**4. Discussion**

With ocean temperatures predicted to rise 3 ◦C by the end of the century [113], the effects of ocean warming on coral reef organisms have received an increasing amount of attention. However, such studies largely ignore the cryptofauna that comprises most of coral reef's biodiversity and biomass, including parasites [67]. In the only long-term monitoring study of any marine parasitic crustacean, Sikkel et al. (2019) [65] reported that during extreme warm-water events in the GBR parasitic gnathiid isopod populations crashed. The findings reported here are consistent with their hypothesis that this may be attributable, in part, to a direct effect of temperature on gnathiid mortality. Such an effect of temperature on the larvae of a tropical ectoparasite has been shown for monogeneans on farmed tropical fish [69].

The present study is the first to examine effects of acute temperature increases on this common reef fish ectoparasite. In our study, gnathiids from both the GBR in Australia and Negros Oriental in the Philippines demonstrated rapid mortality in temperatures raised to above average SST, suggesting that environmental changes in temperature can influence gnathiid survival. In the Philippines, temperatures as little as 2 ◦C (i.e., 32 ◦C) above average seasonal SST (30 ◦C) caused significantly lower survival, with increasingly steep survival curves at 35 ◦C, with the steepest at 36 ◦C, where no gnathiids survived past five days. Unfed gnathiids on the GBR had lower survival at 32 ◦C compared with 29 ◦C, an effect which was consistent across all three juvenile stages. For fed gnathiids on the GBR, the effect of temperature was significant for juvenile stages two and three, with the strongest effect of temperature on stage three, also between the 32 ◦C and 29 ◦C treatments. It is, therefore, likely that gnathiids from both the Philippines and Australia may be living near their thermal limit, as small increases in temperature from the annual seasonal mean have resulted in increased mortality in organisms

from both regions. These results indicate not only that an acute change of temperature to just 32 ◦C decreases the survival of gnathiids, but that the effect of increased temperature is greater on the larger juvenile stages.

In the GBR we found evidence that greater gnathiid headwidth, not just juvenile stage, increased gnathiid survival. For unfed gnathiids, there was a weakly positive relationship between gnathiid survival and gnathiid headwidth, but only in juvenile stage one. For fed gnathiids, there was also a weakly positive relationship between survival and headwidth, but only at 29 ◦C and 32 ◦C. Gnathiid length is correlated with headwidth [95,105] and thus likely with mass also. Both results suggest that even small increases in gnathiid size within a juvenile stage can increase gnathiid survival; these findings also supported our decision to include both headwidth and stage in the statistical model as being important factors to consider when modelling gnathiid juvenile survival. Such a difference in the thermal response related to size may be due to the increased metabolic demand caused by the increase in temperature, an effect which may lead to an energetic deficit for smaller individuals if enough food cannot be obtained, therefore, creating a metabolic mismatch between energy obtained versus energy required [114]. Alternatively, there may possibly be a higher baseline metabolism or higher growth rate at smaller sizes, which then slows down as they reach the maximum size for that stage, resulting in smaller sizes using up their reserves faster than larger sizes. Furthermore, the energetic demands of development may also differ among juvenile life-history stages. Thus increased metabolic demand for basal processes (such as cell maintenance) as a result of increases in temperature, coupled with differential energetic requirements for development may also explain the variation in the thermal response among juvenile stages in our study. Therefore, understanding the effect of increasing temperature on individual metabolism and survival also requires an understanding of food resources and availability [114].

Overall, these results suggest that even with a small increase of 2–3 ◦C above the normal ambient mean, raised temperature can ultimately lead to increased gnathiid mortality. However, there were some key differences in the experimental protocol between the GBR and Philippines studies that should be considered. First, only one species of gnathiid was used in the GBR experiment, compared to at least three (all undescribed, M.O.S. personal observation) species present in the pool of gnathiids used for the Philippines experiment. Second, because these species were unknown, we were unable to confidently separate juvenile gnathiids into their different stages, and so used size class as a proxy. Therefore, we cannot discount the possibility of some among-species and life-stage variation in thermal tolerance. Finally, in contrast to the Philippines, in the GBR study the time the gnathiids were last fed was unknown. This would account for much of the unexplained variation in survivorship in the analysis of the GBR data, as the variation in resources available to the gnathiid, in the form of a blood meal, would decrease over time since their last feeding event. In addition, it should be noted that, in both studies, the gnathiids were not fed for the duration of the experiment, and thus starvation may have been a contributing cause of mortality. While starvation may have influenced mortality of gnathiids among the treatments (as suggested by increased mortality over time in the ambient temperatures), the rates of mortality at higher temperatures were greater, with rapid mortality taking place very early on in the experiments (e.g., one to five days in the Phillipines). This supports the interpretation that increased temperatures influence gnathiid survival directly. It is of relevance that marine "heatwaves" (which are categorised as periods of abnormally high SST lasting for longer than five days [115]) have been predicted to become more frequent, longer and more severe [116,117]. Our observations of rapid gnathiid mortality even after just one day suggests there may be a decline in gnathiid survival from early on in a heatwave, so that gnathiid populations may be heavily impacted if there are more frequent and severe heatwaves in the future.

Our findings appear consistent with data for other tropical marine invertebrates, which have an upper thermal tolerance that is not far above normal sea temperature (reviewed in [4,118]). For example, in a meta-analysis on bivalves, and a study on porcelain crabs, tropical species were found to have upper thermal limits that were closer to the maximum temperature of their habitat than temperate

### *Oceans* **2020**, *1*

species [118,119]. Tropical species of bivalves have also been shown to have a smaller thermal tolerance window than temperate species [118]. This is thought to be due to tropical marine organisms being more sensitive to changes in temperature as they have evolved under relatively invariable thermal conditions [27]. Other studies on marine invertebrates have also shown increased mortality with high SST [120,121], with hermatypic corals being particularly sensitive to increases, with SSTs needing to rise only a few degrees for bleaching to occur [25,26,31,122].

There are a number of studies investigating the potential impact of temperature increase associated with climate change on parasite communities and aquatic parasite-host interactions. The majority of these studies have been on endoparasitic trematodes from temperate regions [123–125]. Temperature was consistently observed to have a significant effect on the survival times of trematodes in their free-living juvenile stage, with survival rates decreasing as temperature increased (e.g., [126–133]). Similarly, temperature has been reported to have an effect on parasitic barnacles (rhizocephalans), with their prevalence decreasing at higher temperatures [68].

In one of the few other studies on ectoparasites, Conley and Curtis (1993) [134] found that, in temperatures of 8–20 ◦C, survival of copepodids was also inversely proportional to temperature. This same trend was observed in the survival rates of monogeneans, and isopods (Cymothoidae) in two studies in sub-tropical regions [135,136] and one study of monogeneans in a tropical region [69]. In all three studies, temperature treatments of 30 ◦C and above had the lowest survival rates [69,135,136]. Similar results were also observed with trematode cercariae from sub-tropical regions [137–139]. Summer temperatures for these lower latitudes parasites are in the range of 30–31 ◦C, which suggests that like gnathiids in the warmer months they are living close to their thermal limits.

Although, this study focused on effects of temperature on mortality, increased temperature can also have sub-lethal effects on marine organisms, impacting their ability to perform essential tasks [140–142]. Based on a review of the literature, Lough (2012) [24] suggested that temperatures between 30–32 ◦C may reflect a potential temperature threshold where a proportion of reef organisms' physiological processes are negatively impacted. Higher temperatures may also affect the ability of parasites with mobile life history stages (such as gnathiids) to successfully detect and associate with a host. To our knowledge there are no studies that specifically examine this. However, the ability to physically reach a host by swimming does appear to be influenced by temperature. For temperate parasitic copepods in their free-living stage, the duration of swimming activity was found to be inversely related to water temperature [134]. For newly emerged cercariae, swimming speed increased in higher temperatures (19–36 ◦C). However, the speed declined over time, with rate of decline increasing with temperature. This resulted in higher swimming speeds, but for shorter durations in water of 30 ◦C and above [143]. In a sub-tropical study, cercariae infectivity also increased with temperature with maximum infectivity occurring at 30 ◦C before declining at 36 ◦C and 40 ◦C [137]. This could be attributable to greater cercariae swimming activity [137]. Although, we did not quantify the effects of temperature on movement, in the Philippines study it was apparent that gnathiids moved more slowly and less frequently at temperatures of 32 ◦C and above, with movement decreasing further as temperature increased, and also with apparent effects greater for the larger size class (M.O.S. personal observation).

Elevated temperatures may also impact host physiology, behavior and survival in ways that impact the balance between parasite and host. For parasitic barnacles (rhizocephalans), the effects of temperature on infected host mortality (and consequent transmission) could threaten their survival, with models showing that just an increase of 2 ◦C in ambient temperature could cause local parasite eradication [68]. In contrast, reef fishes can live further away from their thermal limits than gnathiids were observed to do in this study and in some cases can tolerate temperatures of up to 34–40 ◦C, [3,15,37,47]. However, they can still experience sub-lethal effects with smaller increases in temperatures [3,6,12,14,15,142,144–146], which could also impact host-parasite interactions.

Large hosts, like many reef fishes, can also leave areas of warm water for cooler water, or leave habitat impacted by coral bleaching for other habitats [39,40,43–46,147], depriving gnathiids and other similar ectoparasites, like natatory-stage cymothoid isopods of hosts [148]. The potentially impaired

physiological and swimming ability of the parasite, combined with direct effects on mortality and host availability, could result in a decline in parasite populations. However, the ability of some gnathiids to feed on invertebrate hosts [149,150], combined with weakened immune response for the smaller, less mobile, fish species could leave fish more susceptible to ectoparasites, and thus, compensate for the loss of larger hosts. Indeed, during the 2016–2017 mass bleaching event on the GBR, there was a significant decrease in the numbers of larger, more mobile host fishes in shallow areas, with only smaller, site-attached species remaining [48]. This could have also contributed significantly to the crash in gnathiid populations observed by Sikkel et al. (2019) [65]. However, it should also be noted that as gnathiids are mostly free-living and have a temporary association with their hosts, they too can potentially avoid higher water temperatures. This might happen passively by the gnathiid "hitching a ride" whilst feeding on their host, a process which can last from a few minutes up to a few hours [73]. The gnathiids may, thereby, be transferred to different locations [151,152]. However, as knowledge of the dispersal mechanisms of gnathiids, the infection rates of host fish, and fish movements after disturbances is limited, the proportion of the gnathiid populations that could transfer location with their hosts remains unknown.

Another indirect effect of increased SST may be effects of warming on predators of gnathiids and other ectoparasites' free-living stages. In particular, coral polyps are a major source of predation on juvenile gnathiids [153,154], and thus high coral mortality associated with warm-water events, combined with the loss of cleaner fish [48], which prey on ectoparasites [93], might increase living space and decrease predation on gnathiids. Indeed, once water cools following a bleaching event and most corals are dead, gnathiid populations appear to recover rapidly [65]. While, oceans are also experiencing increased acidification [1,155], Paula et al. (2020) [156] found no effect of acidification on the mortality of the same GBR gnathiid isopod as that studied here.

As parasites have a significant role in ecosystem function, changes in parasite abundance may pose consequences for ecological communities [157–159]. Therefore, while the diversity of coral reef parasites and their hosts makes it difficult to draw general conclusions on how warming events will impact parasite-host interactions, it remains important to further investigate parasite responces to both the direct and indirect effects of warming [67]. Future studies on gnathiids examining sublethal thermal effects on molting, physiology, locomotion, host-detecting mechanisms and reproductive performance will provide a more comprehensive understanding of effects of temperature on host-parasite interactions in coral reef systems.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2673-1924/1/4/16/s1. Table S1: Philippines; Average water temperature of aquaria for five treatments over three aquarium replicates, Table S2: Philippines; Sample size of larval gnathiid isopods < 2 mm and >2 mm in length in five different temperature treatments over three aquarium replicates, Table S3: Great Barrier Reef; Tests of proportionality, using function "cox.zph" in library "coxme" in R 3.2.5, for full model for unfed gnathiid survival among temperature treatments and juvenile stages, Table S4: Great Barrier Reef; Tests of proportionality, using function "cox.zph" in library "coxme", for full model of fed gnathiid survival among temperature treatments and juvenile stages, Table S5: Philippines; Tests of proportionality, using function "cox.zph" in library "coxme" in R 3.2.5, for full model for Trial 1 gnathiid survival among temperature treatments, Table S6: Philippines; Tests of proportionality, using function "cox.zph" in library "coxme" in R 3.2.5, for full model for Trial 2 gnathiid survival among temperature treatments, Table S7: Great Barrier Reef; Summary output for full model for unfed gnathiid survival among temperature treatments and juvenile stages for Cox model, Table S8: Great Barrier Reef; Summary output for full model for fed gnathiid survival among temperature treatments and juvenile stages for Cox model, Table S9: Great Barrier Reef; Analysis of deviance tables (Type II tests) and summary outputs for unfed gnathiid survival for separate Cox models for each juvenile stage, Table S10: Great Barrier Reef; Analysis of deviance tables (Type II tests) and summary outputs for fed gnathiid survival for separate Cox models for each juvenile stage, Table S11: Great Barrier Reef; Analysis of deviance tables (Type II tests) and summary outputs for fed gnathiid survival for separate Cox models for each temperature treatment, Table S12: Philippines; Summary output for full model for Trial 1gnathiid survival among temperature treatments for Cox model, Table S13: Philippines; Summary output for full model for Trial 2 gnathiid survival among temperature treatments for Cox model. Bolded values are ones mentioned in main text, Figure S1: Seawater temperature for Great Barrier Reef data, Figure S2: Great Barrier Reef; Temperatures, measured using a handheld device at 12:00 h, for three replicate aquaria per temperature treatment, Figure S3: Great Barrier Reef; Water temperatures in an aquarium over duration of study for each of the temperature treatments between February 1 and March 2 2018, Figure S4: Great Barrier Reef; Scaled Shoenfeld residual plot for full model testing unfed gnathiid survival relative to time (days), Figure S5: Great Barrier Reef; *Oceans* **2020**, *1*

Scaled Shoenfeld residual plot for full model testing fed gnathiid survival relative to time (days), Figure S6: Philippines; Scaled Shoenfeld residual plot for full model testing gnathiid survival relative to time (days) for Trial 1, Figure S7: Philippines; Scaled Shoenfeld residual plot for full model testing gnathiid survival relative to time (days), for Trial 2.

**Author Contributions:** Conceptualization, P.C.S. and A.S.G.; methodology, P.C.S., A.S.G., M.O.S. and B.D.; validation, A.S.G. and P.C.S.; formal analysis, A.S.G., M.O.S. and A.G.-L.; investigation, M.O.S. and B.D.; resources, P.C.S., A.S.G., M.O.S. and A.G.-L.; data curation, M.O.S. and B.D.; writing—original draft preparation, M.O.S., B.D., P.C.S. and A.S.G.; writing—review and editing, M.O.S., A.G.-L. and P.C.S.; visualization, A.S.G. and M.O.S.; supervision, P.C.S. and A.S.G.; project administration, A.S.G., P.C.S. and M.O.S.; funding acquisition, A.S.G., A.G.-L. and P.C.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Australian Research Council (A00105175, A19937078, ARCFEL010G, DP0557058, DP120102415), Sea World Research and Rescue Foundation Australia (SWR/2/2012), and the US National Science Foundation (OCE-1536794, PC Sikkel, PI).

**Acknowledgments:** We thank the many volunteers and the Lizard Island Research Station (GBR) staff who helped maintain the gnathiid culture and provided equipment and facilities. We also thank Jessica Vorse, who conducted an earlier pilot study on the effects of temperature on gnathiid survival on the Great Barrier Reef; this was invaluable in the development of the final methodology implemented in the present study. We thank the municipality of Sibulan, and Dumaguete City, Negros Oriental, Philippines, for permission to conduct this study (0154-18 DA-BFAR). We also thank Hilconida P. Calumpong, Janet S. Estacion, Rene A. Abesamis, and the staff of the Silliman University Institute for Environmental and Marine Sciences for logistic support, equipment and use of facilities. We thank Jeremiah Gepaya and Lucille Jean Raterta for their field assistance and Dioscoro Inocencio for fish collections and field support. The Authors are also grateful to the three anonymous reviewers for their constructive comments.

**Conflicts of Interest:** The authors declare no conflict of interest.
