Ghost Fishing Impacts on Hydrocorals and Associated Reef Fish Assemblages

Marine Environmental Research 161 (2020) 105129

Contents lists available at ScienceDirect

Marine Environmental Research

journal homepage: http://www.elsevier.com/locate/marenvrev

Ghost fishing impacts on hydrocorals and associated reef fish assemblages

T.M. Beneli a, P.H.C. Pereira b, J.A.C.C. Nunes a,c, F. Barros a,d,* a Laboratorio´ de Ecologia Bentonica,ˆ CIENAM, PPGEcologia, Instituto de Biologia, Rua Barao˜ de Geremoabo s/n., Campus Ondina, Universidade Federal da Bahia, CEP 40170-115, Salvador, BA, Brazil b Projeto Conservaçao˜ Recifal (PCR), Recife-Brasil. PC Maciel Pinheiro, 369 - Sala 01, Boa Vista. Recife, PE, Brazil c Reef Ecology Group, N42, CEP, 40155-010, Salvador, BA, Brazil d Instituto Nacional de Ciˆencia e Tecnologia em Estudo Interdisciplinares e Transdisciplinares em Ecologia e Evoluçao˜ (INCT IN-TREE), Brazil

ARTICLE INFO ABSTRACT

Keywords: Ghost fishingis a threat to many marine environments, as lost or discarded fishinggear (e.g., fishinglines, nets) Fire-coral continues to fishby entangling, damaging or killing various organisms. Among the benthic organisms that live on Fishing line tropical reefs, the group probably most affected, due to their shape, are the branching corals. These corals Coral mortality provide refuge, foraging and breeding sites, especially for fishes and therefore impacts on coral structure could Fish behavior compromise the ecology of associated species. We tested if fishing lines entangled on the branching coral Mil Habitat degradation Coral-fish interaction lepora alcicornis would result in an increase in colony mortality, decrease in abundance and richness of fishesand changes in the behavior of associated reef fish.In the field,we estimated the volume of M. alcicornis colonies and its mortality percentages, and videos were recorded to evaluate abundance and richness of fishassemblages and fish behavior. Our results showed that coral mortality increased with increasing amounts of entangled fishing lines. Fish assemblages were similar in M. alcicornis colonies with or without entangled fishing lines. Never theless, we observed a significant decrease in the frequency of feeding attempts in two herbivore fish species (Acanthurus bahianus and Ophioblennius trinitatis) that play an important role in coral-reef dynamics, controlling algae abundances. Therefore, ghost fishing has negative impacts on shallow reef ecosystems, directly affecting branching corals and important coral-fish interactions. Management of tropical shallow reef environments should consider regulation and monitoring of coastal fisheries to ensure reef integrity.

1. Introduction ghost fishing on sessile invertebrates (Chiappone et al., 2005) and its interactions with fish are virtually unknown. Ghost fishing is the term used when fishing gear that is abandoned, Lost fishinglines entangled on branching corals might damage their discarded or lost, continues to fish or to exert negative effects (e.g. fragile polyps and tissue or even literally rip away branches (Asoh et al., entanglement, ingestion, death) on marine organisms (Gilman et al., 2004; Chiappone et al., 2005; Valderrama Ballesteros et al., 2018). 2016; Matsuoka et al., 2005). Many lost fishing gears, such as traps, When the injury is severe, coral recovery may become debilitated, fishing nets, and lines, can persist in the sea for centuries due to their compromising their survival (Matsuoka et al., 2005; Yoshikawa and resistant and non-bio-degradable composition (Brown and Macfadyen, Asoh, 2004). However, the effects of ghost fishing on branching corals 2007). Ghost fishingeffects on marine megafauna (e.g. turtles, dolphins, have not yet been fully addressed and could be underestimated. whales) are very well known and widely publicized and it has been Branching corals are very important in coral reefs around the world suggested that ghost fishing mortality can affect fish stock assessment due to their complex and delicate architectural structure. The branching models (Gilman, 2015; Gilman et al., 2016). The impacts can also affect hydrocoral species of the genus Millepora are abundant in the South benthic habitats, for instance by damaging benthic communities on Atlantic Ocean, therefore, it is an important habitat-builder on south- vulnerable and significant coastal ecosystems such as coral reefs and western Atlantic reef systems (de Souza et al., 2017). Millepora com seagrass beds (Gilman, 2015; Shester and Micheli, 2011; Yoshikawa and plex tridimensional structure provides habitat and resources for many Asoh, 2004). Nevertheless, studies regarding the potential effects of marine invertebrates and fishes, resulting in complex patterns of

* Corresponding author. Laboratorio´ de Ecologia Bentonica,ˆ CIENAM, PPGEcologia, Instituto de Biologia, Rua Barao˜ de Geremoabo s/n., Campus Ondina, Uni versidade Federal da Bahia, CEP 40170-115, Salvador, BA, Brazil. E-mail address: [email protected] (F. Barros). https://doi.org/10.1016/j.marenvres.2020.105129 Received 23 June 2020; Received in revised form 21 August 2020; Accepted 22 August 2020 Available online 27 August 2020 0141-1136/© 2020 Elsevier Ltd. All rights reserved. T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129 biological interactions (Coni et al., 2012; Garcia et al., 2008; Leal et al., assemblages dominated by turf algae, zoantharians, sponges and echi 2015, 2013; Lewis, 2006; Pereira et al., 2012). Large branching coral noderms (Ferreira et al., 2015). Other abundant benthic organisms colonies (i.e. high volume) exhibit large abundances and high richness found in the area ascidians and coral colonies of Favia gravida, Agaricia associated fishassemblages (Agudo-Adriani et al., 2016; Counsell et al., agaricites, Montastraea cavernosa, Mussismilia hispida, Siderastrea spp., 2018; Johnson et al., 2011; Leal et al., 2015; Pereira and Munday, 2016). and fire corals (Millepora spp.) (Ferreira et al., 2015). Reefs at the TSB Nevertheless, branching corals are likely among the most affected entrance are exploited by human activities such as artisanal fisheries, benthic group by hooks and fishing lines (Chiappone et al., 2005). marine ornamental trade and spearfishing, mostly because it is located Reef fish assemblages are broadly affected by the loss of live corals within the great urban center of Salvador city (Ferreira et al., 2015; (e.g. Lokrantz et al., 2009; Bonin et al., 2011) and different fishspecies Nunes et al., 2012). may have distinct responses (Wilson et al., 2006). For instance, it is expected that impacts on branching corals will likely affect specificfish 2.2. Ghost fishing on hydrocorals assemblages. These assemblages associated with hydrocorals use the colonies mostly for shelter/refuge (Coni et al., 2012), as territory (e.g. Fieldwork was conducted between June and August 2018 by snor territorial herbivores) and as a food source by foraging on colony polyps keling at low tide between 0800 and 1700 h, always with good visibility (Leal et al., 2015, 2013). Hence, fishing lines entangled on hydrocoral (~10 m). A total of 30 colonies of Millepora alcicornis were haphazardly colonies may limit fish use, i.e. if there is loss of shelter or resource sampled, at least 2 m apart. Each colony had their length, width and limitation for fishes that forage among colony polyps. Importantly, the height estimated (cm) using a tape measure. It is well known that presence of resident fishes amid coral colonies could be beneficial for M. alcicornis colonies present a high heterogeneity regarding their shape coral resilience and recovery since they remove competing seaweeds (de Weerdt, 1981; Edmunds, 1999), and that colony volume may have a (Chase et al., 2018; Dixson and Hay, 2012). positive influenceon the abundance and richness of the associated reef Here, we investigate the potential impact of ghost fishing (i.e. fishes( Leal et al., 2015). Therefore, colony volume (m3) was estimated entangled fishinglines) on the branching hydrocoral Millepora alcicornis using the formula V = L. W . H, where L is the length (the longest base and associated reef fish assemblages. We aim to analyze the conse axis of the colony), W is the width (orthogonal to the firstand the second quences of ghost fishingon hydrocoral mortality as well as for associated longest base axis) and H is the height (the height measured from the reef fish assemblages. Specifically, we hypothesized that an increase in substrate to the largest branch of the colony) (adapted from Coni et al., fishing lines entangled on M. alcicornis colonies, (i) would increase the 2012; Leal et al., 2015). All measurements were taken manually by the mortality of hydrocoral colonies (M. alcicornis), (ii) would decrease the same person (TMB) (ESM Fig. 1). The same procedure was applied to abundance and richness of associated reef fish assemblages. Addition measure the entangled fishinglines in each colony. We used colony and ally, we assessed whether fishing lines in M. alcinornis colonies would fishing line volumes to calculate the proportion of fishing line occu change fish behaviors such as sheltering, feeding, agonistic and non- pancy (percentage) in each colony. The mortality of the colonies was occupational swimming. visually estimated by the same observer (TMB) following standard methods (Leao˜ et al., 2015). 2. Materials and methods

2.1. Study site 2.3. Reef fish assemblages

The present study was conducted at Todos os Santos Bay (TSB), Underwater videos were recorded in all 30 sampled colonies. To Salvador, Brazil (Fig. 1), a social and ecologically important natural guarantee the observation of the entire colony two GoPro Hero 3 cam system that present high biodiversity along with several ecosystems, eras, on 25 cm height monopods placed on the substratum (ESM Fig. 2), including coral reefs, estuaries, rocky reefs and mangroves (Barros et al., were simultaneously used. Each camera was placed on opposite di 2008; Costa et al., 2015; Lorders et al., 2018). The studied site was a rections at 50 cm away from the colonies and recorded for 7 min. The shallow reef (Fig. 1), depths between 1 and 8 m, with benthic first and the last minute of footage were excluded from the video anal ysis to avoid potential interference of the diver (Pereira et al., 2016) as

Fig. 1. Sampled site located at Todos os Santos Bay (TSB), Brazil.

2 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129 well as to provide an acclimation period for the fishes regarding the as a plausible option for such cases (Zuur et al., 2009). presence of the cameras. During the video recording, divers have kept at The relationship between the abundance and richness of fisheswith least 2 m apart from the recorded colony. volume of the M. alcicornis colonies and different percentages of fishing To assess the influence of fishing line occupancy on M. alcicornis line occupancy were modelled with GLM fitted with a Poisson distri colonies on the reef fish assemblages, the total abundance and species bution. We chose the Poisson distribution because the response variables richness were evaluated. Two videos were excluded due to poor quality, were count data without an upper limit and their variance was equal to therefore, for the analysis of associated reef fish assemblages n = 28. the mean (Zuur et al., 2009). Furthermore, graphical exploration Recorded videos from the two cameras of each colony were analyzed showed that the data did not exhibit normal residuals and heteroge separately, three times each, by a single trained observer (TMB). The neous variances. We applied the deletion of the interaction between the first analysis focused on large fishes (such as roving herbivores), the predictor variables from the model since they did not show a significant second analysis focused on cryptic fishes(low mobility and small size), relationship with the associated reef fishabundance and richness. Then, and the third analysis was conducted to confirmthe observations made we used the Akaike Information Criterion (AIC) to select the most on the previous two. The abundance was calculated by the maximum plausible model (with or without the deletion of the interaction between number of individuals of each species (maxN) present in a single frame, the covariates). The result confirmedthat the best fittingmodel was the which is a method widely used (Bacheler and Shertzer, 2015; Pereira one not considering the interaction between the covariates because it et al., 2016). Species richness was considered as the total number of had the lowest AIC value (Anderson, 2008; Zuur et al., 2009). For all species observed considering the videos of the two cameras on each GLMs, we calculated the cross-validated r2 value (1 - predictive error colony. sum of squares/total sum of squares) to assess fitted models strength (Zuur et al., 2009). 2.4. Behavioral observations Before applying GLM with a Poisson distribution, a Spearman cor relation test was used to investigate a possible correlation between the We analyzed the behavior of the three most frequent and abundant covariates, which is appropriate for non-normal continuous data (Bish fishspecies among M. alcicornis colonies (Stegastes fuscus, Ophioblennius ara and Hittner, 2012). trinitatis and Acanthurus bahianus; ESM Table 1). We established a min To test for differences in the assemblage composition of associated imum observed frequency of 10 (i.e. a species should be observed in at reef fishes in colonies with and without entangled fishing lines we per least 10 different colonies) to be include in behavioral analyses. We did formed the permutational multivariate analysis of variance (PERMA not consider one of the most abundant fishspecies (Stegastes variabilis) in NOVA). We used Bray-Curtis distance to obtain a dissimilarity matrix the behavior analyses because a large gap was observed in the abun using 9999 permutations (Anderson et al., 2008). Despite having an dance of this species along our measured fishing line occupancy unbalanced design (n = 9 for colonies without fishingline and n = 19 for gradient. colonies with different occupancy percentages of fishingline) no further Behavioral traits were observed in the videos using the focal animal alterations in the analysis were needed since we had only one factor method (Altmann, 1974). As each colony was simultaneously recorded (Anderson et al., 2008). This analysis was performed using Primer-e 6 by two cameras, the chosen video for behavioral observations was al PERMANOVA+1.0 software. ways the one with the greatest abundance of the focal animal. Four To evaluate fish behavior changes, we analyzed the influence of behavioral categories were used (feeding, sheltering, agonistic, different percentages of fishing line occupancy on the total number of non-occupational swimming) considering behavioral traits already events (i.e. behaviors display by fish species associated with described in the literature (Coni et al., 2012; Helfman, 1989; Leal et al., M. alcicornis) using generalized additive models (GAM) with negative 2015, 2013) for the three fish species (ESM Table 1). Ophioblennius tri binomial error distributions since data exploration of residual plots nitatis did not show “non-occupational swimming” and “sheltering” showed heteroscedasticity and overdispersion in all response variables behaviors (Table 1), whereas “sheltering” and “agonistic behavior” (Wood, 2006; Zuur et al., 2009). The validation of all GLM and GAM categories were not displayed by Acanthurus bahianus (Table 1). models were performed through the graphical analysis of the residuals (see ESM Figs. 4–8 for residuals plots of the significant models) (Zuur 2.5. Data analysis et al., 2009). All GLM and GAM models were conducted using the R language for statistical computing with the packages MASS (GLM with The influence of different percentages of fishing line occupancy on negative binomial distribution) (Venables and Ripley, 2002), mgcv mortality of M. alcicornis colonies was analyzed using generalized linear (GAMs) (Wood et al., 2016; Wood, 2011) (R Core Team, 2017). models (GLM) fitted with a negative binomial distribution. GLM was Like in most of the environmental impact studies, in our study an chosen after we verified that residuals were not normal, and variances error type II would be worse than type I (accepting the null hypothesis were heterogeneous, through graphical exploration. The response vari when there is in fact, an effect) (Hatje et al., 2016; Quinn and Keough, able was overdispersed and the negative binomial distribution is known 2002). Therefore, we considered all models, as well as the PERMANOVA analysis, being significant when p < 0.1. Table 1 List of behavior categories and their respective description and metrics. 3. Results

Behavior Description Metric category 3.1. Ghost fishing on hydrocorals

feeding biting algae, polyps or Bites microinvertebrates on M. alcicornis From the total of 30 observed colonies of M. alcicornis, 9 did not have and nearby entangled fishing lines, and 21 colonies had different percentages of sheltering sheltering in the crevices, hollows or sheltering attempts fishing line occupancy, ranging from ≅ 3%–70%. The mean total vol branches of M. alcicornis colonies and ume of the colonies was 30 ± 20 dm3 (mean ± SD) (varying from 8 dm3 nearby to 100 dm3) and the mean proportion of mortality percentages in the agonistic aggressive interactions with other fish agonistic interactions ± that tried to approach or were roving colonies was 22.5 19.43% (varying from 5 to 75%). next to M. alcicornis colonies Percentages of fishing line occupancy in M. alcicornis explained non-occupational swimming near (above and around) swimming above/ 33.68% of the variation in the mortality of the colonies (Table 2A). swimming M. alcicornis colonies around the colony Mortality increased as fishing line occupancy in the colonies increased attempts (Fig. 2, ESM Fig. 10) and the GLM regression model indicated a

3 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129

Table 2 Regression models (GLM) explaining the relationships between A) mortality percentages of M. alcicornis colonies and fishing line occupancy; B) abundance of the associated reef fish assemblages and fishing line occupancy and colony volume; and C) species richness of the associated reef fish assemblages and fishing line oc cupancy and colony volume. SE: Standard error; DE: Deviance explained.

Variables Model Estimate SE z-value Pr (>|z|) DE (%)

A) Colony mortality Negative binomial regression (GLM) 33.68 Intercept 2.73 0.16 16.66 < 0.0001 Fishing line occupancy 0.01 0.01 1.85 0.063 B) Fish abundance Poisson regression (GLM) 29.01 Intercept 1.12 0.16 6.98 < 0.0001 Fishing line occupancy 0.01 0.01 0.41 0.677 Colony volume 9.32 3.70 2.51 0.011 C) Fish species richness Poisson regression (GLM) 20.50 Intercept 0.92 0.18 5.12 < 0.0001 Fishing line occupancy 0.01 0.01 0.25 0.796 Colony volume 7.85 4.28 1.83 0.07 significanteffect on the relationship between fishingline occupancy and mortality of the colonies as well as the intersect of the model (Table 2A). Table 3 Total number of individuals of fishspecies associated with M. alcicornis colonies We also ran a GLM model to explore a possible influenceof an outliner and the relative frequency of occurrence. on the regression line (colony mortality ≅ 7%), but the pattern was very similar. Fish species Individuals (N) Freq. occur. (%) Stegastes fuscus 30 71.4 3.2. Reef fish assemblages Ophioblenius trinitatis 15 46.4 Malacoctenus sp. 13 42.8 Stegastes variabilis 11 32.1 A total of 120 reef fishes from 18 species were observed associated Acanthurus bahianus 11 25.0 with M. alcicornis colonies. The fivemost abundant and frequent species Halichoeres poeyi () 8 21.4 were Stegastes fuscus, Ophioblenius trinitatis, Malacoctenus sp., Stegastes Malacoctenus delalandii 7 25.0 variabilis and Acanthurus bahianus, representing 25%, 12.5%, 10.8%, Chaetodon striatus 6 14.2 Acanthurus chirurgus 4 7.1 9.1% and 9.1% of the total fish assemblages, respectively (Table 3). Holocentrus adscensionis 3 10.7 Myrichthys ocellatus and Sphoeroides spengleri represent new records for Cantherhines pullus 3 7.1 reef fishes associated with the hydrocoral Millepora spp. Halichoeres brasiliensis 2 7.1 Fishing line occupancy and colony volume explained 29.01% of the Halichoeres penrosei 2 7.1 variation in fish abundance (Table 2B) and 20.05% of the variation in Abudefduf saxatilis 1 3.5 Cephalopholis fulva 1 3.5 richness (Table 2C). Nevertheless, the Poisson regressions (GLMs) Myrichthys ocellatus 1 3.5 showed significanteffects for colony volume but no significanteffect of Sphoeroides spengleri 1 3.5 fishing line occupancy on fish abundance or richness (Fig. 3). Labrisomus kalisherae 1 3.5 The reef fish assemblage composition associated with M. alcicornis was not significantly different among colonies with and without the 3.3. Behavioral analysis presence of fishinglines (ESM Table 2); indicating that fishassemblages frequenting colonies regardless of the presence or not of entangled We observed a decline in the total number of bites of A. bahianus and fishing lines were similar. O. trinitatis in coral colonies with greater volumes occupied by fishing lines (Fig. 4). Fishing line occupancy was a good predictor of the total variation in the total number of bites, with high levels of deviance explained (55.5% for A. bahianus and 40.6% for O. trinitatis) (Table 4) For S. fuscus, the smoother for fishing line occupancy was not sig nificant regarding all the four behaviors analyzed (feeding as total number of bites, sheltering as total sheltering attempts, non- occupational swimming as total swimming attempts and agonistic as agonistic interactions) (Table 4, ESM Fig. 9).

4. Discussion

Increasing areas of entangled fishing lines in branching coral col onies showed an increase in colony mortality. Therefore, we corroborate previous studies with massive corals (Yoshikawa and Asoh, 2004) and mainly branching corals (Valderrama Ballesteros et al., 2018) observing ghost fishing effects on corals reefs. Although Millepora, as other hydrocorals, is considered relatively tolerant to physical disturbances, predation and diseases, they are also vulnerable to competition with macroalgae and zoantharians (Lewis, 1989) as well as bleaching (Marshall and Baird, 2000; Banaszak et al., 2003; Fitt, 2012). Addi tionally, branching corals are certainly more vulnerable than massive corals, to anchoring, aquarium trade (Giglio et al., 2017), recreational Fig. 2. Relationship between the mortality of M. alcicornis colonies (%) and diving (Giglio et al., 2015; Lyons et al., 2015) and trampling (Hannak fishing line occupancy in relation to the total volume of the colonies (%). et al., 2011). Such factors may explain the mortality herein observed in Dashed lines represent 95% confidence interval bands.

4 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129

Fig. 3. Relationship between (A) total fishabundance per colony and fishingline occupancy in relation to the total volume of the colony (%) showing no significant effect, (B) total fish abundance per colony and colony volume, (C) fish species richness and fishing line occupancy in relation to the total volume of the colony (%) showing no significant effect, and (D) fish species richness and colony volume. Dashed lines represent 95% confidence interval bands.

Fig. 4. Estimated smoothing curves for GAM representing the relationship between fishingline occupancy and feeding behavior (total number of bites/individual) of A: Acanthurus bahianus, and B: Ophioblennius trinitatis. Dashed lines represent 95% confidence interval bands.

Table 4 GAM output explaining the relationship between fishing line occupancy and behavior categories for different fish species associated with M. alcicornis colonies. Deviance explained; s: variable smoother.

Models Acanthurus bahianus Ophioblennius trinitatis Stegastes fuscus

d.f p value D.E (%) d.f p value D.E(%) d.f p value D.E (%)

feeding ~ s(fishing line) 2.2 0.022 55.5 2.0 0.038 40.6 1.0 0.88 0.06 sheltering ~ s(fishing line) _ _ 1.0 0.98 0.003 swimming ~ s(fishing line) _ _ 1.0 0.25 4.37 agonistic ~ s(fishing line) _ _ 1.0 0.72 0.45

5 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129 some colonies with low fishing line occupancy. We observed a few sponges colonizing the fishing lines. The depletion of habitat quality is M. alcicornis colonies overgrowing fishinglines (ESM Fig. 11), indicating likely to affect coral specialized fish species (Munday, 2004; Wilson a potential recovery. M. alcicornis can overgrow other animals (e.g., et al., 2006) while generalist species are less prone to be affected by Wahle, 1980), but few coral species are known to overgrow fishinggear such. For instance, a decline in the abundance of specialist fishes in (Hoeksema and Hermanto, 2018). Further studies would be necessary to Caribbean reefs was observed between 1980 and 2006, due to reef better investigate the recovery potential of Millepora alcicornis colonies structure simplification, while the abundance of generalist species after a ghost fishing event and how this interacts with other anthropo remained stable (Alvarez-Filip et al., 2015) genic disturbances (e.g. coral bleaching, recreational diving). Ghost fishing has shown severe impacts on invertebrates that play Millepora species play an important role in the dynamic and function key roles for reef fishassemblages and reef dynamics (Link et al., 2019). of reefs (Brown and Edmunds, 2013), especially in sites with low di We presented the effect of entangled fishing lines in branching corals versity and continuous decline of scleractinian coral cover such as and its associated reef fishes. There are many challenges for research Southwestern Atlantic reefs (Cruz et al., 2018). They are important reef regarding ghost fishing impacts, but we emphasize that hydrocorals builders and provide a complex three-dimension habitat (Glynn and should be a priority concern in coastal reef environments with low coral Enochs, 2011; Leal et al., 2015). Therefore, ghost fishing affects such cover. habitat complexity, which supports high biodiversity and complex bio Increasing the awareness and efforts to avoid abandonment and logical interactions (e.g., Montano et al., 2020) disposal of fishing gears are important strategies to avoid ghost fishing The hypothesis that fishing lines in M. alcicornis colonies would impacts. Another solution could be the replacement of traditional fish decrease the total abundance and species richness of the associated reef ing gears for biodegradable ones (Gilman, 2016). Additionally, regula fish assemblages was not corroborated. We found no changes in the tion and monitoring of artisanal fisheriesin reef environments would be abundances and richness of the associated fish assemblages, regardless valuable strategies to guarantee reef integrity. of the amount of fishinglines entangled on the colonies. We argue that the presence of fishinglines does not affect most of the associated fishes, Credit author contribution statement but it could affect a certain species that requires one or more resources provided by Millepora colonies. T.M. Beneli: Conceptualization, Data curation, Formal analysis, We observed entangled fishing lines colonized by filamentous, Field work, Lab work, Methodology, Writing - original draft, Writing - calcareous algae, turf algae and/or sponges (ESM Fig. 10). These or review & editing. P.H.C. Pereira: Conceptualization, Methodology, ganisms may provide resources to the associated fishes like territorial Supervision, Writing - review & editing. J.A.C.C. Nunes: Conceptuali damselfishes(e.g. S. fuscus and S. variabilis) and roving herbivores (e.g. zation, Data curation, Fish behavior analyses Supervision, Methodology, A. bahianus, A. chirurgus, C. striatus and S. axillare). On the other hand, Supervision, Writing - review & editing. F. Barros: Conceptualization, macroalgae is a direct and aggressive competitor of corals in reef envi Funding acquisition, Methodology, Project administration, General Su ronments (Clements et al., 2018; Dixson and Hay, 2012) and its grown pervision, Writing - review & editing. on fishinglines could compromise Millepora spp. growth and reduce its competitive advantage. Our results support that the abundance and richness of reef fishesare Declaration of competing interest positively affected by the volume of the branching coral M. alcicornis, (Coni et al., 2012; ). The larger the colony volumes, the higher is the The authors declare that they have no known competing financial habitat structure as observed by previous studies (Carvalho and Barros, interests or personal relationships that could have appeared to influence 2017; Darling et al., 2017; Leal et al., 2015). Millepora spp. colonies are the work reported in this paper. feeding ground for several reef fish species (Pereira et al., 2012, 2016) and we show that O. trinitatis and A. bahianus decreased the number of Acknowledgments bites as fishingline percentages in the colonies increased. Colonies with great amounts of fishing lines entangled showed high mortality and We thank LEB members for fieldassistance, and E. Mariano Neto and could become an unattractive resource for foraging fishes. Therefore, P. Dodonov from the Federal University of Bahia for valuable insights ghost fishing may affect the foraging behavior of territorial and roving and assistance on data analyses. We thank CAPES for the master’s herbivores, both important fish trophic groups for coral reef dynamics. scholarship granted to T.M.B.. F.B. was supported by CNPq (PQ 306332/ Surgeonfishes (e.g., Acanthurus bahianus (ESM Fig. 12), controls algae 2014-0; 304907/2017-0) and INCT in-TREE, CNPQ (465767/2014-1). biomass (Cheal et al., 2010; Marshell and Mumby, 2015), especially in degraded reef systems (Plass-Johnson et al., 2015). Furthermore, Appendix A. Supplementary data coral-algae interactions are widely affected by climate change (Del Monaco et al., 2017) and coral reefs have demonstrated severe decline Supplementary data to this article can be found online at https://doi. worldwide (Bellwood et al., 2004; Hughes et al., 2018, 2017). Thus, the org/10.1016/j.marenvres.2020.105129. influence of ghosts fishing on the ecological role of herbivores control ling this process seems to be critical in times of climate change. The Compliance with ethical standards cryptic fish O. trinitatis is predominantly herbivorous, compete for high-quality territories and, therefore, have their density dependent on Conflictof interest All authors declare that they have no conflictof available substrates where structural complexity is particularly impor interest. tant (Medeiros et al., 2014). The sharp decrease in the number of bites (feeding attempts) in colonies with large amounts of fishinglines can be References related to lower-quality territories, which require larger grounds and foraging over longer distances. However, this needs to be formally Agudo-Adriani, E.A., Cappelletto, J., Cavada-Blanco, F., Croquer, A., 2016. Colony tested, since little is known about cryptic fish behavior. geometry and structural complexity of the endangered species Acropora cervicornis partly explains the structure of their associated fishassemblage. PeerJ 4. https://doi. Ghost fishinglines did not alter the feeding, sheltering, the agonistic org/10.7717/peerj.1861 e1861. and non-occupational swimming behavior of S. fuscus (ESM Fig. 12). Altmann, J., 1974. Observational study of behavior: sampling methods. Behaviour 49, This is a generalist species, capable to explore a great variety of food and 227–266. https://doi.org/10.1163/156853974X00534. Alvarez-Filip, L., Paddack, M.J., Collen, B., Robertson, D.R., Cotˆ e,´ I.M., 2015. territorial resources (Feitosa et al., 2012; Ferreira et al., 1998). We Simplification of Caribbean reef-fish assemblages over decades of coral reef observed an opportunistic behavior of feeding on macroalgae and degradation. PloS One 10, 1–14. https://doi.org/10.1371/journal.pone.0126004.

6 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129

Anderson, D.R., 2008. Model Based Inference in the Life Sciences: a Primer on Evidence. algal-dominated reefs. J. Sea Res. 70, 42–49. https://doi.org/10.1016/j. Springer, New York. seares.2012.03.006. Anderson, M., Gorley, R.N., Clarke, K.R., 2008. PERMANOVA + for PRIMER : by, Guide Ferreira, C.E.L., Goncalves, J.E.A., Coutinho, R., Peret, A.C., 1998. Herbivory by the to Software and Statistical Methods. PRIMER-E Ltd, Plymouth, UK. https://doi.org/ dusky damselfishStegastes fuscus (Cuvier, 1830) in a tropical rocky shore: effects on 10.1016/j.isatra.2014.07.008. the benthic community. J. Exp. Mar. Biol. Ecol. 229, 241–264. Asoh, K., Yoshikawa, T., Kosaki, R., Marschall, E.A., 2004. Damage to cauliflower coral Ferreira, C.M., Coni, E.O.C., Medeiros, D.V., Sampaio, C.L.S., Reis-Filho, J.A., Barros, F., by monofilamentfishing lines in Hawaii. Conserv. Biol. 18, 1645–1650. https://doi. Loiola, M., Nunes, Jos´e de Anchieta Cintra da Costa, 2015. Community structure of org/10.1111/j.1523-1739.2004.00122.x. shallow rocky shore fish in a tropical bay of the southwestern Atlantic. Braz. J. Bacheler, N.M., Shertzer, K.W., 2015. Estimating relative abundance and species richness Oceanogr. 63, 379–396. https://doi.org/10.1590/S1679-87592015074706304. from video surveys of reef fishes. Fish. Bull. 113, 15–26. https://doi.org/10.7755/ Fitt, W.K., 2012. Bleaching of the fire corals Millepora. Proc. 12th Int. Coral Reef Symp. FB.113.1.2. 9–13. Banaszak, A.T., Ayala-Schiaffino, B.N., Rodríguez-Roman,´ A., Enríquez, S., Iglesias- Garcia, T.M., Matthews-cascon, H., Frankling-Junior, W., 2008. Macrofauna associated Prieto, R., 2003. Response of millepora alcicornis (milleporina: milleporidae) to two with branching firecoral Millepora alcicornis (Cnidaria: Hydrozoa). Thalass. An Int. bleaching events at puerto morelos reef, Mexican caribbean. Rev. Biol. Trop. 51, J. Mar. Sci. 24, 11–19. https://doi.org/10.1590/S1679-87592009000200009. 57–66. https://doi.org/10.15517/rbt.v51i4.26429. Giglio, V.J., Luiz, O.J., Schiavetti, A., 2015. Recreational diver behavior and contacts Barros, F., Hatje, V., Figueiredo, M.B., Magalhaes,˜ W.F., Dorea,´ H.S., Emídio, E.S., 2008. with benthic organisms in the abrolhos national marine park , Brazil. Environ. The structure of the benthic macrofaunal assemblages and sediments characteristics Manag. 57, 637–648. https://doi.org/10.1007/s00267-015-0628-4. of the Paraguaçu estuarine system, NE, Brazil. Estuar. Coast. Shelf Sci. 78, 753–762. Giglio, V.J., Ternes, M.L.F., Mendes, T.C., Cordeiro, C.A.M.M., Ferreira, C.E.L., 2017. https://doi.org/10.1016/j.ecss.2008.02.016. Anchoring damages to benthic organisms in a subtropical scuba dive hotspot. Bellwood, D.R., Hughes, T.P., Folke, C., 2004. Confronting the coral reef crisis. Nature J. Coast Conserv. 21, 311–316. https://doi.org/10.1007/s11852-017-0507-7. 429, 827–833. Gilman, E., 2015. Status of international monitoring and management of abandoned, lost Bishara, A.J., Hittner, J.B., 2012. Testing the significance of a correlation with and discarded fishinggear and ghost fishing.Mar. Pol. 60, 225–239. https://doi.org/ nonnormal data: comparison of pearson, spearman, transformation, and resampling 10.1016/j.marpol.2015.06.016. approaches. Psychol. Methods 17, 399–417. https://doi.org/10.1037/a0028087. Gilman, E., 2016. Biodegradable fishing gear: part of the solution to ghost fishing and Bonin, M.C., Almany, G.R., Jones, G.P., 2011. Contrasting effects of habitat loss and marine pollution. Anim. Conserv. 19, 320–321. https://doi.org/10.1111/acv.12298. fragmentation on coral-associated reef fishes. Ecology 92, 1503–1512. https://doi. Gilman, E., Chopin, F., Suuronen, P., Kuemlangan, B., 2016. Abandoned, Lost and org/10.1890/10-0627.1. Discarded Gillnets and Trammel Nets. Methods to Estimate Ghost Fishing Mortality, Brown, D., Edmunds, P.J., 2013. Long-term changes in the population dynamics of the and the Status of Regional Monitoring and Management. Food and Agriculture Caribbean hydrocoral Millepora spp. J. Exp. Mar. Biol. Ecol. 441, 62–70. https://doi. Organization of the United Nations, Rome. org/10.1016/j.jembe.2013.01.013. Glynn, P.W., Enochs, I.C., 2011. Invertebrates and their roles in coral reef ecosystems. In: Brown, J., Macfadyen, G., 2007. Ghost fishing in European waters: impacts and Dubinsky, Z., Stambler, N. (Eds.), Coral Reefs: an Ecosystem in Transition. Springer, management responses. Mar. Pol. 31, 488–504. https://doi.org/10.1016/j. New York, USA, pp. 273–325. marpol.2006.10.007. Hannak, J.S., Kompatscher, S., Stachowitsch, M., Herler, J., 2011. Snorkelling and Carvalho, L.R.S., Barros, F., 2017. Physical habitat structure in marine ecosystems: the trampling in shallow-water fringing reefs: risk assessment and proposed meaning of complexity and heterogeneity. Hydrobiologia. https://doi.org/10.1007/ management strategy. J. Environ. Manag. 92, 2723–2733. https://doi.org/10.1016/ s10750-017-3160-0. j.jenvman.2011.06.012. Chase, T.J., Pratchett, M.S., Frank, G.E., Hoogenboom, M.O., 2018. Coral-dwelling fish Hatje, V., De Souza, M.M., Ribeiro, L.F., Eça, G.F., Barros, F., 2016. Detection of moderate bleaching susceptibility of coral hosts. PloS One 13, 1–20. https://doi.org/ environmental impacts of shrimp farming through multiple lines of evidence. 10.1371/journal.pone.0208545. https://doi.org/10.1016/j.envpol.2016.06.056. Cheal, A.J., MacNeil, M.A., Cripps, E., Emslie, M.J., Jonker, M., Schaffelke, B., Helfman, G.S., 1989. Threat-sensitive predator avoidance in dameselfish-trumpetfish Sweatman, H., 2010. Coral – macroalgal phase shifts or reef resilience : links with interactions. Behav. Ecol. Sociobiol. 24, 47–58. diversity and functional roles of herbivorous fishes on the Great Barrier Reef. Coral Hoeksema, B.W., Hermanto, B., 2018. Plastic nets as substrate for reef corals in Lembeh Reefs 1005–1015. https://doi.org/10.1007/s00338-010-0661-y. Strait, Indonesia. Coral Reefs 37, 631. https://doi.org/10.1007/s00338-018-1686-x. Chiappone, M., Dienes, H., Swanson, D.W., Miller, S.L., 2005. Impacts of lost fishinggear Hughes, T.P., Barnes, M.L., Bellwood, D.R., Cinner, J.E., Cumming, G.S., Jackson, J.B.C., on coral reef sessile invertebrates in the Florida Keys National Marine Sanctuary. Kleypas, J., van de Leemput, Ingrid A., Lough, J.M., Morrison, T.H., Palumbi, S.R., Biol. Conserv. 121, 221–230. https://doi.org/10.1016/j.biocon.2004.04.023. Nes, E.H. van, Scheffer, M., 2017. Coral reefs in the anthropocene. Coral Reefs Anthr. Clements, C.S., Rasher, D.B., Hoey, A.S., Bonito, V.E., Hay, M.E., 2018. Patial 546, 1–5. https://doi.org/10.1007/978-94-017-7249-5. andtemporal limits of coral macroalgal competition: the negative impactsof Hughes, T.P., Connolly, S.R., Eakin, C.M., Dietzel, A., McWilliam, M.J., Hoey, A.S., macroalgal density, proximity, and history of contact. Mar. Ecol. Ser. 586, 11–20. Pears, R.J., Liu, G., Skirving, W.J., Heron, S.F., Pratchett, M.S., Baird, A.H., Stella, J. Coni, E.O.C., Ferreira, C.M., de Moura, R.L., Meirelles, P.M., Kaufman, L., Francini- S., Hoogenboom, M.O., Torda, G., Kerry, J.T., 2018. Global warming transforms Filho, R.B., 2012. An evaluation of the use of branching fire-corals (Millepora spp.) coral reef assemblages. Nature 556, 492–496. https://doi.org/10.1038/s41586-018- as refuge by reef fish in the Abrolhos Bank, eastern Brazil. Environ. Biol. Fish. 96, 0041-2. 45–55. https://doi.org/10.1007/s10641-012-0021-6. Johnson, M.K., Holbrook, S.J., Schmitt, R.J., Brooks, A.J., 2011. Fish communities on Costa, P., Dorea,´ A., Mariano-Neto, E., Barros, F., 2015. Are there general spatial patterns staghorn coral: effects of habitat characteristics and resident farmerfishes. Environ. of mangrove structure and composition along estuarine salinity gradients. Estuar. Biol. Fish. 91, 429–448. https://doi.org/10.1007/s10641-011-9802-6. Coast. Shelf Sci. https://doi.org/10.1016/j.ecss.2015.08.014. Leal, I.C.S., Pereira, P.H.C., De Araújo, M.E., 2013. Coral reef fish association and Counsell, C.W.W., Donahue, M.J., Edwards, K.F., Franklin, E.C., Hixon, M.A., 2018. behaviour on the firecoral Millepora spp. in north-east Brazil. J. Mar. Biol. Assoc. U. Variation in coral-associated cryptofaunal communities across spatial scales and K. 93, 1703–1711. https://doi.org/10.1017/S0025315413000234. environmental gradients. Coral Reefs 37, 827–840. https://doi.org/10.1007/ Leal, I.C.S., de Araújo, M.E., da Cunha, S.R., Pereira, P.H.C., 2015. The influenceof fire- s00338-018-1709-7. coral colony size and agonistic behaviour of territorial damselfishon associated coral Cruz, I.C.S., Waters, L.G., Kikuchi, R.K.P., Leao,˜ Z.M.A.N., Turra, A., 2018. Marginal reef fish communities. Mar. Environ. Res. 108, 45–54. https://doi.org/10.1016/j. coral reefs show high susceptibility to phase shift. Mar. Pollut. Bull. 135, 551–561. marenvres.2015.04.009. https://doi.org/10.1016/j.marpolbul.2018.07.043. Leao,˜ Z.M.A.N., Minervino-Neto, A., Ferreira, B.P., Feitosa, C.V., Sampaio, C.L.S., Costa- Darling, E.S., Graham, N.A.J., Januchowski-Hartley, F.A., Nash, K.L., Pratchett, M.S., Sassi, C.F., Neves, E.G., Freire, F.A.M., Silva, G.-O.M., Strenzel, G.M.R., Wilson, S.K., 2017. Relationships between structural complexity, coral traits, and Sovierzoski, H.H., Oliveira, J.E.L., Mendes, L.F., Soares, M.O., Araujo, M.-E., reef fish assemblages. Coral Reefs 36, 561–575. https://doi.org/10.1007/s00338- Oliveira, M.D.M., Maida, M., Correia, M.D., Rosa, R.S., Sassi, R., Johnsson, R., 017-1539-z. Francini-Filho, R.B., Kikuchi, R.K.P., Leite, T.S., 2015. Monitoramento dos Recifes e de Souza, J.N., Nunes, F.L.D., Zilberberg, C., Sanchez, J.A., Migotto, A.E., Hoeksema, B. Ecossistemas Coralinos. In: Turra, A., Denadai, M.R. (Eds.), Protocolos Para o W., Serrano, X.M., Baker, A.C., Lindner, A., 2017. Contrasting patterns of Monitoramento de Habitats Bentonicosˆ Costeiros. Instituto Oceanografico´ da connectivity among endemic and widespread fire coral species (Millepora spp.) in Universidade de Sao˜ Paulo, Sao˜ Paulo, pp. 155–179. https://doi.org/10.7476/ the tropical Southwestern Atlantic. Coral Reefs 36, 701–716. https://doi.org/ 9788598729251. 10.1007/s00338-017-1562-0. Lewis, J.B., 1989. The ecology of Millepora - a review. Coral Reefs 8, 99–107. https:// de Weerdt, W.H., 1981. Transplantation experiments with Caribbean Millepora species doi.org/10.1007/BF00338264. (Hydrozoa, Coelenterata), including some ecological observations on growth forms. Lewis, J.B., 2006. Biology and Ecology of the hydrocoral Millepora on coral reefs. Adv. Bijdr. tot Dierkd. 51, 1–19. Mar. Biol. 50, 1–55. https://doi.org/10.1016/S0065-2881(05)50001-4. Del Monaco, C., Hay, M.E., Gartrell, P., Mumby, P.J., Diaz-Pulido, G., 2017. Effects of Link, J., Segal, B., Miguel, L., 2019. Abandoned, lost or otherwise discarded fishinggear ocean acidificationon the potency of macroalgal allelopathy to a common coral. Sci. in Brazil: a review. Perspect. Ecol. Conserv. 17, 1–8. https://doi.org/10.1016/j. Rep. 7, 1–10. https://doi.org/10.1038/srep41053. pecon.2018.12.003. Dixson, D.L., Hay, M.E., 2012. Corals chemically cue mutualistic fishes to remove Lokrantz, J., Nystrm, M., Norstrm, A.V., Folke, C., Cinner, J.E., 2009. Impacts of artisanal competing seaweeds. Science (80-. ) 338, 804–807. https://doi.org/10.1126/ fishing on key functional groups and the potential vulnerability of coral reefs. science.1225748. Environ. Conserv. 36, 327–337. https://doi.org/10.1017/S0376892910000147. Edmunds, P.J., 1999. The role of colony morphology and substratum inclination in the Lorders, F.L., Miranda, R.J., Nunes, J.A.C., Barros, F., 2018. Spongivory by fishes on success of Millepora alcicornis on shallow coral reefs. Coral Reefs 18, 133–140. Southwestern Atlantic coral reefs: No evidence of top-down control on sponge https://doi.org/10.1007/s003380050167. assemblages. Front. Mar. Sci. 5 https://doi.org/10.3389/fmars.2018.00256. Feitosa, J.L.L., Concentino, A.M., Teixeira, S.F., Ferreira, B.P., 2012. Food resource use Lyons, P.J., Arboleda, E., Benkwitt, C.E., Davis, B., Gleason, M., Howe, C., Mathe, J., by two territorial damsel fish (Pomacentridae: Stegastes) on South-Western Atlantic Middleton, J., Sikowitz, N., Untersteggaber, L., Villalobos, S., 2015. The effect of

7 T.M. Beneli et al. Marine Environmental Research 161 (2020) 105129

recreational SCUBA divers on the structural complexity and benthic assemblage of a Plass-Johnson, J.G., Ferse, S.C.A., Jompa, J., Wild, C., Teichberg, M., 2015. Fish Caribbean coral reef. Biodivers. Conserv. 24, 3491–3504. https://doi.org/10.1007/ herbivory as key ecological function in a heavily degraded coral reef system. Limnol. s10531-015-1009-2. Oceanogr. 60, 1382–1391. https://doi.org/10.1002/lno.10105. Marshall, P.A., Baird, A.H., 2000. Bleaching of corals on the Great Barrier Reef : Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists, differential susceptibilities among taxa. Coral Reefs 19, 155–163. Experimental Design and Data Analysis for Biologists. https://doi.org/10.1016/ Marshell, A., Mumby, P.J., 2015. The role of surgeon ish (Acanthuridae) in maintaining S0022-0981(02)00278-2. algal turf biomass on coral reefs. J. Exp. Mar. Biol. Ecol. 473, 152–160. https://doi. R Core Team, 2017. R: A Language and Environment for Statistical Computing. R Found. org/10.1016/j.jembe.2015.09.002. Stat. Comput, Vienna. Matsuoka, T., Nakashima, T., Nagasawa, N., 2005. Review Article A review of ghost Shester, G.G., Micheli, F., 2011. Conservation challenges for small-scale fisheries: fishing: scientific approaches to evaluations and solutions. Fish. Sci. 691–702. bycatch and habitat impacts of traps and gillnets. Biol. Conserv. 144, 1673–1681. Medeiros, P.R., Rada, D.P., Rosa, R.S., 2014. Abundance and behavioural ecology of the https://doi.org/10.1016/j.biocon.2011.02.023. blenny Ophioblennius trinitatis (Teleostei: blenniidae) at an oceanic archipelago of Valderrama Ballesteros, L., Matthews, J.L., Hoeksema, B.W., 2018. Pollution and coral Brazil (Atlantic). Sci. Mar. 78, 203–212. https://doi.org/10.3989/ damage caused by derelict fishing gear on coral reefs around Koh Tao, Gulf of scimar.03979.30G. Thailand. Mar. Pollut. Bull. 135, 1107–1116. https://doi.org/10.1016/j. Montano, S., Reimer, J.D., Ivanenko, V.N., García-Hernandez,´ J.E., van Moorsel, G.W.N. marpolbul.2018.08.033. M., Galli, P., Hoeksema, B.W., 2020. Widespread occurrence of a rarely known Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S. Springer, ISBN 0- association between the hydrocorals stylaster roseus and millepora alcicornis at 387-95457-0. Bonaire. South./ Caribbean. Div. 12 https://doi.org/10.3390/D12060218. Wahle, C.M., 1980. Detection, pursuit, and overgrowth of tropical gorgonians by Munday, P.L., 2004. Habitat loss, resource specialization, and extinction on coral reefs. milleporid hydrocorals: perseus and Medusa revisited. Science (80-. ) 209, 689–691. Global Change Biol. 10, 1642–1647. https://doi.org/10.1111/j.1365- https://doi.org/10.1126/science.209.4457.689. 2486.2004.00839.x. Wilson, S.K., Graham, N.A.J., Pratchett, M.S., Jones, G.P., Polunin, N.V.C., 2006. Nunes, Jos´e de Anchieta Cintra da Costa, Medeiros, D.V., Reis-Filho, J.A., Sampaio, C.L. Multiple disturbances and the global degradation of coral reefs: are reef fishesat risk S., Barros, F., 2012. Reef fishes captured by recreational spearfishing on reefs of or resilient? Global Change Biol. 12, 2220–2234. https://doi.org/10.1111/j.1365- Bahia State, northeast Brazil. Biota Neotropica 12, 179–185. https://doi.org/ 2486.2006.01252.x. 10.1590/S1676-06032012000100014. Wood, S.N., 2006. Generalized Additive Models: an Introduction with R. Chapman Hall/ Pereira, P.H.C., Munday, P.L., 2016. Coral colony size and structure as determinants of CRC, Boca Raton, Florida, USA. https://doi.org/10.1198/jasa.2007.s188. habitat use and fitness of coral-dwelling fishes. Mar. Ecol. Prog. Ser. 553, 163–172. Wood, S.N., 2011. Fast stable restricted maximum likelihood and marginal likelihood https://doi.org/10.3354/meps11745. estimation of semiparametric generalized linear models. J. Roy. Stat. Soc. 73, 3–36. Pereira, P.H.C., Leal, I.C.S., De Araújo, M.E., Souza, A.T., 2012. Feeding association Wood, S.N., Pya, N., Safken,¨ B., 2016. Smoothing Parameter and Model Selection for between reef fishes and the fire coral Millepora spp. (Cnidaria: Hydrozoa). Mar. General Smooth Models. J. Am. Stat. Assoc. 111, 1548–1563. https://doi.org/ Biodivers. Rec. 5, 1–4. https://doi.org/10.1017/S1755267212000073. 10.1080/01621459.2016.1180986. Pereira, P.H.C., Leal, I.C.S., de Araújo, M.E., 2016. Observer presence may alter the Yoshikawa, T., Asoh, K., 2004. Entanglement of monofilament fishing lines and coral behaviour of reef fishes associated with coral colonies. Mar. Ecol. 37, 760–769. death. Biol. Conserv. 117, 557–560. https://doi.org/10.1016/j.biocon.2003.09.025. https://doi.org/10.1111/maec.12345. Zuur, A.F., Ieno, E.N., Walker, N.J., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects Models and Extensions in Ecology with R. Springer, Berlin, Germany. https://doi. org/10.1007/978-0-387-87458-6.