island, Great Barrier Reef. Proc. 5th Int. Coral Reef Symp., Tahiti 4: Harriott VJ (1992) Recruitment patterns of scleractinian corals in an isolated sub-tropical reef system. Coral Reefs 11: Harriott VJ, Banks SA (1995) Recruitment of scleractinian corals in the Solitary Islands Marine Reserve, a high latitude coral-dominated community in eastern Australia. Mar. Ecol. Prog. Ser. 123: Harriott VJ, Fisk DA (1987) A comparison of settlement plate types for experiments on the recruitment of scleractinian corals. Mar. Ecol. Prog. Ser. 37: Harriott VJ, Fisk DA (1988) Recruitment patterns of scleractinian corals: a study of three reefs. Aust. J. Mar. Freshwater Res. 39:

16 Harrison PL, Wallace CC (1990) Reproduction, dispersal and recruitment of scleractinian corals. In: Dubinsky Z (ed.) Ecosystem of the world 25, Coral Reefs. Elsevier, New York (pp ) Harrison PL, Ward S (2001) Elevated levels of nitrogen and phosphorus reduce fertilization success of gametes from scleractinian reef corals. Coral Reefs 139: Hodgson G (1990) Sediment and the settlement of larvae of the reef coral Pocillopora damicornis. Coral Reefs 9: Hughes TP (1985) Life histories and population dynamics of early successional corals. Proc. 5th Int. Coral Reef Cong., Tahiti 4: Hughes TP, Baird AH, Dinsdale EA, Moltschaniwskyj N, Pratchett MS, Tanner JE, Willis B (1999) Patterns of recruitment and abundance of corals along the Great Barrier Reef. Nature 397: Hughes TP, Baird AH, Dinsdale EA, Moltschaniwskyj NA, Pratchett MS, Tanner JE, Willis BL (2000) Supply-side ecology works both ways: the link between benthic adults, fecundity, and larval recruits. Ecology 81: Hughes TP, Tanner JE (2000) Recruitment failure, life histories, and long-term decline of Caribbean corals. Ecology 81: Kojis BL, Quinn NJ (2001) The importance of regional differences in hard coral recruitment rates for determining the need for coral restoration. Bull. Mar. Sci. 69: Kuo KM (2001) Study on the recruitment, growth and survival of juvenile corals at Nanwan. Institute of Marine Biology. National Sun Yat-sen University, Kaohsiung, Taiwan Lee HJ, Chao SY, Fan KL (1999) Tide-induced eddies and upwelling in a semienclosed basin: Nanwan. Estuar. Coast. Shelf Sci. 49: Leu Y (2001) Adaptation mechanism of eclosion date dimorphism in the marine midge Pontomyia oceana (Diptera: Chironomidae). Institute of Marine Biology. National Sun Yat-sen University, Kaohsiung (51) Levin S (1992) The problem of pattern and scale in ecology. Ecology 73:

17 Maida M, Sammarco PW, Coll JC (1995) Effects of soft corals on scleractinian coral recruitment. I: directional allelopathy and inhibition of settlement. Mar. Ecol. Prog. Ser. 121: Michalek-Wagner K, Willis BL (2001) Impacts of bleaching on the soft coral Lobophytum compactum. I. Fecundity, fertilization and offspring viability. Coral Reefs 19: Miller MW, Weil E, Szmant AM (2000) Coral recruitment and juvenile mortality as structuring factors for reef benthic communities in Biscayne National Park, USA. Coral Reefs 19: Mundy CN (2000) An appraisal of methods used in coral recruitment studies. Coral Reefs 19: Nzali LM, Johnstone RW, Mgaya YD (1998) Factors affecting scleractinian coral recruitment on a nearshore reef in Tanzania. Ambio 27: Pineda J, Caswell H (1997) Dependence of settlement rate on suitable substrate area. Mar. Biol. 129: Randall RH, Cheng Y-M (1977) Recent corals of Taiwan. Part I. Description of reefs and coral environments. Acta Geol. Taiwanica 19: Randall RH, Cheng Y-M (1979) Recent corals of Taiwan. Part II. Description of reefs and coral environments. Acta Geol. Taiwanica 22: 1-32 Rogers CS, Fitz HCI, Gilnack M, Beets J, Hardin J (1984) Scleractinian coral recruitment patterns at Salt River submarine canyon, St. Croix, U.S. Virgin Islands. Coral Reefs 3: Roughgarden J, Gaines S, Possingham H (1988) Recruitment dynamics in complex life cycles. Science 241: Rylaarsdam KW (1983) Life histories and abundance patterns of colonial corals on Jamaican reefs. Mar. Ecol. Prog. Ser. 13: Sammarco PW (1985) The Great Barrier Reef vs. the Caribbean: comparisons of grazers, coral recruitment patterns and reef recovery. Proc. 5th Int. Coral Reef Cong. 4: Sammarco PW, Andrews JC (1988) Localized dispersal and recruitment in Great Barrier Reef corals: the helix experiment. Science 239:

18 Scheltema RS (1986) On dispersal and planktonic larvae of benthic invertebrates: an eclectic overview and summary of problems. Bull. Mar. Sci. 39: Stephenson TA (1931) Development and the formation of colonies in Pocillopora and Porites. Part I. Sci. Rep. Great Barrier Reef Exped : Sutherland JP (1990) Recruitment regulates demographic variation in a tropical intertidal barnacle. Ecology 71: Underwood AJ, Chapman MG (1996) Scales of spatial patterns of distribution of intertidal invertebrates. Oecologia 107: Veron JEN, Don TJ (1979) Corals and coral communities of Lord Howe Island. Aust. J. Mar. Freshwater Res. 30: 1-34 Wallace CC (1985) Seasonal peaks and annual fluctuations in recruitment of juvenile scleractinian corals. Mar. Ecol. Prog. Ser. 21: Williams MD, Wolanski E, Andrews JC (1984) Transport mechanisms and the otential movement of planktonic larvae in the central region of the Great Barrier Reef. Coral Reefs 3: Wittenberg M, Hunte W (1992) Effects of eutrophication and sedimentation on juvenile corals. I. Abundance, mortality and community structure. Mar. Biol. 112: Yoshioda PM (1996) Variable recruitment and its effects on the population and community structure of shallow-water gorgonians. Bull. Mar. Sci. 59:

19 Figure Legend Figure 1. Recruitment density (No./plate surface) of pocilloporids in Taiwan. The settlement plates were recovered at sites 1, 2, 4, 5, 6, 7, 9, 10 in 1997 and 1998; at sites 2, 3, 4, 5, 6, 7, 9, 11 in 1999; and at sites 1, 3, 4, 5, 6, 8, 9 in Site labels on the X-axis were deleted for those sites with no plates retrieved. 19

20 Yungan Penghu Northeast Northeast * 24 Green Island Shitzwan * Yungan * Shitzwan Green Island Hsiaoliuchiu Hsiaoliuchiu Orchid Island Hengchun Orchid Island Hengchun Site Site 20

21 Table 1 The set up of plates and racks in each year. Plate Material Ceramic tiles PVC plates Carbonic plates PVC plates Orientation Horizontal Horizontal Vertical Vertical No. of plates/rack

22 Table 2 Summary of preliminary comparisons of settlement densities of coral recruits Comparison Average density Number of Statistics (number/panel) ±SE plates used Vertical/Horizontal 38.0±2.2/25.0± p<0.01, MWU Lower/Upper 20.5±3.8/2.7± p<0.01, WSR Ceramic/PVC 58.3±9.7/39.8± p=0.08, MWU 2-plate/11-plate 9.6±4.9/5.4± p=0.36, MWU (Horizontal) 2-plate/11-plate 54.7±4.0/38± p<0.01, MWU (Vertical) Facing outward/ Facing each other 2.7±1.2/43.9± p<0.01, WSR MWU: Mann-Whitney U Test; WSR: Wilcoxon Signed-rank Test 22

23 Table 3 Average settlement densities of pocilloporids on the lower surfaces of plates at different depths at two sites Depth (m) Number of Densities (no./panel) plates in each site Shiju Tiaoshipi ± ± ± ± ± ± ±0.2 0 p (Kruskal-Wallis Test)

24 Table 4. Nested-ANOVA table of pocilloporid recruitment in Hengchun Penninsula, Southern Taiwan. Scale of Area: 2-5 km, Site: 1 km, Location: 10 m. All the data was transformed by using log(n+1). (a) 1997/4/ /5/23 Levels d.f. SS F-value p Area Site <0.01 Location Residual (b) 1998/1/ /2/28 Levels d.f. SS F-value p Area Site <0.01 Location <0.01 Residual (c) 1999/4/ /6/27 Levels d.f. SS F-value p Area Site Location <0.01 Residual (d) 2000/4/2-2000/8/8 Level d.f. SS F-value p Area <0.01 Site Location <0.01 Residual

25 Table 5. Nested-ANOVA table of pocilloporid recruitment in Taiwan. Scale of Region: about 100 km, Area/Site: 1-5 km, Location: 10 m (1) 1999/4/ /6/27 Level d.f. SS F-value P Region Area/Site Location <0.01 Residual (2) 2000/4/2-2000/8/8 Level d.f. SS F-value P Region Area/Site <0.01 Location <0.01 Residual