Ecology, conservation and management of Hong Kong freshwater marshlands

Abstract

Hong Kong freshwater marshlands are threatened by anthropogenic impacts and natural processes (terrestrialization and seawater intrusion). However, studies of marshland ecology and conservation are scant. This research provides recommendations on the conservation and management of marshes in Hong Kong, based on a territory-wide survey of marshland macroinvertebrate and fishes, and a fish reintroduction project.

First, I used macroinvertebrate and fish community data to evaluate the effectiveness and representativeness of protected areas in conserving marshland biodiversity, complemented by a habitat vulnerability assessment. I found that sites protected by conservation zonings maintained similar levels of macroinvertebrate α-diversity currently (2014-17) to that present two decades earlier, and were representative in terms of species richness and rarity. However, sites that were highly vulnerable to development or natural degradations were not well represented. Several marshes had been filled or reclaimed in recent years, and the majority of the remainder were also threatened by terrestrialization.

I then used the same 2014-17 dataset to investigate whether the presence of bovids (feral cattle and water buffalo) affected marshland macroinvertebrate communities in both the wet and dry seasons. Bovid presence affected site-scale macroinvertebrate composition, and sites with bovids also had higher macroinvertebrate richness at the site- and microhabitat- scales. No seasonal difference in the effects of bovids was detected, but seasonality itself significantly affected macroinvertebrate composition, α-diversity and species turnover. Overall, these data allowed me to identify several vulnerable marshes with high biodiversity for statutory protection. I also recommend that bovid translocation can be employed for managing marshland communities, given the evident benefits for biodiversity and presumed effects of bovid grazing and tramping in slowing terrestrialization.

A reintroduction project was implemented within a recently revitalized area of marsh at Lai Chi Wo (Hong Kong) for the locally threatened curved-back ricefish (Oryzias curvinotus: Adrianichthyidae). During the reintroduction, I employed environmental DNA (eDNA) methods with the use of species-specific primers to monitor the post-release population of captive-bred individuals. This necessitated development of species-specific primers and standardized protocols. Primer specificity has been supposed to be ensured solely by maximizing the number of nucleotide mismatches between primers and non-target templates. However, I discovered that specific types of mismatches (purine-purine mismatches) reduced non-specific amplification more strongly than other mismatch types. In addition, I found that the widely-employed BLAST program provided misleading predictions on primer specificity during in silico primer development.

After designing ricefish-specific primers, I investigated the effects of ricefish abundance and different environmental parameters on eDNA concentration, and examined the natural diffusion of eDNA in the field with a trial release. Ricefish abundance was the only significant parameter in predicting eDNA concentration and the spread of eDNA was limited. Based on these findings, eDNA methodologies were standardized, and reintroduction proceeded. One year of post-release monitoring using eDNA methods suggested that the reintroduced population had grown 7.5 times in abundance, and expanded its range to adjacent habitats. This success demonstrates that well-planned reintroductions (or translocations) have the potential to conserve threatened aquatic species and can be combined with eDNA methods for post-release monitoring.