Habitat determinants of Chaetodon butterflyfish and fishery-targeted coral reef fish assemblages in the central Philippines

Habitat availability and suitability are key factors determining the local abundance of a species and the structure of species assemblages. Nowhere is this more evident than in coral reef systems, where the availability, condition, and complexity of scleractinian corals are major determinants of the abundance, growth, and survival of both specialist and generalist coral reef fishes. Unfortunately, these shallow water habitats are also extremely vulnerable to destructive natural events such as cyclones, as well as anthropogenic environmental disturbances such as reduced water quality, overfishing and destructive fishing, and increased temperatures and reduced ocean pH associated with climate change. The deterioration of shallow water habitats is particularly pronounced in poor, developing nations such as those in the Coral Triangle, where marine biodiversity is high, but so are human population density and reliance on resource extraction. No-take marine reserves (NTMRs) are used throughout the region to reduce human impacts on coastal areas, and produce benefits for conservation, fisheries, or even both. This thesis aims to determine the habitat characteristics that drive the abundance, species richness, growth rates, and assemblage structure of coral reef fishes in order to inform placement of NTMRs, and to identify potential environmental stressors that may require more complex management strategies. Identifying the full range of habitat characteristics that can influence coral reef fish is potentially highly complex. Thus, work reported early in this thesis was carried out on Chaetodon butterflyfish, which exhibit relatively well-known habitat associations, are highly sensitive to changes in habitat condition and habitat loss, and are not normally targeted by fishers, such that a Chaetodon response to habitat condition should not be confounded by fishing effects. Many Chaetodon butterflyfish exhibit a particularly strong association with live coral cover, such that the abundance of obligate corallivorous Chaetodon butterflyfish has often been promoted for use as an indicator of coral reef health. The work presented in Chapter 2 of this thesis tested the influence of a wide range of bottom-up (benthic resource) and top-down characteristics on the abundance of corallivorous and generalist Chaetodon, and overall Chaetodon species richness. This research reinforced certain well-established associations between Chaetodon butterflyfish and their habitats, such as increased abundance of obligate corallivores with increased availability of branching and tabular live coral and with increasing benthic rugosity, but also elucidated and quantified several less-known influences on Chaetodon abundance and diversity. One of these is that the negative effect of macroalgal cover on Chaetodon abundance and species richness is stronger and more directly influential than has previously been reported in the literature. Furthermore, Chaetodon were not affected by the density of large predators, but corallivorous Chaetodon were negatively influenced by the density of small predators (mesopredators), which are most likely affecting new recruits and juveniles. These results encourage the management of coral reefs to limit macroalgal growth (e.g. improve water quality and reduce fishing pressure on herbivores) and to control the abundance of mesopredators. Habitat characteristics can influence not only the presence and abundance of a species, but also its physiology and development, with potential consequences for the survival and reproduction of each individual. Chapter 3 extended previous research on the habitat associations of Chaetodon butterflyfishes to explore the effects of habitat condition on the growth rates of newly-settled fishes, using Chaetodon vagabundus as a model species. The data collected in this chapter also provided the first measure of pelagic larval duration and size-at-age estimates for C. vagabundus in this region, which has previously been used to model larval connectivity of coral reef fishes. I report that larval characteristics of C. vagabundus were consistent across the 80 km study region, but that site-level (1-5km) differences in habitat condition caused significant differences in early post-settlement growth rates for this species. High growth rates were associated with the known habitat preferences of juvenile C. vagabundus (rocky, low rugosity reef flats). However, the strongest habitat association was an unexpected, strongly negative influence of macroalgal cover on early growth rates, as well as evidence of reduced growth rates when density of juvenile conspecifics was high (i.e. negative density dependence). The significant negative influence of macroalgal cover in this region again emphasises the importance of management to control macroalgae, such as by improving water quality and reducing fishing pressure on herbivores. Strong associations between coral reef fishes and their habitats can be used to assess fish responses to environmental disturbances and also to assess the duration of recovery periods of fish and benthos after disturbance. This is particularly true of Chaetodon butterflyfishes and measures of live coral cover. In Chapter 4, I identify the major drivers of the abundance of corallivorous and generalist Chaetodon, as well as overall Chaetodon species richness, and use them to assess the Chaetodon response to a range of environmental disturbance events and subsequent recovery periods over the course of 31 years of almost annual monitoring at two NTMRs and two fished sites at offshore islands in the Bohol Sea, in the central Philippines. Macroalgae was not present at the long-term monitoring sites. Instead, the abundance of generalist Chaetodon and Chaetodon species richness both exhibited a strong negative association with sand and rubble cover. In addition, the uniquely long and consistent monitoring provided the opportunity to quantify the effects of several major disturbance events (typhoons, crown of thorns (COTS) outbreaks, coral bleaching, and destructive fishing) and recovery periods, which saw disturbance events reduce cover of branching and tabular corals and the abundance of corallivorous Chaetodon on average by half in just 1-2 years, while recovery periods saw branching and tabular coral cover and abundance of corallivorous Chaetodon on average double in 11-12 years. The order of magnitude difference in the rate of recovery compared to the rate of decline is an important concern for reef managers and users. Furthermore, the NTMR status of the study sites did not significantly affect benthic cover or Chaetodon abundance, species richness, or assemblage structure, which serves as an important reminder that NTMRs are not a panacea for increasingly disturbed coral reefs. Many coral reef fishes have much more complex habitat requirements than Chaetodon butterflyfishes. For example, some use a range of reef and non-reef habitats at different periods in their lives, and the spatial extent (i.e. area) of each habitat type, and their proximity to each other (i.e. connectedness) have been identified as significant drivers of the abundance and species richness of these coral reef fishes. Chapter 5 combined extensive in situ collection of traditional habitat condition metrics, such as depth and percent cover of massive and encrusting corals, with remotely-sensed imagery that quantified the spatial extent and proximity of major habitat types (coral reef, seagrass beds, mangrove stands), to determine which variables most strongly influenced the presence/absence and biomass of several major coral reef fishery taxa at 19 inshore sites along the coast of Negros Oriental in the central Philippines. Study taxa included both heavily and lightly targeted species groups in order to provide relevant information to coastal resource managers in this region. I determined that NTMR protection and measures of habitat condition specific to each fish grouping (e.g. percent rubble cover for goatfish) were consistently strong predictors of species presence and biomass, while measures of habitat extent and connectedness were rarely significant drivers in this system. These results encourage the management of coastal resources to promote habitat health, and the placement of NTMRs in areas of high habitat quality rather than in marginal habitats in order to maximize their benefits to local fisheries. This thesis emphasises the importance of managing coastal habitats, and coral reefs in particular, to prioritise habitat health and condition, such as improving water quality. In addition, it encourages the placement of NTMRs on healthy rather than degraded habitats in order to generate the greatest benefits to conservation and fisheries