<i>Supporting data for</i><i> Spatio-temporal community dynamics of Hong Kong rocky shores</i>

<p dir="ltr">This dataset is submitted for JWT Lau's thesis submission, which will be kept private for ~ 3 years. For people who are interested in the dataset, please email me at jacksonlau620(at)gmail.com</p><p dir="ltr"><br></p><p dir="ltr">Community ecology has been described as a “mess” (Lawton, 1999) due to the variations in community structure which occur across multiple dimensions (e.g. space, time) at different scales. Elucidating patterns of spatio-temporally variable community structure requires thorough documentation of community dynamics over different scales. This thesis focuses on establishing quantitative baselines of variation in community structure (at various sites and across timescales from months to four years) on Hong Kong rocky shores to identify key drivers of their complex dynamics.</p><p dir="ltr">Firstly, 24 rocky shores around Hong Kong waters were surveyed in the two monsoon seasons (winter and summer) over two years. Using beta diversity analysis, three major community structures could be described: wave-sheltered; exposed; and low salinity and high nutrient (= Northwestern) groups. Such classification conforms with previous qualitative descriptions in Hong Kong, showing rocky intertidal communities exhibit high species replacement along the wave-exposure and east-west salinity gradients, and strong seasonal changes characterized by consistent summer-loss and winter-gain patterns.</p><p dir="ltr">Secondly, to further investigate the temporal changes, communities of three shores were monitored monthly for 49 months. This community dataset documented compositional changes at a finer temporal scale (monthly) over more annual cycles (four years), compared to the 24-site surveys. Using time series analysis, the within-year variations were demonstrated to be caused by strong seasonal changes in population dynamics of individual species. This strong inter-annual variation was mainly caused by stochasticity in supply-side ecology (i.e. varying intensity of pelagic propagule supply of different species across years) and summer mortalities.</p><p dir="ltr">Finally, the impact of the summer-loss patterns described above were explored for ecologically important species such as the foundation species: the oyster Saccostrea cuccullata and mussel Mytilisepta virgata. These two species form shell beds, increasing structural complexity on bare rock which support a diversity of associated species. From laboratory and field experiments, the mussels suffered an earlier onset and a greater magnitude of mortality than the oysters. Mortality in the case of the mussels removed the shell matrix, whereas for oyster mortalities the dead shells did not detach over the summer, and thus their facilitative functions persist. Mussel mortality, unlike oyster mortality, caused declines in associated species number, demonstrating that these two bivalves from the same guild have differential responses that impact local species diversity, thus supporting the caveat that species identity matters when studying ecological processes.</p><p dir="ltr">Overall, this thesis unravels part of the “mess” of community ecology, by providing quantitative documentation of rocky shore dynamics across multiple spatial and temporal scales in Hong Kong. The community dynamics described reflect Hong Kong’s biogeographic position within the tropics, but under a strong monsoonal influence, exhibiting characteristics of both tropical-like rapid community turnover, and temperate-like seasonal dynamics. By identifying wave exposure, salinity gradients, seasonal mortalities and recruitment patterns, this thesis establishes a quantitative framework for understanding complex dynamics in the relatively understudied monsoonal tropics. These findings contribute to positioning Southeast Asian intertidal communities within broader global patterns of community ecology and provide important baselines for predicting how future climate changes may affect regional biodiversity.</p>