No abstract

Few studies have investigated the long-term temporal dynamics of seagrass beds, especially in Southeast Asia. Remote sensing is one of the best methods for observing these dynamic patterns, and the advent of deep learning technology has led to recent advances in this method. This study examined the feasibility of applying image classification methods to supervised classification and deep learning methods for monitoring seagrass beds. The study site was a relatively natural seagrass bed in Hat Chao Mai National Park, Trang Province, Thailand, for which aerial photographs from the 1970s were available. Although we achieved low accuracy in differentiating among various densities of vegetation coverage, classification related to the presence of seagrass was possible with an accuracy of 80% or more using both classification methods. Automatic classification of benthic cover using deep learning provided similar or better accuracy than that of the other methods even when grayscale images were used. The results also demonstrate that it is possible to monitor the temporal dynamics of an entire seagrass area, as well as variations within sub-regions, located in close proximity to a river mouth.

Broad scale studies in seagrass benthic macrofauna are important for future regional marine conservation. We examined spatial variation in the community structure of seagrass‐associated benthic macroinvertebrates collected by sediment coring in 2010 at six seagrass sites of Japan covering the latitudinal range of 24°–43°N. Total species richness and ES(50) at site level did not show clear site variations and relationship with latitude. At core level, site variations of mean species richness, ES(50), Simpson diversity and abundance showed inconsistent pattern, but with more cases of statistically significant association with latitude. Variations were generally influenced by the seagrass species, often among subtropical species, among temperate Zostera species, and between Zostera and subtropical species. Finally, the community composition differed significantly across all sites and community similarity decreased rapidly with geographic distance, with only 5% similarity retained at the distance of 400 km. The dissimilarity among sites was higher with the similar distance compared to other types of coastal communities such as rocky intertidal assemblages, which is associated with minor occurrence of species with broad distributional range.

The catch composition and economic impacts of ghost fishing in the fishing grounds near Suan Son Beach, Rayong province, Thailand, were examined based on interviews of squid fishers and experiments using ghost-fishing squid traps. In the province, 27 fisher families are engaged in squid trap fishing, and each family operates 100-300 squid traps. The catch per unit effort (CPUE) of the traps is 30-40 kg/100 traps/trip. The average price of a squid trap is 135 Thai baht (US$ 4.50), and traps have a lifespan of 1-2 months. Squid traps can ghost fish when they are lost in bad weather conditions, due to gear conflict, or when the trap materials deteriorate. The catches of experimental ghost-fishing traps were examined during two periods in 2017: May-July and August-October. The traps caught cephalopods and other species, both commercial and non-commercial. The commercial species included bigfin reef squid (Sepioteuthis lessoniana), other squids, cuttlefishes, groupers, snappers, and blue swimming crabs. The total number of aquatic animals, total weight, and total economic value of the catches during May-July and August-October 2017 were 51 and 38 specimens; 12 000 g and 7250 g; and 6318.0 baht/27 traps and 5302.5 baht/28 traps, respectively.

Seagrass beds support highly productive and diverse animal communities. Extreme physical disturbances such as tsunamis are expected to affect the abundance and diversity of both the seagrass and the infaunal community. To examine the effects of a tsunami on infaunal seagrass communities, we compared the communities in seagrass beds before and after the tsunami that hit the Andaman Sea coast of Thailand on December 26th, 2004. Infauna were collected in 2001 and again in 2005 from inside and outside seagrass vegetation. Polychaetes were the most abundance taxa, followed by crustaceans. Density of macrobenthic animals varied greatly among sites, among areas inside and outside the vegetated areas and between the two years. The density increased over time in the areas with seagrass while it deceased in the non-vegetated areas, suggesting that the tsunami impacted the two areas differently. Multivariate analysis on the polychaete assemblage revealed that temporal changes in assemblage structure differed between vegetated and non-vegetated areas, and that the degree of temporal changes in assemblage structure was not necessarily related to the magnitude of the tsunami.