3-1 Introduction
Coral reefs are often recognized as one of the most diverse marine ecosystems with high ecological and economic values (Connell 1978; Knowlton 2001). In recent decades, coral reef systems worldwide are declined rapidly and dramatically by natural and anthropogenic disturbances (Knowlton 2001; Hughes et al. 2003; Wilkinson 2004), such as strong tropical cyclones or hurricanes (Hughes 1994, 1996; Connell et al. 1997;
Hughes and Connell 1999; Guillemot et al. 2010), mass bleaching and the population outbreak of the crown-of-thorns starfish. Recent studies have suggested that global warming could increase the frequency and intensity of strong hurricanes (Webster et al.
2005; Elsner et al. 2006a, b). If the frequency and intensity of tropical cyclones and typhoons increase as the hurricanes in Atlantic, there may be no enough time for coral assemblages to recovery under such frequent disturbances (Connell 1997; Hughes and Connell 1999). In such a context, it is an important issue to better understand the dynamics of coral reef ecosystems under increasing threats of climate change and to
storms have been shown to cause severe damage and mortality to coral colonies (Connell 1997; Connell et al. 1997). Tropical storms may also have negative effects on coral recruitment and may result in rapid changes of coral communities (Hughes and Connell 1999; Crabbe et al. 2002; Mallela and Crabbe 2009). The recurrence of tropical storms has been shown to be responsible for the change of coral communities and result in the dominance of morphologically resistant species such as massive corals (Connell et al. 1997; Hughes and Connell 1999; Loya et al. 2001; Guillemot et al. 2010). For example, coral reefs of Rio Bueno, northern Jamaica have been struck by Hurricane Allen in 1980, the covers of two branching species, Acropora palmata and A.
cervicorinis which accounted for half or more of the total cover, dropped from 30% to
nearly zero, while massive species, Colpophyllia natans and Montastrea annularis remained virtually unchanged. By 1993, branching Acropora remained rare while C.
natans and M. annularis accounted for 40% of the total cover (Hughes and Connell
1999).
A number of long-term studies have been conducted to assess the impacts of tropical storms on coral reefs and the following trajectory of recovery (e.g., Connell et al. 1997;
Hughes and Connell 1999; Guillemot et al. 2010). However, long-term studies on the effects of typhoons on marginal coral communities are scarce. Similarly, selective effect of physical disturbances on coral species with different growth forms in subtropical
areas have been reported (Nozawa et al. 2008; Tam and Ang 2008), but its long-term effects have rarely been studies. Such a study is useful to better assess the fate of subtropical coral communities under increasing threat of climate change.
In this study, the data of a long-term monitoring program in 6 years including pre- and post-typhoons was presented. The objectives are (1) to examine the impacts of typhoons on marginal coral communities; (2) to assess the responses of different coral species under the impact of typhoons; and (3) to evaluate the recovery of coral communities in northern Taiwan.
3-2 Materials and methods
The study area and sampling methods have been described in Chapter 2.
Data analysis
In order to determine the temporal changes of coral cover at each site, the differences of coral cover among surveys and the trend of changes were evaluated by
communities in 12 surveys. In addition, all coral colonies in each quadrat were classified into four size classes: < 50 cm2, 50-100 cm2, 100-250 cm2, > 250 cm2. Differences between size classes and temporal trend of each size class at each site among censuses were also analyzed using repeated measures ANOVA and post-hoc pair comparisons to assess the temporal patterns of each size class.
3-3 Results
Changes of coral cover at three reefs
The changes of coral covers at three reefs showed a similar trend from August 2003 to September 2009 (Fig. 3.1). The coral cover at Sites A, B, and C declined from 23.9 ± 6.7 (mean ± SE), 16.8 ± 1.5, 10.0 ± 1.1% in August 2003 to 9.1 ± 0.9, 8.3 ± 1.0, 2.2 ± 0.3% in July 2007, then increased gradually to 14.1 ± 0.5, 13.2 ± 1.2, and 2.9
± 0.5% in September 2009, respectively (Fig. 3.1). Changes of coral cover among 12 surveys were highly significant at Sites B and C, but not significant at Site A, mainly due to the large variations among quadrats (Table 3.1). The decline of coral cover was the greatest during the period from July to November 2005, and there were three major typhoons (namely Haitan, Talim, and Longwang, that hit the study area on July 18, August 31, and October 1, successively, data from the Central Weather Bureau, Taiwan.
http://rdc28.cwb.gov.tw/data.php) in summer 2005 (Table 3.2). The trends of decline in
coral cover were similar at three sites, although that at Site A was not significant (Table 3.1). In addition, the recoveries of coral cover between July 2007 and September 2009were significant at Sites A and B (Table 3.1), but not significant at Site C.
Changes of species composition of coral communities
Non-metric multidimensional scaling (nMDS) analysis of coral communities among 12 consecutive surveys showed distinct grouping (Fig. 3.2). Three periods could be identified in the course of this study, i.e., August 2003-July 2005, November 2005-September 2006, and July 2007-September 2009. During these periods, the changes of coral cover varied among genera and resulted in conspicuous changes in taxonomic composition of coral communities (Fig. 3.3). Coral genera with major changes included encrusting Montipora which decreased by 95.8-96.6% at three sites and foliaceous Pachyseris which was totally disappeared at Sites A and C during the study period. Besides, the branching Stylophora at Sites A and B decreased by 71.0-71.2% after the typhoon disturbances in 2005, however, it recovered to 41.1% and
Partial and whole colony mortality varied greatly among coral genera after the remarkable typhoon disturbances in 2005 (Fig. 3.4). Branching, foliaceous and encrusting species (Stylophora, Pachyseris, and Montipora) suffered higher whole mortality rates (37-67%). In contrast, massive faviids (Favia, Favites, Montastrea,
Platygyra, and Cyphastrea) suffered much lower whole mortalities (0-42%), but higher
partial mortalities (0-69%) following typhoon disturbances. There was a significant difference of living coral cover among growth forms throughout the survey period, especially those coral with susceptible forms. The covers of most corals in foliaceous and encrusting forms decreased to less than 1% after the disturbances (Fig. 3.5). In contrast, massive corals remained abundant at most study sites except at Site C. Hence, a notable change of community structure after typhoon disturbances was detected.
Temporal variation of colony size class
During the study period, differences among four size classes of hard coral colonies were significant at three sites (Fig. 3.6, Table 3.3). Moreover, all size classes of coral colonies showed similar trends at three reefs except for the smallest size (< 50 cm2).
Between 2003 and 2009, the colony number of smallest colony size (< 50 cm2) at Sites A and B (Trend Analysis, p < 0.05) increased gradually, meanwhile, the smallest colonies decreased in the beginning, then increased at Site C (Trend Analysis, p < 0.01).
The proportion of coral colonies of smallest colony size ranged from 40.7%~51.1% in August 2003 to 57.1%~68.1% in September 2009. Meanwhile, the proportion of largest coral colonies (> 250 cm2) decreased from 13.3%~8.5% in August 2003 to 4.6%~2.4%
in September 2009 (Table 3.3, trend Analysis, p = 0.187, p < 0.05, p < 0.05). The proportion of median colony sizes (50~100, 100~250 cm2) increased gradually at Sites A and B, but decreased in the beginning then increased slightly at Site C. Although most of the largest size colonies suffered whole or partial mortalities at all sites after typhoon disturbances in 2005, the number of coral colonies at Sites A and B increased mainly due to fragmentation of colonies. Nevertheless, at Site C, total colonies of all size classes decreased 41.4% even though the percentage of small colonies was increasing. This indicated that most coral colonies at Site C suffered partial mortality during this period.
3-4 Discussion
Effects of typhoon disturbances on coral communities
scleractinian corals were likely due to the damage of strong waves and heavy sedimentation associated with typhoon disturbances. Tropical storms (cyclone, hurricane, and typhoon) have been considered one of the large scale disturbances that may cause extensive and severe impacts on coral reefs (Moran et al. 1988; Birkeland and Lucas 1990; Connell 1997; Connell et al. 1997; Hoegh-Guldberg 1999; Wilkinson 1999; Hughes et al. 2003; Osborne et al. 2011). In addition, the damage induced by strong waves of storms seems to be the most common cause of coral mortality in high latitude reefs (Harriott and Smith 2000; Nozawa et al. 2008). Since there was no other major disturbance during the same period, the typhoon disturbances were considered as the major factor responsible for changes of coral communities (Harriott and Banks 2002;
Harriott and Smith 2000; Nozawa et al. 2008).
Changes of species composition of coral communities
The results of nMDS analysis showed distinct temporal variations of coral communities among 12 consecutive surveys (Fig. 3.2) and the results highlighted the variability of typhoon disturbances among coral species in Yenliao Bay. The composition of coral communities shifted from a combination of branching, foliaceous, encrusting and massive corals in 2003 to those dominated by massive corals in 2009 (Fig. 3.5). This change was consistent with the observations worldwide (Hughes and
Connell 1999) and could be attributed to the selective effects of typhoons on different growth forms. The growth forms of coral colonies are thought to be an important character for their resistance and survival under the impacts of severe disturbances such as recurrent typhoons or high seawater temperatures (Connell et al. 1997; Hughes and Connell 1999; Loya et al. 2001; Guillemot et al. 2010). For example, at Solitary Islands, Australia, a cyclone in 1976 caused a decrease of > 40% in delicate tabular corals, while massive and encrusting corals showed little change (Connell et al. 1997). At Sesoko Island, Japan, branching corals (mostly Acropora and Pocillopora) were more susceptible to bleaching induced by higher seawater temperatures, while massive and encrusting corals (mainly faviids and poritids) survived (Loya et al. 2001). These evidences showed that physical disturbances caused significantly damage to more vulnerable forms and favored resistant massive corals (Harriott and Smith 2000;
Nozawa et al. 2008).
In Yenliao Bay, differential susceptibility of corals to typhoon disturbances was clearly linked to colony morphology. During the major typhoon disturbances in 2005,
stable trajectories following typhoon disturbances. Unlike other high latitude areas, most of the reduction of coral cover in Yenliao Bay was attributable to the decrease of encrusting species (e.g., Montipora), rather than branching species Acropora and
Pocillopora in southwest Japan (Nozawa et al. 2008) and Southern Australia (Harriott
and Smith 2000). The low abundance of branching corals at Yenliao Bay may be due to frequent impacts of strong wave actions generated by typhoons and monsoon that constraint the growth of branching corals and favor the growth of corals with compact forms (Veron 1993; Harriott 1999a). This morphological selection is a feature of disturbances, a large number of studies have provided evidences of selective mortality among corals induced by tropical storms (e.g., Woodley et al. 1981; Hughes and Connell 1999; Guillemot et al. 2010). If the frequency and intensity of storm disturbances increase as projected (Webster et al. 2005; Elsner et al. 2006a, b), a shift of community structure to the dominance of more resistant massive corals may occur on subtropical or marginal coral communities.
Recovery of the coral communities
From 2007 to 2009, the cover of scleractinian corals increased gradually and varied among sites in Yenliao Bay. At Sites A and B, coral cover showed significant increase (Table 3.1, Repeated measures ANOVA, pairwise comparison: p < 0.05, p < 0.01
respectively), while Site C remained almost no change throughout this period. Since Site C was close to the coastline and facing the input terrestrial sediment from Shiding Stream, the lack of recovery at Site C might be attributable to heavy sedimentation that impeded coral growth (Nugues and Roberts 2003; Wilkinson 2000). In addition, the competition with macroalgae might also constrain or limit the growth and recruitment of corals (Nozawa et al. 2008; Nugues and Roberts 2003; van Woesik et al. 1995). The nutrients enriched water from Shiding River seems to cause the bloom of macroalgae at Site C and resulted in extensive overgrowth of coral colonies.
The recovery of coral communities could be verified by comparing the trajectory of different colony sizes. In the study period, all large coral colonies (> 250 cm2) at Sites A and B decreased (Fig. 3.6) and showed no significant change (Fig 3.6, Table 3.3), while small colonies (< 50 cm2) increased remarkably after the typhoon disturbances in 2005 (p > 0.05). The mortality of large colonies by strong wave actions is likely a limiting factor for the development of coral communities and reef accretion (Grigg 1998;
Harriott and Smith 2000; Nozawa et al. 2008). In the meantime, the number of median
contrast, even though small colonies (< 50 cm2) were increasing after 2005 at Site C, most of the largest size colonies suffered whole or partial colony mortality and median size colonies decreased significantly. Besides, the total number of colonies of all size classes decreased as well. Therefore, the coral community at Site C did not show any sign of recovery.
In conclusion, significant declines of coral cover and changes of species composition of coral communities were detected at three reefs in Yenliao Bay from 2003 to 2009. These changes were resulted from differential mortality and contrasting trajectories of different coral genera during and after the typhoon disturbances. It is also noted that the recovery of coral communities might be impeded by localized stresses such as those at Site C and this is likely responsible for the spatial heterogeneity of coral communities in this subtropical environment.
Table 3.1 Results of the repeated measures ANOVAs, Trend analysis, and Pairwise comparisons for detecting the changes of coral cover among 12 surveys at the three sites.
Site A B C
Coral coverage change among surveys n.s. (p = 0.087) p < 0.001 *** p < 0.001 ***
Trend analysis of coral coverage n.s. (p = 0.076, quadratic) p < 0.05 * (quadratic) p < 0.01 ** (quadratic)
Pairwise comparisons
August 2003, July 2007 n.s. (p = 0.128) p < 0.01 ** p < 0.01 **
July 2007, September 2009 p < 0.05 * p < 0.01 ** n.s. (p = 0.135)
Levels of significance (p values) for the repeated measures ANOVAs and trend analysis: n.s.: not significant, * P < 0.05, ** P < 0.01, *** P <
0.001
Table 3.2 A list of 14 typhoons passed or close to Yen-laio Bay during study years. Sources: The Central Weather Bureau, Taiwan.
Date Name Maximum gust speed
near the center (m/s)
Table 3.3 Results of the repeated measures ANOVAs, Trend analysis, and Pairwise comparisons for detecting the changes of coral cover among 12 surveys at the three sites.
Site A B C Levels of significance (p values) for the repeated measures ANOVAs and trend analysis: n.s.: not significant, * P < 0.05, ** P < 0.01, *** P <
0.001
Fig. 3.1 Changes of hard coral cover (mean±SE) at three sites from 2003 to 2009.
Arrows indicate the major typhoon events and the arrow size represents the relative intensity of typhoon disturbances.
Site A
Aug_03 Apr_04 Jan_05 Jul_05 Nov_05 May_06 Sep_06 Jul_07 Jun_08 Sep_08 May_09 Sep_09
Pachyseris
Aug_03 Apr_04 Jan_05 Jul_05 Nov_05 May_06 Sep_06 Jul_07 Jun_08 Sep_08 May_09 Sep_09
Hydnophora
Site C
0 5 10
Aug_03 Apr_04 Jan_05 Jul_05 Nov_05 May_06 Sep_06 Jul_07 Jun_08 Sep_08 May_09 Sep_09
Goniastrea Pachyseris Cyphastrea Favia Montipora Platygyra Favites
Fig. 3.3 Stacked histograms of hard coral cover of major genera at three sites from 2003 to 2009.
Fig. 3.4 Mortality rates of coral genera at three sites after major disturbance in 2005. Black bar: whole mortality; white bar: partial mortality.
Fig. 3.5 Changes of relative abundance of major coral genera at three sites between August 2003 and September 2009.
Fig. 3.6 Changes of coral colony number (mean ± SE) of four size classes at three sites from 2003 to 2009.