Assessing the potential for recovery of a Sargassum siliquastrum community in Hong Kong Yu Hin Leung & Chung Wing Yeung & Put O. Ang Jr. Received: 26 May 2013 /Revised and accepted: 17 July 2013 /Published online: 8 September 2013 # Springer Science+Business Media Dordrecht 2013 Abstract Tung Ping Chau Marine Park in Hong Kong was designated in 2001 with aims to protect its highly-diverse marine habitats from destructive human activities. Sargassum spp. (mainly Sargassum siliquastrum) formed a dense canopy in Lung Lok Shiu (LLS) in southwestern side of the Marine Park. This Sargassum bed, however, disappeared and a barren ground developed 7 years after the marine park designation. The increase in density of the short-spined sea urchin Anthocidaris crassispina , and thus the grazing pressure, as a result of restriction on its commercial fishery within the Ma- rine Park was hypothesized as the immediate cause of the barren ground formation. In this study, we tested this hypoth- esis and manipulated the density of A. crassispina in exclu- sion cages underwater in LLS, where S. siliquastrum canopy originally existed. Transplantation of S. siliquastrum thalli and ceramic tiles with their recruits was carried out in these cages in FebruaryâMay 2012 and May 2013. The results showed that the transplanted Sargassum thalli could not sur- vive outside exclusion cages, whereas the induced recruits were almost wiped out in all tiles even inside exclusion cages. This provides evidence that the sea urchin A. crassispina could have overgrazed S. siliquastrum thalli and decimated the Sargassum canopy in LLS. It also suggested that other than A. crassispina , other grazers also contributed to the removal of S. siliquastrum recruits. Therefore, natural recov- ery of Sargassum bed seems unlikely if the grazing pressure remains high in this site. Proactive management strategies may need to be adapted to restore this Sargassum bed. Keywords Anthocidaris crassispina . Herbivory . HongKong . Phase-shift . Barren ground formation Introduction Hong Kong is located in the southern part of China where both tropical and temperate species can coexist. Tung Ping Chau, a remote island at the northeastern part of Hong Kong, supports a high biodiversity of marine organisms. Tung Ping Chau Marine Park (TPCMP) was, therefore, designated in 2001 to conserve its rich marine communities, including the high coverage of hard corals and the extensive macroalgal (mainly Sargassum) beds from any destructive human activ- ities. An extensive Sargassum bed could be found in Lung Lok Shui (LLS) on the southern side of the island. Various Sargassum spp. provide subtidal canopy to support highly- diverse zooplankton and epiphytic flora and fauna (Ng 2009). The abundance of Sargassum spp. is highly seasonal with different growth phases at different times of the year. The largest biomass and longest thallus length of Sargassum spp. were usually attained in winter, while most individuals would dieback with only the holdfast left on the substratum in summer (Ang 2006). At least eight species of Sargassum were found in LLS intertidal and subtidal waters. The algal cover in shallow (0â3 m chart datum (CD)) and deep (3â6 m CD) subtidal zones were considerably different (Ang 2006). For example, Sargassum hemiphyllum (Turner) C. Agardh and Sargassum henslowianum C. Agardh dominated in the shallow zone, while Sargassum siliquastrum (Mertens ex Turner) C. Agardh and Sargassum patens C. Agardh domi- nated in the deeper zone. However, starting from 2006, the Sargassum bed in LLS started to diminish. This coincided with a bloom in density of short-spined sea urchin Anthocidaris crassispina (A. Agas- siz), one of the major grazers of Sargassum spp. in Hong Kong (Ang et al. 2010). A barren ground was ultimately formed in LLS in 2008 and various Sargassum spp., including S. siliquastrum were no longer found since then, although patches of S. hemiphyllum individuals could still be found in the lower intertidal and shallow subtidal. Yu Hin Leung and Chung Wing Yeung equally contributed to this work. Y. H. Leung : C. W. Yeung : P. O. Ang Jr. (*) Marine Science Laboratory, School of Life Sciences, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong SAR, China e-mail:
[email protected] J Appl Phycol (2014) 26:1097â1106 DOI 10.1007/s10811-013-0097-1 Phase shift or state shift is the sudden drastic change in a dominant state to an alternative state of the environment (Scheffer et al. 2001), e.g., from coral-dominated state to dominance by macroalgae (Hughes 1994; McCook 1999; Scheffer et al. 2001; Hughes et al. 2007). Here, the designation of marine park was suspected to be an indirect cause of the phase shift from Sargassum bed to barren ground in LLS in recent years. Commercial fishery in TPCMP was prohibited, and thus the fishing pressure on A. crassispina was released in the area. Although this sea urchin was not the main target species for conservation, its density was found to increase significantly 5 years after legal protection of TPCMP took effect. Its density remains relatively high until now. Herbivory has been shown to affect the distribution of foliose algae such as Sargassum spp. (Fletcher 1987; McCook 1996; 1997), and large-scale grazing on macroalgae by sea urchin that led to urchin-barren formation is well documented in many places around the world (Watanabe and Harrold 1981; Vasquez et al. 1984; Himmelman and Nedelec 1990; Gagnon et al. 2004). In order to prove the causeâeffect relationship between sea urchin density and formation of barren ground, caging exper- iment was carried out in the present study to manipulate the sea urchin density in order to evaluate the grazing pressure of this sea urchin on S. siliquastrum , the main subtidal canopy spe- cies. The cages were set in the subtidal area in LLS previously dominated by S. siliquastrum . Both adult Sargassum thalli and Sargassum recruits on ceramic tiles were transplanted in ur- chin exclusion and other control cages to evaluate the specific role of A. crassispina . Furthermore, the possibility of using these transplantation strategies to facilitate the recovering of Sargassum bed in LLS was also investigated. Materials and methods TPCMP, at the northeastern part of Hong Kong (Fig. 1), is relatively remote and therefore, enjoys less anthropological disturbances. It supports a high biodiversity and is of high ecological value for conservation. LLS and Lung Lun Tsui at the southwestern side of the island used to have extensive Sargassum bed. After the outbreak of A. crassispina in 2006, both sites now have become barren ground (Ang et al. 2010). This study was carried out in LLS (22°540â²046â²â²N; 114°430â² 053â²â²E) in TPCMP. This site experiences strong wave action especially during summer monsoon. Since S. siliquastrum was no longer found in LLS, all of the Sargassum thalli used in this study were collected from Lo Fu Ngam (LFN) (22°356â² 684â²â²N; 114°278â²559â²â²E) in Sai Kung, Eastern Hong Kong (Fig. 1), less than 40 km from the study site. This is one of the closest sites where extensive S. siliquastrum bed is known to be found. Density and size of A. crassispina and S. siliquastrum in LLS The density and mean size of the short-spined sea urchin A. crassispina in LLS were recorded every month for 1 year from Aug. 2011 to Jul. 2012. Eight 20 m transects were laid haphazardly perpendicular to the shore at a depth of around â3 to â5 m CD. A 0.5 mÃ0.5 m quadrat was laid at every 2 m on the left hand side along each transect. All individuals of A. crassispina within the quadrat were counted, and test diame- ter of each was measured to the nearest millimeter using a vernier caliper. The density of A. crassispina was calculated by dividing the number of individuals over the total area of Fig. 1 Map of Hong Kong showing the location of Tung Ping Chau Marine Park (TPCMP) in NE Hong Kong and Sai Kung in eastern side of Hong Kong. Caging experiments were set-up in Lung Lok Shui (LLS) (marked by a cross in the right enclosed map) and Sargassum thalli were collected from Lo Fu Ngam (LFN) in Sai Kung (arrow in lower enclosed map) 1098 J Appl Phycol (2014) 26:1097â1106 quadrats laid along each transect. Density of Sargassum spp. was also monitored using the same transects and quadrats. Cage setup The experimental setups consisted of stainless steel cages (30 cmÃ30 cmÃ10 cm) with 1 cmÃ1 cm mesh size mounted by anchor blots on the sedimentary barren rock (mainly shale) underwater at about â3 m CD. The size of cages was kept small because of the strong waves in LLS. Five cage designs were used in the experiment as follows: (1) full cage with sea urchin completely excluded (full cage without urchin), (2) full cage with one randomly obtained sea urchin enclosed (full cage with urchin), (3) half cage with two sides opened (two- side open cage), (4) half cage with top opened (top-open cage), and (5) natural control, with no cage (natural control). Each cage setup had five replicates. The full cage without urchin was designed to provide a sea urchin-free area, and the other designs acted as controls to compare the cage effect while allowing entry of sea urchins. The full cage with urchin enclosed served to measure the grazing impact of sea urchin. The two-side open cage served as a control against current flow while the top-open cage served as the control against shading effect of the cage cover. These setups were used in the two sets of transplantation experiments and the recruitment experiment to be explained below. Adult transplantation experiment The transplantation experiments on adult S. siliquastrum were carried out twice using the cage setups in LLS. The first experiment lasted for 4 weeks in February 2012, while the second experiment was conducted from mid-March to early May 2012 and lasted for 6 weeks. More than 170 individuals of adult S. siliquastrum with holdfast were collected from LFN in January 2012 (for the first experiment) and in Febru- ary 2012 (for the second experiment). They were first maintained in the outdoor flow-through culture tank in the Marine Science Laboratory of the Chinese University of Hong Kong. The collected individuals were trimmed before being transplanted to LLS to avoid overcrowding inside the cages. In the first experiment, all collected thalli were trimmed to 100 mm in length. Six thalli were placed in each replicate of each cage setup. Therefore, a total of 150 thalli were transplanted into the field. For the second experiment, all thalli of Sargassum were similarly trimmed to 100 mm in length, but at the same time, the biomass of each thallus was also trimmed down to 6 g (wet weight). The trimmed thalli in both experiments were tied on nylon ropes and then fixed inside each cage. Thereafter, the maximum length of each thallus was measured to the nearest millimeter once every 1â2 weeks. The number and sizes of the sea urchin A. crassispina en- countered in the cage setups were also recorded andmeasured. Recruitment experiment Reproductive S. siliquastrum thalli were collected from LFN inmid-February 2012. Theywere cultured in an outdoor flow- through culture tank in the laboratory. Thirty-five ceramic tiles with rough surface were placed below the plants to allow S. siliquastrum germlings to settle on them. The thalli were removed after 14 days, and the new recruits on the tiles were allowed to grow in the tank for four more weeks. They were moved to a closed-system water table for another 4 weeks to allow the recruits to stabilize. The number of established recruits on each tile was then counted under the dissecting microscope and recorded. Twenty-five tiles with recruits in their best condition were chosen, transferred to the field, and mounted in the same cage setups in LLS. The tiles were collected back and the number of surviving recruits on each tile was counted. This experiment was carried twice, once in 2012 and repeated in 2013. Statistical analyses All analyses were carried out using SPSS 16.0 with significant value set at p =0.05. For the adult transplantation experiment, each cage served as a replicate, so the maximum lengths of all Sargassum thalli within each cage were measured and aver- aged. Two-way analysis of variance (ANOVA) with repeated measures was then used to detect any significant differences in Sargassum thallus length under the different cage setups over time. One-way ANOVA was further used to compare differ- ences in Sargassum thallus length under the different cage setups at each time point in order to find out the critical time at which differences in thallus length started to appear. Should significant differences be detected, post hoc test was carried out to find the significant groupings. For the recruitment experiment, the percentages of Sargassum recruits lost in different cage designs were compared to each other with one-way ANOVA to see if any significant difference existed between cage designs. Results Density and size of A. crassispina and Sargassum spp. in LLS The mean (± SD) density of A. crassispina in LLS peaked at 8.5±5.4 ind.mâ2 in August 2011 and dropped to 4.75±2.1 ind.mâ2 in July 2012, with a mean density of 6.56±1.21 ind. mâ2 throughout the year of monitoring (Fig. 2a). This is consistent with the earlier study (Ang et al. 2010) showing the density of A. crassispina to stay above 4.0 ind.mâ2 since 2006 (Fig. 2b). Furthermore, mean (±SD) test diameter of A. crassispina was comparable throughout the monitoring peri- od at 30.92±1.24 mm. No Sargassum plant was found within J Appl Phycol (2014) 26:1097â1106 1099 the quadrats throughout this monitoring period, and this has been the case since 2008 (Fig. 2b). Adult transplantation experiment In the first adult transplantation experiment, Sargassum thalli in the full cages without sea urchin did not show significant drop in their length over time, whereas those in other cage setups all showed different degrees of decline which are statistically significant (two-way ANOVAwith repeated mea- sures, p HSD test, p =0.275). The mean maximum thallus length of S. siliquastrum in the first and second experiments exhibited different growth patterns. The increasing trend of thallus length in full and top-open cages in the second experiment was not observed in the first experiment. Biomass change was not considered in the first experiment. In the second experiment, the pattern of percentage change in wet weight of the transplanted Sargassum (Fig. 4) was shown to be similar to that observed for thallus length. Sargassum thalli in full cage without urchin and in top-open cage exhibited a positive increase in their biomass, whereas those in the other three cage setups all showed a decrease in their biomass. By comparing the percentage change in thallus length and that in wet weight in different cage setups, those in full cages without urchins were shown to be significantly different (independent sample t test, p =0.005). The absence of grazing pressure from sea urchins probably allowed the thallus to grow more lateral branches and more healthily and led to a higher biomass-to-length ratio of the Sargassum thalli in these full cages without urchin than in the other cage setups. Table 1 shows the number of sea urchins encountered in the cage areas of each type of cage setup in both experiments. In full cages with and without urchin, the number of urchins was always one and zero, respectively, indicating that exclusion Fig. 3 a Results of the first adult transplantation experiment showing decrease in mean (±SD) maximum thallus length over time in all experimental cages except full cages without urchins. However, only those in full cages and two-side open cages showed significant difference at week 4 (marked with asterisk). b Results of the second adult transplantation experiment showing changes in Sargassum mean maximum thallus length were significantly different among cage setups after week 3. In week 6, cage setups with Sargassum thallus lengths which are not statistically significantly different are indicated with the same letter (one-way ANOVA, Tukey HSD test p >0.05) J Appl Phycol (2014) 26:1097â1106 1101 and enclosure of sea urchin were successful. On the other hand, A. crassispina urchins had free access to the other three cage setups, and the average number of urchins in all these cage setups was mostly higher than one. But the average numbers of urchins in these three cage setups in the second experiment were all lower than those found in the first experiment. A direct positive relationship between sea urchin density and loss in Sargassum thallus length per week can be obtained (Fig. 5). The higher the density of sea urchins, the greater is the rate of loss in Sargassum thallus length. Recruitment experiment There was a large variation in the number of germlings that settled on the recruitment tiles, and hence the number of recruits that eventually stabilized on the tile surface also varied significantly among tiles (ranged from 65 to 972 recruits per tile). The tiles were grouped in five sets for the five cage setups with five replicates each. The mean number of recruits per set was kept comparable with the mean (±SD) density of recruits being 348.6±40 recruits per set. These were then laid out in the field in different cage setups on May 15, 2012. The Fig. 4 The mean (+SD) percentage change in wet weight of transplanted thalli after 6 weeks in the second adult transplantation experiment. Percentage changes in cage setups that are not significantly different are marked with the same letter Fig. 5 The relationship between loss in thallus length per week and the number of sea urchin A. crassispina found in the cage as expressed in a regression analysis y=â3.2778xâ7.8817, R2=0.7764, n=8. The last data point represents data for mean sea urchin number that ranged from 9 to 12 1102 J Appl Phycol (2014) 26:1097â1106 mortality (lost) rate was very high within the first week (May 21, 2012) and ranged from 88.5% in natural control to 98.4 % in top-open cages (Fig. 6), with no significant differences among cage setups. Among individual tiles, the mortality (lost) rate ranged from 70.4 to 100 %. The experiment was repeated in 2013. The tiles were again grouped in five sets with three replicates each. The number of recruits per set was also kept comparable with the mean (± SD) density being 52.5 ±13.4 recruits per set. The tiles were laid out in their respec- tive cage setups on May 8, 2013, and within the first week, all recruits (100 %) were lost (data not shown). Discussion In the caging experiment on adult thalli, full cages with sea urchin enclosed provided very clear evidence that sea urchins pose significant grazing impact on Sargassum thalli. Mean thallus lengths in full cages with urchin were significantly smaller than those in full cages without urchin (week 5: one- way ANOVA, p Meanwhile, the densities of the sea urchin found inside the cages were actually lower than those found naturally in LLS. Sea urchins in the natural environment tend to be patchy in their distribution, with many of them clustering in holes and crevices (Nelson and Vance 1979; Linga et al. 2010). The difference in growth pattern of Sargassum observed between the two experiments may partly be explained by seasonal effects. No drifting green algae Ulva spp. were observed during the first experimental period, but plenty of them were found drifting around the experimental area during the second experimental period. Algae in Hong Kong waters, including Ulva spp., showed high seasonal variation in their abundance (Hodgkiss 1984; So 2005; Wong 2005). Their growth is correlated with seasonal water temperature change. They were in high abundance during the second experimental period but not during the first. The drifting thalli of Ulva around the cage area were therefore likely to have come from the shallow subtidal or lower intertidal area in LLS or nearby areas during the second experimental period. Some of them even drifted into the cages, and individuals of A. crassispina were found to be grazing on them. Although sea urchins may not be selective on their food (Taylor and Steinbergh 2005), the presence of Ulva spp. in LLS, especially in areas around the cage setups, could have reduced the grazing pressure on transplanted Sargassum , especially in the top-open cages and natural controls. Transplanted Sargassum in full cages, which were completely isolated from sea urchins, also showed different growth patterns between the two experiments. That being the case, there must be other factors affecting Sargassum growth. Sargassum spp. usually show a cycle of growth phases which is regulated by seasonal change in seawater temperature (Arenas and Fernández 2000; Rao 2002; Ang 2006). Since the two experiments were carried out in different months (first one started in January and the second one started in late March), Sargassum thalli were likely to be at different phases of their growth (Ang 2006) with the mean seawater tempera- ture recorded in TPCMP in March and April (16.9 and 21.2 °C, respectively) being higher than that in February (15.9 °C). This temperature difference, however, did not ap- pear to have significant effect on sea urchin activities. Higher density of urchins clearly exerted a higher grazing pressure on Sargassum thalli (Fig. 5). This is clearly shown in the regression analysis. This regression analysis utilizes mean urchin density data from all cage setups and assumes that the number of sea urchins found in each cage at the time of measurement represented the number of sea urchins staying in the cage area in between sampling periods. Any loss in algal thalli could, therefore, be attributed to the grazing impact exerted by this number of urchins. It further assumes that different sea urchin individuals, irrespective of their sizes, exert the same grazing pressure on Sargassum thallus, and that the grazing rate was the same throughout the experimental period. The designs of different cage setups did not affect the grazing pressure, as evidenced by the comparable rate of loss of Sargassum thalli in different cage setups having compara- ble density of sea urchins. The mean (± SD) density of A. crassispina in LLS recorded in the present study remained high at 6.56±1.21 ind.mâ2. Between 2000 and 2004, when high coverage of S. siliquastrum was still found in LLS, the mean urchin density was 3.9 ind.mâ2. The mean density doubled to 8.7 ind.mâ2 between 2006 and 2007when S. siliquastrum was observed to be overgrazed (Ang et al. 2010) (see also Fig. 2b). This dropped a bit to 6.4 ind.mâ2 from 2008 onwards after the S. siliquastrum community completely disappeared (Fig. 2b). Although monitoring of A. crassispina density was not con- tinuous in the last 12 years, nonetheless, the change in mean urchin density in the three sampling periods (i.e., 2000 to 2004, 2006 to 2007, and 2008 to 2012) was statistically significant (one-way ANOVA, p 80 to 100 %) of Sargassum recruits in LLS even under full cages, where sea urchins were excluded. Sea urchins were not the only grazers in LLS. Other gastropods and hermit crabs could also graze heavily on algae (So et al. 2007) and could be the major factor leading to the high mortality of Sargassum recruits. Our cage setups were not designed to exclude these smaller grazers. In other natural macroalgae communities, algal recruits are protected from 1104 J Appl Phycol (2014) 26:1097â1106 grazers and wave actions by nearby adult thalli (Dayton 1985; Gagnon et al. 2003). This protection is obviously absent in LLS, with adult Sargassum thalli being long overgrazed by sea urchins. Macroalgal recruits may also settle between bar- nacles, mussels, and other sessile organisms that provided refuges, hence survivorship of algal recruits induced to settle on recruitment tiles might be increased when tiles with more rugged surface were used (Norton 1983; Hixon and Brostoff 1985; Diaz-Pulido and McCook 2004). However, given that the natural rock (mainly shale) surfaces in LLS are very smooth, use of rugged surface recruitment tiles would not reflect the natural situation in the present study site. What can be concluded from our present study is that sea urchins can readily graze down adult Sargassum in LLS and only thalli in exclusion cages can be spared. A longer-term (at least a year) study needs to be set up to evaluate the fate of these thalli to see if they could become reproductive and could persist into the following growth season. On the other hand, survival of Sargassum new recruits is extremely low in a barren ground. Young recruits need to face not only grazing pressures from sea urchins but also from other smaller grazers. The latter may in fact be more important than the sea urchins. Throughout the present sampling period, no new natural Sar- gassum recruit was observed in LLS. The absence of a supply source of Sargassum germlings may partly explain for this. But given the high mortality of Sargassum recruits, it would be difficult to re-establish the Sargassum bed even by artificial seeding of the natural substrata with young recruits. TPCMP was set up in 2001 in order to protect the high biodiversity in the marine park. The decimation of its Sargas- sum community in LLS ran contrary to this objective. It provides a good lesson to be learned about setting up of marine protected areas in a highly disturbed locality like Hong Kong. Exploitative fishery has often been shown to decrease the abundance of urchin predators and resulted in the forma- tion of sea urchin barrens (Jackson et al. 2001; Bonaviri et al. 2009). Therefore, setting up of marine protected area could reverse this trend. We do not have any previous information on sea urchin predator in LLS, so whether earlier populations of sea urchins were regulated through top-down control is not known. Nonetheless, the earlier exploitative fishery was on the urchins themselves, so harvesting pressure from human may have served the role of the top-down predators. With the release of this pressure after the establishment of TPC as a marine park, this protection did not lead to an increase in other top predators, notably those that could prey on sea urchins such as fish (Lau et al. 2011) and sea stars (Yatsuya and Nakahara 2004). With sea urchin density remaining high, it seems unlikely that the Sargassum bed would be able to recover on its own nor would transplantation of adult plants or artificial seeding of new recruits be a practical option to facilitate its recovery. There is no sign to indicate the appear- ance of urchin top predators either. With their high abundance in Hong Kong waters, sea urchins were never a target for protection. Allowing sea urchin harvest in LLS and the sur- rounding areas may be the only option left to restore the Sargassum community in LLS. There is, however, no guar- antee of Sargassum recovery even in the absence of sea urchins, as evidenced by the difficulty for Sargassum recruits to establish. Additional restoration strategies using transplan- tation of adult plants may need to be put in place to serve as the seed stock. The government and other stakeholders should consider seriously the objective of establishing the protected area in order to adapt the most effective and proactive ap- proach for its restoration and sustenance. Acknowledgments This study was partly supported by Research Grant Council General Research Fund 460010 to PO Ang. We thank Hong KongAgriculture, Fisheries and Conservation Department for permission to work in TPCMP. References Ang PO (2000) Relative importance of recruitment versus regeneration in maintaining the populations of Sargassum henslowianum in Ping Chau, Hong Kong. J Phycol 36(Suppl):2â3 Ang PO (2006) Phenology of Sargassum spp. in Tung Ping ChauMarine Park, Hong Kong SAR, China. J Appl Phycol 18:629â636 Ang PO, So KY, Li TW, Yeung FF, Wong SY, Woo CK (2010) Balance between harvesting pressure and protection: Implications for barren ground formation in Hong Kong. Bull Fish Res Agency 32:1â10 Arenas F, Fernández C (2000) Size structure and dynamics in a popula- tion of Sargassum muticum (Phaeophyceae). J Phycol 36:1012â 1020 Bonaviri C, Fernández TV, Badalamenti F, Gianguzza P, Di Lorenzo M, Riggio S (2009) Fish versus starfish predation in controlling sea urchin populations in Mediterranean rocky shores. Mar Ecol Prog Ser 382:129â138 Chiu ST (1985) Feeding biology of the short-spined sea urchin Anthocidaris crassispina (A. Agassiz) in Hong Kong. In: Keegan BF, OâConnor BDS (eds) Proceedings of the Fifth International Echinoderm Conference. Balkema, Boston, pp 223â232 Dayton PK (1985) Ecology of kelp communities. Annu Rev Ecol Syst 16:215â245 Diaz-Pulido G, McCook LJ (2004) Effects of live coral, epilithic algal communities and substrate typeon algal recruitment. Coral Reefs 23:225â233 Fletcher WJ (1987) Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecol Monogr 57:89â109 Gagnon P, Himmelman JH, Johnson LE (2004) Temporal variation in community interfaces: kelp-bed boundary dynamics adjacent to persistent urchin barrens. Mar Biol 144:1191â1203 Gagnon P, Himmelman JH, Johnson LE (2003) Algal colonization in urchin barrens: defense by association during recruitment of the brown alga Agarum cribrosum . J Exp Mar Biol Ecol 290:179â196 Himmelman JH, Nedelec H (1990) Urchin foraging and algal survival strategies in intensely grazed communities in eastern Canada. Can J Fish Aquat Sci 47:1011â1026 Hixon MA, Brostoff WN (1985) Substrate characteristics, fish grazing, and epibenthic reef assemblages of Hawaii. Bull Mar Sci 37:200â 213 J Appl Phycol (2014) 26:1097â1106 1105 Hodgkiss IJ (1984) Seasonal patterns of intertidal algal distribution in Hong Kong. Asian Mar Biol 1:49â57 Hughes TP (1994) Catastrophes, phase shifts, and large-scale degradation of a Caribbean coral reef. Science 265:1547â1551 Hughes TP, Rodrigues MJ, Bellwood DR, Ceccarelli DM, Hoegh- Guldberg O,McCook L,Moltschaniwskyj N, Pratchett MS, Steneck RS, Willis B (2007) Phase shifts, herbivory, and the resilience of coral reefs to climate change. Curr Biol 17:1â6 Jackson JBC, Kirby MX, Berger WH, Bjorndal KA, Botsford LW, Bourque BJ, Bradbury RH, Cooke R, Erlandson J, Estes JA, Hughes TP, Kidwell S (2001) Historical overfishing and the recent collapse of coastal ecosystems. Science 293:629â637 Kendrick GA, Walker DI (1994) Role of recruitment in structuring beds of Sargassum spp. (Phaeophyta) at Rottnest Island, Western Aus- tralia. J Phycol 30:200â208 Lau DCC, Dumont CP, Lui GCS, Qiu J (2011) Effectiveness of a small marine reserve in Southern China in protecting the harvested sea urchin Anthocidaris crassispina : A mark-and-recapture study. Biol Conserv 144:2674â2683 Linga SD, Ibbottb S, Sandersona JC (2010) Recovery of canopy-forming macroalgae following removal of the enigmatic grazing sea urchin Heliocidaris erythrogramma. J Exp Mar Biol Ecol 395:135â146 McCook LJ (1996) Effects of herbivores and water quality on Sargassum distribution on the central Great Barrier Reef: cross-shelf trans- plants. Mar Ecol Prog Ser 139:179â192 McCook LJ (1997) Effects of herbivory on zonation of Sargassum spp. within fringing reefs of the central Great Barrier Reef. Mar Biol 129:713â722 McCook LJ (1999) Macroalgae, nutrients and phase shifts on coral reefs: scientific issues andmanagement consequences for the Great Barrier Reef. Coral Reefs 18:357â367 Nelson BV, Vance RR (1979) Diel foraging patterns of the sea urchin Centrostephanus coronatus as a predator avoidance strategy. Mar Biol 51:251â258 Ng KY (2009) The faunal assemblage structure in the seaweed bed of Sargassum siliquastrum. The Chinese University of Hong Kong, Hong Kong, M.Phil Thesis Norton TA (1983) The resistance to dislodgement of Sargassummuticum germlings under defined hydrodynamic conditions. J Mar Biol Assoc UK 63:181â193 Rao AS (2002) Seasonal growth pattern in Sargassum polycystum C. Agardh (Phaeophyta, Fucales) occurring at Visakhapatnam, east coast of India. Indian J Mar Sci 31:26â32 Scheffer M, Carpenter SR, Foley JA, Folke C, Walker B (2001) Cata- strophic shifts in ecosystems. Nature 413:591â596 So KY (2005) Algal-herbivore interactions in coastal communities in Tung Ping Chau. The Chinese University of Hong Kong, Hong Kong, Hong Kong. M.Phil Thesis So KY, Yeung FF, Ang PO (2007) Different scales of herbivory as a structuring force in subtidal rocky shore in Hong Kong. 19th Inter- national Seaweed Symposium Japan, Mini-symposium 19, 2007 Taylor RB, Steinbergh PD (2005) Host use by Australasian seaweed messograzers in relation to feeding preferences of larger grazers. Ecology 86:2955â2967 Vasquez JA, Castilla JC, Santelices B (1984) Distributional patterns and diets of four species of sea urchins in giant kelp forest (Macrocystis pyrifera) of Puerto Toro, Navarino Island, Chile. Mar Ecol Prog Ser 19:55â63 Wai TC (2012) Stock and ecological status of sea urchins in Hong Kong: evaluation of the effectiveness of marine protected areas using sea urchins as model organisms. Retrieved from http://www.ecf.gov.hk/ en/approved/200704.html on 30 July 2012 Watanabe JM, Harrold C (1981) Destructive grazing by sea urchins Strongylocentrotus spp. in a central California kelp forest: Potential roles for recruitment, depth, and predation. Mar Ecol Prog Ser 71:125â141 Wong TC (2005) Spatial and temporal distribution of heavy metals in Hong Kong seaweeds with an analysis on the effect of heavy metals on the reproduction of the green algaUlva lactuca . PhD Thesis, The Chinese University of Hong Kong, Hong Kong Yatsuya K, Nakahara H (2004) Density, growth and reproduction of the sea urchin Anthocidaris crassispina (A. Agassiz) in two different adjacent habitats, the Sargassum area and Corallina area. Fish Sci 70:233â240 1106 J Appl Phycol (2014) 26:1097â1106 http://www.ecf.gov.hk/en/approved/200704.html http://www.ecf.gov.hk/en/approved/200704.html Assessing the potential for recovery of a Sargassum siliquastrum community in Hong Kong Abstract Introduction Materials and methods Density and size of A. crassispina and S. siliquastrum in LLS Cage setup Adult transplantation experiment Recruitment experiment Statistical analyses Results Density and size of A. crassispina and Sargassum spp. in LLS Adult transplantation experiment Recruitment experiment Discussion References