Received: 1 November 2019 | Revised: 3 February 2020 | Accepted: 12 February 2020 DOI: 10.1111/gcb.15048

PRIMARY RESEARCH ARTICLE

It’s a wormy world: Meta-analysis reveals several decades of change in the global abundance of the parasitic nematodes Anisakis spp. and Pseudoterranova spp. in marine fishes and invertebrates

Evan A. Fiorenza1 | Catrin A. Wendt1 | Katie A. Dobkowski1,2 | Teri L. King3 | Marguerite Pappaionou4 | Peter Rabinowitz4 | Jameal F. Samhouri5 | Chelsea L. Wood1

1School of Aquatic and Fishery Sciences, University of Washington, Seattle, WA, USA Abstract 2Department of Biology, Bates College, The Anthropocene has brought substantial change to ocean ecosystems, but whether Lewiston, ME, USA this age will bring more or less marine disease is unknown. In recent years, the ac- 3Washington Sea Grant, Seattle, WA, USA celerating tempo of epizootic and zoonotic disease events has made it seem as if 4Department of Environmental and Occupational Health Sciences, School of disease is on the rise. Is this apparent increase in disease due to increased observa- Public Health, University of Washington, tion and sampling effort, or to an actual rise in the abundance of parasites and patho- Seattle, WA, USA gens? We examined the literature to track long-term change in the abundance of two 5Conservation Biology Division, Northwest Fisheries Science Center, parasitic nematode genera with zoonotic potential: Anisakis spp. and Pseudoterranova National Oceanographic and Atmospheric spp. These anisakid nematodes cause the disease anisakidosis and are transmitted Administration, Seattle, WA, USA to humans in undercooked and raw marine seafood. A total of 123 papers published Correspondence between 1967 and 2017 met our criteria for inclusion, from which we extracted 755 Chelsea L. Wood, School of Aquatic and Fishery Sciences, University of Washington, host–parasite–location–year combinations. Of these, 69.7% concerned Anisakis spp. Seattle, WA, USA. and 30.3% focused on Pseudoterranova spp. Meta-regression revealed an increase Email: [email protected] in Anisakis spp. abundance (average number of worms/fish) over a 53 year period Funding information from 1962 to 2015 and no significant change in Pseudoterranova spp. abundance University of Washington Royalty Research Fund; University of Washington Innovation over a 37 year period from 1978 to 2015. Standardizing changes to the period of Award; Washington Sea Grant, University of 1978–2015, so that results are comparable between genera, we detected a signifi- Washington, Grant/Award Number: R/SFA/ PD-2; National Science Foundation, Grant/ cant 283-fold increase in Anisakis spp. abundance and no change in the abundance of Award Number: OCE-1829509; Alfred P. Pseudoterranova spp. This increase in Anisakis spp. abundance may have implications Sloan Foundation for human health, marine mammal health, and fisheries profitability.

KEYWORDS anisakiasis, anisakidosis, cod worm, dolphins, fish, herring worm, parasite burden, seals, whales, zoonoses

1 | INTRODUCTION have complex life cycles that connect disparate host species across a sprawling food web (Figure 1). People have interrupted most of these Fifty-eight percent of the infectious diseases affecting humans are zoo- transmission cycles in the developed world by disentangling human activ- notic, passing from animals to people (Woolhouse & Gowtage-Sequeria, ities from nature: our water is purified, our sewers direct waste away from 2005). Among the parasites that cause these zoonotic diseases, many water and food sources, and our food is processed. But some activities

2854 | © 2020 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/gcb Glob Change Biol. 2020;26:2854–2866. FIORENZA et al. | 2855

(a) FIGURE 1 (a) Anisakid nematodes in blue whiting (Micromesistius poutassou). Image by Gonzalo Jara and reproduced with permission via Shutterstock (stock photo ID 656259196). (b) Life cycle of Anisakis spp. Eggs from the adult worms are released in the feces of their cetacean definitive hosts and infect small crustaceans (e.g., krill) that consume the parasite larvae (Anderson, 2000). These first-intermediate crustacean hosts are then consumed by fish or squid, where they encyst and await trophic transmission to a marine mammal—and where they may then be incidentally ingested by humans (Acha & Szyfres, 2003). Anisakis spp. can proceed directly from the crustacean intermediate host to a marine mammal definitive (b) host, although they often use fish or cephalopod paratenic (i.e., “transport”) hosts (Hays et al., 1998). (c) Life cycle of Pseudoterranova spp. Eggs from the adult worms are released in the feces of their pinniped definitive hosts and infect microcrustaceans (e.g., copepods) that consume the parasite larvae (McClelland, 2002). Although Anisakis spp. can proceed directly from the crustacean intermediate host to the marine mammal definitive host, Psuedoterranova spp. must proceed through several additional life stages. First, the microcrustacean host is consumed by a macrocrustacean (e.g., amphipod, caprellid, shrimp, isopod, mysid), which is in turn consumed by a benthic-feeding fish (McClelland, 2002). Once a marine mammal consumes an (c) infected fish, Pseudoterranova spp. matures in the stomach, sexually reproduces, and releases eggs with the mammal's feces. Intermediate host fish may be consumed by larger fishes (paratenic hosts), where they encyst in fish tissues and await trophic transmission to a marine mammal. Vector images courtesy of Kim Kraeer, Lucy Van Essen- Fishman, Tracey Saxby, Joanna Woerner, Dieter Tracey, Jane Hawkey, Integration and Application Network, University of Maryland Center for Environmental Science (ian.umces.edu/imagelibra​ ry/)​

keep humans intimately connected with ecosystems, and therefore risk of raw, pickled, smoked, undercooked, or improperly frozen wild ensnaring people in parasite life cycles. marine seafood (Deardorff, Kayes, & Fukumura, 1991). The disease As one of the few zoonoses that can be transmitted to humans is caused by nematode parasites in the family Anisakidae (hereaf- from marine animals, anisakidosis is contracted through consumption ter, anisakids), most commonly the genera Anisakis (“herring worm” 2856 | FIORENZA et al. or “whale worm”; Mattiucci et al., 1997) and Pseudoterranova (“cod Past efforts to track change in parasite abundance over time worm” or “seal worm”; Paggi et al., 1991), which are globally dis- have used “vote-counting” meta-analytic approaches (Rudel, 2008). tributed and ubiquitous in the flesh of marine fishes (Figure 1a). For example, Ward and Lafferty (2004; see also Wood, Lafferty, & Eggs from the adult worms are released in the feces of their marine Micheli, 2010, appendix S1) used the proportion of total reports of a mammal definitive hosts (i.e., cetaceans for Anisakis, pinnipeds for taxon that addressed infectious disease (i.e., number of papers about Pseudoterranova), and infect small crustaceans that consume the par- disease in a taxon/total number of papers about that taxon) to as- asite eggs or larvae (Anderson, 2000; McClelland, 2002; Figure 1b,c). sess whether rates of disease in marine hosts had changed between These first-intermediate crustacean hosts are then consumed by 1970 and 2001. Fey et al. (2015) used a similar approach to show that, fishes, cephalopods, or other crustaceans, where they encyst and among animals, the number of mass mortalities caused by infectious await trophic transmission, either to another intermediate host, to disease increased between 1940 and 2000. However, because these a paratenic (i.e., “transport”) host, or to a marine mammal definitive approaches depend on counting the number of reports on a particu- host—and where they may then be incidentally ingested by humans lar topic, they are susceptible to literature bias (Fey et al., 2015). (Figure 1b,c; Acha & Szyfres, 2003). The worms cannot successfully Because we were interested in a single, well-studied parasite reproduce in the human intestinal tract, but they can survive there clade, we were able to perform a true meta-analysis, in which we ex- temporarily and cause substantial pathology (Bouree, Paugam, & tracted data from studies rather than tallying the number of studies Petithory, 1995). showing a particular effect. We sought to assess change in anisakid Human seafood consumers often discover their infection sta- nematode abundance over the past several decades. Although no tus when they find live worms in their phlegm, mucus, vomit, or study before ours has assessed temporal change in anisakid abun- feces (Acha & Szyfres, 2003), and in serious cases, symptoms in- dance at a global scale, thousands of studies have empirically quan- clude acute abdominal pain, nausea, vomiting, and diarrhea, which tified the abundance of anisakids in fish and invertebrate hosts at can persist for months (Acha & Szyfres, 2003; Bouree et al., 1995). specific times and locations across the world. We assimilated these Misdiagnosis is common, as symptoms resemble those of many studies into our meta-analysis, harnessing the statistical power of gastrointestinal ailments (Acha & Szyfres, 2003; Khan & Williams, many studies conducted in many regions across many host species. 2016; Shibata, Ueda, Akaike, & Saida, 2014; Shrestha et al., 2014). We specifically focused on studies of fish and invertebrate interme- Restaurants and supermarkets in the United States are required to diate and paratenic hosts, as they have been systematically assessed comply with regulations in the US Food and Drug Administration's for parasites by many prior studies. Food Code, which stipulates that fishery products must be frozen to There are several ecological differences between Anisakis spp. and kill anisakids before raw product can be sold or served to the public Pseudoterranova spp. that could cause the genera to exhibit divergent (FDA, 2009). Similar regulations are in effect in the European Union responses to global change. First, while Anisakis spp. use cetaceans as (European Commission, 2011). Nonetheless, existing evidence sug- definitive hosts, Psuedoterranova spp. use pinnipeds (Measures, 2014). gests that anisakid exposure is substantial on a global scale among If the trajectories of change in abundance for these two definitive host individuals who consume raw fish (Bao, Pierce, Pascual, et al., 2017; groups differ, or if the two parasite genera differ in their degree of de- Garcia-Palacios et al., 1996; Uga, Ono, Kataoka, & Hasan, 1996). pendence on the definitive host (e.g., if cetaceans are “life-cycle bottle- Even after worms have been killed by freezing, they can elicit gas- necks” for Anisakis spp., but pinnipeds are not “life-cycle bottlenecks” trointestinal symptoms if consumed (Audicana, Ansotegui, Corres, for Pseudoterranova spp., sensu Lafferty, 2012), the two genera might & Kennedy, 2002; Audicana & Kennedy, 2008; Daschner, Alonso- differ in their change in abundance over time. As a cetacean specialist Gomez, Cabanas, Suarez-de-Parga, & Lopez-Serrano, 2000) and with larvae that swim in the water column, Anisakis spp. are pelagic respiratory or dermatological symptoms through nonconsumptive (Measures, 2014; Smith & Wootten, 1978), whereas Pseudoterranova exposure, as in fishermen and fish-processing workers (Armentia spp.—pinniped specialists with larvae that sink and stick to benthic et al., 1998; Nieuwenhuizen et al., 2006; Purello-D’Ambrosio et al., habitat—are associated with demersal habitats (McClelland, 2002; 2000; Scala et al., 2001). Measures, 2014). Anthropogenic pressures affecting shallow, coastal Anisakidosis has been labeled an emerging zoonosis (Bao, Pierce, habitats (e.g., chemical pollution, nutrient pollution, coastal develop- Pascual, et al., 2017) due to recent upticks in reports of the disease ment) would therefore be likelier to influence Pseudoterranova spp. from Japan, New Zealand, coastal areas of Europe, the United States, than Anisakis spp. Finally, Anisakis spp. can proceed directly from the Canada, Brazil, Chile, and Egypt (Audicana et al., 2002; Audicana & crustacean intermediate host to a marine mammal definitive host, al- Kennedy, 2008; Chai, Murrell, & Lymbery, 2005; Hochberg, Hamer, though they often use fish or cephalopod paratenic (i.e., “transport”) Hughes, & Wilson, 2010). The rising incidence of anisakidosis could hosts (Figure 1b; Hays, Measures, & Huot, 1998; Measures, 2014), be due to: more accurate detection and diagnosis (Lohi et al., 2007), whereas Psuedoterranova spp. must infect a fish or cephalopod in- increased rates of consumption of raw or undercooked fish (Nawa, termediate host in order to complete its development (Figure 1c; Hatz, & Blum, 2005), or an increased abundance of the parasites in McClelland, 2002; Measures, 2014). We might therefore expect the commercially fished hosts. We sought to assess whether the risk Anisakis spp. to be more resilient to long-term biodiversity loss than of anisakidosis to human consumers has increased over time due to Pseudoterranova spp., since parasites with many obligately required increasing abundance of anisakid parasites in wild seafood. hosts are more likely to lose one of those hosts as biodiversity loss FIORENZA et al. | 2857 proceeds than are parasites with few obligately required hosts (Wood anisakids in humans, birds, or marine mammals and titles that clearly et al., 2015; Wood, Sandin, Zgliczynski, Guerra, & Micheli, 2014). indicated that the paper concerned a different suite of parasite spe- We conducted a meta-analysis of change over time (1967–2017) cies. After title screening, we retained 1,336 papers for the next in the abundance of anisakid nematodes in fish and invertebrate stages of the screening process. We then screened abstracts to hosts collected in any location around the world, finding that further focus the dataset, excluding publications that examined ani- Anisakis spp. increased over time while Pseudoterrnova spp. remained sakids in humans, birds, or marine mammals, that performed experi- unchanged. These findings have important implications or human mental manipulation of parasites or hosts, or that were reviews. This anisakidosis risk, as well as implications for marine mammal health process resulted in the retention of 576 papers, which were read in and fisheries profitability. full to determine their eligibility for inclusion in the study. To qualify for final inclusion, papers were required to contain information on host species identity, parasite genus identity, location of host collec- 2 | METHODS tion (either as a named place or latitude and longitude), collection year or year range, size of the host, how the hosts were examined 2.1 | Literature search and data extraction for anisakids, and either prevalence and intensity of infection with an error estimate or abundance with an estimate of error. Ultimately, To identify papers estimating anisakid abundance, we conducted a we extracted data from 123 papers, resulting in 755 data points search in ISI Web of Science. We used the search string: TS = (anisak* (i.e., unique estimates of parasite abundance for a host species in or “herring worm” or “herringworm” or Pseudoterranova or whale- a particular location at a particular time). Of these 755 data points, worm or “whale worm” or phocanema or “whale-worm”), on October 69.7% described Anisakis spp. (Figure 2b) and 30.3% described 10, 2017, which yielded 2,284 papers. We then used a systematic Pseudoterranova spp. (Figure 2c). For each data point, we extracted screening process to eliminate nonrelevant papers (Figure 2a). The information on host species identity, collection location, year of col- screening process was performed by EAF, CAW, and KAD. We lection, portion of the host examined (i.e., all, viscera, fillet, alimentary first screened titles for suitability, excluding papers that quantified tract), host examination method (i.e., standard visual assessment/

(a) (b)

(c)

FIGURE 2 (a) Flow diagram for the process of screening papers for inclusion in meta-analysis. For all full-text articles, we assessed eligibility for inclusion and ultimately retained 123 articles. From these 123 articles, we were able to extract 755 unique records of anisakid abundance. Geographic density of unique estimates of (b) Anisakis spp. and (c) Pseudoterranova spp. abundance from our meta-analytic dataset 2858 | FIORENZA et al. candling, microscopy, UV light, acid digestion), parasite genus (i.e., 2.3 | Data analysis Anisakis or Pseudoterranova), parasite prevalence (i.e., proportion of hosts infected), parasite intensity (i.e., average number of parasites To assess change over time in anisakid abundance, we developed per infected host), error associated with parasite intensity (i.e., stand- two parallel meta-analytic regression models for the two genera ard deviation, range, standard error, or confidence intervals), parasite of anisakid nematodes contained in our dataset (i.e., Anisakis and abundance (i.e., average number of parasites per host for all hosts), Pseudoterranova). For each model, we fourth-root transformed the and error associated with abundance (i.e., standard deviation, range, abundance and standard deviation of the abundance to meet nor- standard error, or confidence intervals). mality assumptions of the model. In our data, we had observations of anisakid abundance with associated error equal to zero. This arose if only a single host was examined, if reported anisakid abun- 2.2 | Data standardization dance was zero, or if anisakid abundance was greater than zero, but all the fish that were sampled had the same anisakid burden. Since We standardized data prior to analysis. For example, to standard- meta-regression uses the inverse of variance to weight each ob- ize host length, which was reported in different ways (i.e., standard servation, we corrected the variance to account for observed zero length, total length, fork length) across studies, we used the stand- error. We chose to add 1 to every variance to prevent overweight- ard linear conversion equation, ing (i.e., adding a small value like 0.000001 to every zero-error observation will overweight these observations). Adding a very = a+b ∗ Lengthstandard Lengthreported. small variance correction would give many orders of magnitude more weight to observations with zero variance compared to any We obtained the a and b parameters for each fish species from observation that had greater than 1 variance. By adding 1 to the FishBase (Froese & Pauly, 2000), using the average a and b if multiple variance, we prevent this issue and observations with zero error values were provided by FishBase. get full weight while all other observations receive progressively Parasite abundance was quantified as the number of parasites less weight. We included year of collection and host length as per host, including uninfected hosts (Bush, Lafferty, Lotz, & Shostak, moderators in the meta-regression model. We also included host 1997). When parasite abundance was not reported, but parasite species, portion of the host examined nested in host species, FAO intensity and prevalence were, we calculated parasite abundance major fishing area (to account for geographic clustering of points), by multiplying intensity and prevalence and propagating the error method for detecting parasites, and paper ID as random effects through in quadrature (the square root of the sum of squares). If using the rma.mv() function in the package metafor (Viechtbauer, the range of intensity or abundance was provided but the standard 2010) in R. The final models took the form: deviation was not, we estimated standard deviation by optimizing 1∕4 1∕4 the negative binomial distribution for the dispersion parameter. We Anisakis_abundanceijkl or Pseudoterranova_abundanceijkl assumed that the maximum abundance or intensity was the 95th ∼ host_lengthijkl +yearijkl + 1 host_speciesijklportion_of_fishijkl quantile of the negative binomial distribution, since parasites often  + 1 FAO_regionijkl+ 1 method_of_detectionijkl follow a negative binomial distribution (Shaw, Grenfell, & Dobson,   + 1998). With this assumption, we used the supplied mean intensity 1 manuscript_IDijkl,  or abundance as the mean of the negative binomial distribution with the dispersion parameter unknown. We then used an optimi- where the response variableijkl represents a measurement of parasite zation algorithm to estimate the dispersion parameter that best fit abundance from the ith study in the jth location at the kth time in the the supplied mean and 95th percentile. With the given mean and lth host species. estimated dispersion parameter, we were able to calculate the error We were also interested in testing which fish species, geographic of the mean given that the variance of a negative binomial distri- regions, detection techniques, and portions of host examined drove bution is the sum of the mean and the squared mean divided by the patterns observed in the above meta-regression, so we per- the dispersion parameter. We also converted other forms of error formed four subanalyses. First, we sequentially excluded each host reported in the studies (i.e., standard error, confidence intervals) species and reran the model. We extracted an updated estimate of back to standard deviation. To convert from standard error to stan- change over time and compared the estimate for the effect of time dard deviation, we multiplied the standard error by the square root to that of the full model, which included all host species. This al- of the sample size. To convert from a confidence interval, we took lowed us to determine whether a single host species was driving the the difference between the upper bound of the confidence interval patterns we observed. We then performed the same analysis, but and the mean and then divided that difference by the appropriate instead sequentially excluded each geographic region to determine z-score and multiplied by the square root of the sample size. For whether a single geographic region was driving the patterns we ob- location information, we grouped data points based on major FAO served. To test the influence of detection technique, we excluded fishing region (FAO, 2008) using ESRI ArcGIS (ESRI, 2011), to match studies that used techniques that came into common usage over the our data to FAO major fishing region. study period: digestion of fish tissues with acids and use of UV light FIORENZA et al. | 2859 for parasite detection. We sequentially excluded (a) all studies that that controls for host length; estimate = +0.0206, SE = 0.0069, used digestion, including digestion combined with other techniques, z = +2.9922, n = 526, p = .0028) in abundance of Anisakis spp. like microscopy and (b) all studies that used UV, including UV com- (Table 1a; Figure 3a) and no change (estimate = –0.0019, SE = 0.0093, bined with other techniques. We then extracted an updated esti- z = –0.2063, n = 229, p = .8366) in abundance of Pseudoterranova mate of change over time and compared the estimate for the effect spp. (Table 1b; Figure 3b). For Anisakis spp., we detected a 283- of time to the full model that included all detection techniques. fold increase in abundance across the standardized time period of Finally, to test the influence of the portion of host examined, we se- 1978–2015. Change in abundance over this time frame is measur- quentially excluded all studies that quantified anisakid burden in the able for both anisakid genera; thus, constraining our effect estimates (a) alimentary tract, (b) fillet, (c) viscera, and (d) whole body. We then to this period allows for fair comparison of results between anisakid extracted an updated estimate of change over time and compared genera. In terms of fish length-corrected modeled abundance, we the estimate for the effect of time to the full model that included all observed an increase from an average of less than 1 anisakid per categories of portion of host examined. 100 hosts examined (mean = 0.0037, 95% CI = 0.0017–0.2328) to more than 1 anisakid in every host examined (mean = 1.0356, 95% CI = 0.1273–4.0681) over the time period for which we have data 3 | RESULTS from both parasite genera (1978–2015). We evaluated the influence of high-leverage data points, as well We obtained 755 unique host–parasite–location–year combina- as the influence of host species, geographic region, and detection tions (i.e., data points; Anisakis spp. n = 526, Pseudoterranova spp. technique on these trends. For the Anisakis spp. analysis, the record n = 229) across 123 papers, 215 host species, 53 years, and four from 1962 was a temporal outlier, but did not qualitatively alter our oceans. A total of 56,778 fish were examined across all studies, re- results (with the 1962 record, the effect of time is +0.0206 with a sulting in enumeration of 446,615 anisakid nematodes. For Anisakis standard error of 0.0069, p = .0028; when the record from 1962 is ex- spp., data were obtained across 15 FAO regions and 182 host spe- cluded, the effect of time is +0.0222 with a standard error of 0.0070, cies (Figure 2a); 31 data points were obtained by standard visual as- p = .0017). Sequentially removing single host species from our model sessment/candling, 25 by standard visual assessment/candling plus did not change our results either, as the effect of time was always sig- microscopy, 69 by digestion, 13 by digestion plus microscopy, eight nificantly positive for Anisakis spp. (Figure S1). Despite not changing by digestion plus UV, 363 by microscopy, 10 by microscopy plus the overall temporal patterns, the host species that leveraged the larg- UV, and seven by UV. For Pseudoterranova spp., data were obtained est absolute change to the temporal effect size for Anisakis spp. was across nine FAO regions and 92 host species (Figure 2b); seven data the black scabbardfish, Aphanopus carbo; the exclusion of this species points were obtained by standard visual assessment/candling, nine nonsignificantly decreased the effect of time (without A. carbo, the by standard visual assessment/candling plus microscopy, four by di- effect of time is +0.0169 with a standard error of 0.0068, p = .0132). gestion plus microscopy, 206 by microscopy, and three by micros- Sequentially removing individual geographic regions from our model copy plus UV. also did not substantially change our conclusions. The effect of time In our meta-regression models, we detected an increase through for Anisakis spp. was always significantly positive, except when we ex- time (all estimates = regression coefficients for the effect of time cluded FAO region 27 (northeastern Atlantic), which resulted in the on fourth-root transformed anisakid abundance, from a model effect becoming marginally significant (without region 27, the effect

TABLE 1 Regression coefficients from meta-regression of fourth-root transformed parasite abundance for (a) Anisakis spp. and (b) Pseudoterranova spp. Year of collection and host length are included as moderators in the meta-regression model. We also included host species, portion of the host examined nested in host species, FAO major fishing area (to account for geographic clustering of points), method for detecting parasites, and paper ID as random effects. Models were run using the rma.mv() function of the package metafor (Viechtbauer, 2010). We calculated pseudo-R2 for each model as the amount of variability captured by the full model (including moderators year of collection and host length + random effects) that was not captured by the null model (including only random effects). Pseudo-R2 = .14 for the Anisakis spp. model and .00 for the Pseudoterranova spp. model

Estimate SE z-value p-value CI lower CI upper

(a) Anisakis spp. Intercept −40.5379 13.7751 −2.9428 .0033 −67.5366 −13.5392 Year 0.0206 0.0069 2.9922 .0028 0.0071 0.0341 Host_length 0.0127 0.0021 6.1656 <.0001 0.0087 0.0168 (b) Pseudoterranova spp. Intercept 4.4327 18.5596 0.2388 .8112 −31.9435 40.8089 Year −0.0019 0.0093 −0.2063 .8366 −0.0202 0.0163 Host_length 0.0048 0.0048 0.9926 .3209 −0.0047 0.0142 2860 | FIORENZA et al.

FIGURE 3 (a) Change over time (1962–2015) in Anisakis spp. abundance with 95% confidence interval. Points represent extracted, length-corrected mean abundances. Shapes indicate which portion of the hosts was examined and color indicates host species (n host species = 182). (b) Change over time (1978–2015) in Pseudoterranova spp. abundance with 95% confidence interval for all hosts. Shapes indicate which portion of the hosts was examined and color indicates host species (n host species = 92) of time is +0.0140 with a standard error of 0.0072 and p = .0503; we found a long-term increase in the abundance of Anisakis spp. and no Figure S2). FAO region 27 encompassed a substantial proportion of long-term change in the abundance of Pseudoterranova spp. This finding the data in our Anisakis spp. dataset (123 data points of 526), but was implies a rising risk of anisakidosis for humans and cetaceans, and an only the second most data-rich FAO region in our dataset (Figure S2). unchanging risk of anisakidosis for pinnipeds; it also suggests that the For Pseudoterranova spp., we detected no change in abundance be- profitability and sustainability of fisheries could be compromised by the tween 1978 and 2015, and sequential exclusion of individual species rising rates of Anisakis spp. infection. (Figure S3) and geographic regions (Figure S2) did not alter this finding. On average, we estimated that Anisakis spp. abundance increased Finally, we also explored the potential influence of changing detection from less than 1 anisakid per 100 hosts in 1978 to more than 1 ani- techniques on the change in anisakid burden over time. Neither the sakid in every host examined in 2015. This pattern was not driven exclusion of studies using digestion as a technique for anisakid detec- by any single host species. The host that had the largest influence tion (without digestion, the effect of time is +0.0202 with a standard on the temporal increase in Anisakis abundance was the black scab- error of 0.0078, p = .0097) nor the exclusion of studies using UV light bardfish (A. carbo), but the influence of this one species did not drive (without UV, the effect of time is +0.0175 with a standard error of the temporal pattern observed across all hosts (Figure S1). One FAO 0.0068, p = .0107) eliminated the significant, positive trend over time region did appear to influence the outcome of our results; exclusion for Anisakis spp. Similarly, neither the exclusion of studies using diges- of data points from FAO Region 27 (northeastern Atlantic) caused tion (without digestion, the effect of time is –0.0045 with a standard the change over time in Anisakis spp. abundance to become margin- error of 0.0090, p = .6164) nor the exclusion of studies using UV light ally nonsignificant (p = .0503). In short, the increase in Anisakis spp. (without UV, the effect of time is –0.0015 with a standard error of abundance across the time frame of our study appears to be robust 0.0097, p = .8800) shifted the trend over time for Pseudoterranova spp. across host species, but the positive temporal trend we reveal here from nonsignificant to significant. was strongly driven by the northeastern Atlantic. The northeastern Atlantic FAO region encompasses the Barents, Norwegian, Baltic, North, and Irish Seas and represents a major 4 | DISCUSSION source of the data points on Anisakis spp. included in this study (n = 123 out of 526, the second most well-represented region in the From our analysis of 123 manuscripts and 755 data points summarizing study, after FAO region 37, the Meditteranean, and Black Seas). It 56,778 fish and 446,615 anisakid nematodes over almost four decades, is possible that temporal increases in anisakid burden are likelier FIORENZA et al. | 2861 to be detected where data are ample. It is also possible that recent spp. and Pseudoterranova spp. that might make data on one genus more increases in cetacean populations have been especially influential susceptible to literature bias than data on the other. For example, the in the northeastern Atlantic. This region is dominated by Anisakis two genera are morphologically similar (Hurst, 1984), but Anisakis spp. simplex sensu stricto, whose definitive hosts include minke whales might be slightly more challenging to detect than Pseudoterranova spp., (Balaenoptera acutorostrata), common dolphins (Delphinus delphis), given that Anisakis spp. larvae tend to be small and translucent while long-finned pilot whales (Globicephala melas), white-beaked dol- Pseudoterranova spp. larvae are large and dark in color (Measures, phins (Lagenorhynchus albirostris), killer whales (Orcinus orca), and 2014; Smith & Wootten, 1978). If detection techniques have improved striped dolphins (Stenella coeruleoalba; Bilska-Zajac et al., 2015; over time, this could lead to an apparent increase in abundance for Mattiucci et al., 2014, 2018; Pierce et al., 2018). Some of these Anisakis spp. and little change for Pseudoterranova spp. However, we cetacean species have increased in abundance since International explored the influence of detection technique on our results, and Whaling Commission regulations came into effect in the mid-1980s our findings show that no one detection technique drove the change (e.g., Hammond et al., 2013; Murphy, Pinn, & Jepson, 2013; Skaug, over time in Anisakis spp. burden. Additionally, we anticipated that, Oien, & Bothun, 2004). Other world regions have probably also ex- because both Anisakis spp. and Pseudoterranova spp. pose threats to perienced increases in marine mammal abundance due to legisla- human health, there would be roughly equal incentives for scientific tive or regulatory protections enacted within the timeframe of our study of the two genera (Buchmann & Mehrdana, 2016). However, study, but the northeastern Atlantic has a particularly long history because Anisakis spp. tends to be more pathogenic in human hosts of intense exploitation of marine mammals (Clapham & Link, 2006); than is Psuedoterranova spp., it might be the case that there is more relaxation of this pressure might therefore have produced an es- incentive for the study of Anisakis spp., and this could drive an increas- pecially robust increase over time in Anisakis spp. Although the in- ing number of reports over time of high Anisakis spp. burdens. We are crease over time that we detected was strongly influenced by this unable to test this hypothesis with our current dataset, but we encour- region, the positive relationship between time and Anisakis burden age others to ground-truth the patterns we document here using, for remained marginally significant (p = .0503) when data points from example, estimates of parasite burden in fluid-preserved fishes held the northeastern Atlantic region were excluded, suggesting that this in natural history collections (Harmon, Littlewood, & Wood, 2019; temporal increase is not a phenomenon unique to that region. Howard, Davis, Lippert, Quinn, & Wood, 2019) or long-term monitor- In contrast, we did not detect a similar increase over time in the ing of Anisakis spp. burdens at individual sites. abundance of Pseudoterranova spp. Because it is ecologically similar Our study is correlational and precludes conclusions regarding to Anisakis spp. and subject to many of the same long-term environ- causation, but we have developed two hypotheses that could si- mental changes (e.g., climate change), we expected to observe simi- multaneously explain the dramatic increase in Anisakis spp. and the lar patterns between the two genera. Finding no significant change lack of change in Pseudoterranova spp. First, as complex life-cycle in Pseudoterranova spp. could indicate either that this genus is unaf- parasites, anisakids can respond to changes in the abundance of fected by the changes experienced by ocean ecosystems over the any intermediate host (copepod, krill, fish), paratenic host (fish, past few decades, or that the factors determining its abundance are cepahlopod), or definitive host (marine mammal; Arneberg et al., responding to global ocean change antagonistically, such that any 1998). We do not know which host is the most important determi- gain in abundance (e.g., from increases in pinniped populations), is nant of anisakid abundance (i.e., which host serves as the life-cycle counteracted by a negative effect on abundance (e.g., decreases in bottleneck; sensu Lafferty, 2012), but we do know that several ani- larval parasite survival with increasing temperatures). Excluding in- sakid hosts have changed in abundance over the past half-century. dividual host species, geographic regions, and detection techniques Since 1972, all marine mammals have been protected in the United did not alter the estimate of the change over time for the abundance States under the Marine Mammal Protection Act and many coun- of Pseudoterranova spp. tries adhere to the moratorium on commercial whaling imposed The fact that we observed different temporal patterns between by the International Whaling Commission in 1982. This protection the Anisakis and Pseudoterranova genera gives us some confidence has allowed for the recovery of many cetacean and pinniped pop- that these patterns are not driven exclusively by literature bias. All ulations (Magera, Flemming, Kachner, Christensen, & Lotze, 2013). meta-analyses are susceptible to literature bias (Borenstein, Hedges, Marine mammal-driven increases in anisakids have been observed Higgins, & Rothstein, 2009), and we envisioned that improvements at small spatial scales; for example, fish collected near seal haul-out in anisakid detection capacity, increasing interest in anisakids as a sites have been found to harbor a greater number of Pseudoterranova research subject, or an increase in “file-drawer” bias over time could spp. than fish collected near non-haul-out sites (Jensen & Idas, 1992; generate artifactual temporal increases in anisakids. However, any of Marcogliese & McClelland, 1991). As many marine mammal popula- these sources of literature bias should apply roughly equally to both tions recover their former abundance, the increased number of de- Anisakis spp. and Pseudoterranova spp. To observe a dramatic increase finitive hosts could support larger parasite populations. One of the in one genus and no change in the other suggests that literature bias biggest differences between Anisakis spp. and Pseudoterranova spp. is not entirely responsible for the temporal change observed, and that concerns their use of cetaceans and pinnipeds as definitive hosts, the increase in Anisakis spp. is therefore likely to reflect an actual tem- respectively (Figure 1b,c). Perhaps it is the case that definitive mam- poral change in nature. There are some differences between Anisakis malian hosts are an important life-cycle bottleneck for Anisakis spp. 2862 | FIORENZA et al. only, and that increases in cetacean populations have alleviated a bottleneck for Anisakis spp. and not for Pseudoterranova spp., or if bottleneck for Anisakis spp., whereas increases in pinniped popula- only the crustacean hosts to which Anisakis spp. is specific respond to tions have not done the same for Pseudoterranova spp. While this nutrient pollution with increases in abundance, then this could lead may be perceived as a recent increase in anisakid abundance, our to temporal increases in the abundance of Anisakis spp. and stasis data cover only the period of time since the enforcement of the for Pseudoterranova spp. However, if nutrient pollution is influential, Marine Mammal Protection Act in 1972, and thus the increase we one might expect it to drive a pattern opposite to the one actually observe in Anisakis spp. abundance could be a return to the histor- observed: nutrient-driven increases in crustacean first intermediate ical anisakid abundance levels present before exploitation-driven hosts should benefit coastal anisakids (i.e., Pseudoterranova spp., declines in marine mammal abundance. Second, it is possible that, which uses pinniped hosts and is associated with the host's coastal because Anisakis spp. marine mammal hosts (i.e., cetaceans) tend to habitat), rather than pelagic anisakids (i.e., Anisakis spp., which uses be vagile and migratory and Pseudoterranova spp. marine mammal cetacean hosts and is associated with their pelagic habitat). hosts (i.e., pinnipeds) tend to remain within a limited home range, Whatever the drivers of the increase in anisakid abundance, such it may be easier to detect broad-scale change in the abundance of an increase could have substantial economic, ecological, and human Anisakis spp. than in Pseudoterranova spp. Our meta-analysis was not health impacts on a global scale. First, increases in anisakid abun- geographically finely resolved (Figure 2b,c), and could have missed dance could increase the risk of anisakidosis in people (Bao, Pierce, areas where Pseudoterranova spp. abundances are changing. Long- Pascual, et al., 2017). The more anisakid worms that are present in term monitoring of individual sites is the most direct way to detect commercially exploited fish and invertebrates, the greater the like- temporal change in Pseudoterranova spp., if these changes are only lihood that the resulting seafood meal contains a worm. While the detectable in the geographic areas where pinnipeds congregate symptoms of anisakidosis are rarely life-threatening, increased in- (i.e., Jensen & Idas, 1992; Marcogliese & McClelland, 1991). cidence of anisakidosis could increase hospitalizations, represent- Other factors might also have driven the temporal increase in ing a burden on public health (Bouree et al., 1995). Occasionally, Anisakis spp., but it is not as clear how they would simultaneously people exposed to anisakid worms will develop hypersensitivity to leave the abundance of Pseudoterranova spp. unaffected. For ex- worm antigens, leading to severe anaphylactic reactions upon fur- ample, as fisheries exploitation has increased over the past sev- ther anisakid exposures (Alonso, Daschner, & Moreno-Ancillo, 1997; eral decades, intermediate fish host density may be on the decline Audicana et al., 2002; Audicana & Kennedy, 2008; Daschner et al., (Christensen et al., 2014). If fish hosts are not a transmission bot- 2000; del Pozo et al., 1996). These allergic reactions can be initiated tleneck (sensu Lafferty, 2012), and reduced fish density leads to by both live and dead worms (Alonso et al., 1997; Audicana et al., a relaxation in competition for krill and an increase in the number 2002; Audicana & Kennedy, 2008; Daschner et al., 2000) and even of krill intermediate hosts consumed per fish, a reduction in fish through nonconsumptive exposure, as in fishermen and fish-pro- density could actually result in an increase in the per-fish burden cessing workers (Armentia et al., 1998; Nieuwenhuizen et al., 2006; of anisakid parasites, as worms become more concentrated in the Purello-D’Ambrosio et al., 2000; Scala et al., 2001). Allergies to ani- few remaining fish hosts. However, given that both Anisakis spp. and sakids are regularly misdiagnosed as allergies to fish or shellfish (Acha Pseudoterranova spp. are host generalists for their fish intermediate/ & Szyfres, 2003; del Pozo et al., 1996) and, in some regions, anisakid paratenic hosts (Buchmann & Mehrdana, 2016), this would not ex- allergy may be more common than true seafood allergy (Kimura, plain the dramatic increase in one genus and the lack of temporal 2001). Such severe anaphylactic reactions are a grave and growing change in the other. Similarly, long-term climate change could also concern, given the increasing abundance of Anisakis spp. worms in influence the abundance of anisakid nematodes, as increases in commercially exploited marine fish species that we reveal here. temperature can increase host susceptibility to infection by compro- In addition to the threat they present to public health, anisakids mising the ability of fish hosts to immunologically or behaviorally can also cause disease in their definitive marine mammal hosts. resist infection (Burge et al., 2014; Harvell et al., 2002). Increasing Anisakids are commonly found in necropsied marine mammals, in- temperatures can also facilitate faster growth and shorter gener- cluding species that are endangered (Dailey & Stroud, 1978; Gibson ation times in aquatic parasites (Marcogliese, 2001a, 2001b). This et al., 1998). The presence of anisakids in marine mammals can cause could lead to increases in Anisakis spp. and not Pseudoterranova spp. gastric obstruction, ulceration, and, in some cases, death (Geraci & if Anisakis spp. are able to capitalize on their hosts’ compromised St. Aubin, 1987; Spraker, Lyons, Tolliver, & Bair, 2003; Young & Lowe, immunity, while Pseudoterranova spp. are not, or if increasing tem- 1969). These worms also divert energy that would otherwise be used perature shortens the generation times of Anisakis spp. while leav- for host survival, growth, and reproduction (Dailey & Stroud, 1978); ing those of Pseudoterranova spp. unaffected. Finally, as agriculture for example, intestinal parasites (including anisakids) may be hamper- and logging intensify, nutrients are released into coastal ecosystems ing the recovery of endangered populations of killer whales (Krahn (e.g., Cuo, Lettenmaier, Alberti, & Richey, 2009), which can fuel phy- et al., 2002). Anisakid infections have also been empirically linked to toplankton blooms (e.g., Beman, Arrigo, & Matson, 2005) that in a mass mortality of sea otters in Cordova, Alaska (Ballachey, Gorbics, turn increase the abundance of copepods and other zooplanktonic & Doroff, 2002). Although our study suggests that pinnipeds are not crustaceans (e.g., Siokou-Frangou & Papathanassiou, 1991; Uriarte & at risk of increasing rates of anisakidosis (because the abundance of Villate, 2004). If crustacean first intermediate hosts are a life-cycle Pseudoterranova spp. remained unchanged throughout the time period FIORENZA et al. | 2863 of our study), cetaceans now face a 283-fold greater risk of contracting studies of relatively short temporal duration is their inability to capture Anisakis spp. from consuming the fishes assessed in our meta-analysis, short-term or small spatial-scale fluctuations in the response variable and this risk could be an important threat to the population viability of of interest. For example, empirical studies in the Gulf of St. Lawrence, these vulnerable and—in some cases—endangered species. Canada document an increase in Pseudoterranova decipiens from 1981 Finally, anisakids present a threat to the health of their fish in- to 1990, followed by a precipitous decline from 1990 to 1992; abun- termediate hosts and associated fisheries. In Atlantic salmon (Salmo dances of A. simplex remained stable during this time period (Boily & salar), red vent syndrome—which causes fish vents to become red, Marcogliese, 1995; Marcogliese, 2001a, 2001b; McClelland & Martell, swollen, hemorrhaging, and prolapsed—is attributed to A. simplex 2001). While the long-term increase was attributed to increases in infection (Beck et al., 2008; Noguera et al., 2009), and can in some gray seal (Halichoerus grypus) populations, the short-term decline was cases lead to fatal opportunistic secondary infection by pathogenic attributed to declining water temperatures between 1986 and 1994, bacteria (Beck et al., 2008). Stomach crater syndrome in Atlantic which might have inhibited the development and hatching of P. de- cod (Gadus morhua)—which causes damage to the stomach wall and cipiens eggs (Boily & Marcogliese, 1995; Marcogliese, 2001a, 2001b; mucosa—is also attributed to A. simplex (Levsen & Berland, 2012). McClelland & Martell, 2001). This vignette highlights that the temporal Anisakids reduce swimming efficiency of fish when encysted in the window selected for a meta-analytic study can strongly influence the musculature (Sprengal & Luchtenberg, 1991), reducing survival by perception of whether anisakid burdens are increasing or decreasing, increasing the likelihood that the fish will be consumed a predator and suggests that the long-term increase in anisakid burden docu- (Persson, 1991). Anisakids also have the potential to alter public per- mented here could be reversed if conditions change. ceptions of seafood safety (Karl, 2008), reducing the market value of Although disease ecologists suspect a long-term increase in the a vast number of commercially fished species (Bao, Pierce, Strachan, frequency and severity of infectious disease outbreaks due to ongoing et al., 2017). For example, in 1987 German news broadcasters dis- global change, few sources of data exist to test this hypothesis. The played a video of living anisakids nematodes emerging from a filet meta-analysis we conducted yielded temporally resolved, long-term of fish, driving an immediate 80% drop in seafood sales and loss of data on the abundance of two ecologically and economically import- fishery-based jobs in Germany (Karl, 2008). In Atlantic Canada, ani- ant parasites, and revealed a long-term increase in the abundance of sakid-related downgrading and discard has been estimated to cost Anisakis spp. and long-term stasis in the abundance of Pseudoterranova processors US$26.6–50 million annually (McClelland, 2002). For spp. However, this is the story of only two parasite species among Pacific cod (Gadus macrocephalus), it is estimated that monitoring millions that are extant, and we encourage others to use historical and detection of anisakids constitutes up to 50% of production costs ecology approaches (e.g., meta-analysis, parasitological dissection of (Choudhury & Bublitz, 1994). In Spain, a majority of consumers are museum specimens) to track change across a diversity of marine para- willing to pay more for fish that is guaranteed to be free of anisakids site species. Only then will we have the data to indicate whether con- (Bao, Pierce, Strachan, et al., 2017). Rising abundances of Anisakis temporary oceans are facing a “rising tide” of marine disease. spp. therefore threaten the viability of fish populations and the prof- itability of the fisheries that depend upon them. ACKNOWLEDGEMENTS Like all meta-analyses, our study is limited in its temporal scope The authors thank Hyejoo Ro for assistance with data extrac- by the available literature. We used the database ISI Web of Science tion. E.A.F. was supported by a Program Development Grant from to survey peer-reviewed publications, and found that the available lit- Washington Sea Grant. C.L.W. was supported by a grant from the erature is heavily weighted to the past 50 years (Figure S4). This con- National Science Foundation (OCE-1829509), a Sloan Research strains meta-analyses to near history (~1960s to present), by which Fellowship from the Alfred P. Sloan Foundation, a UW Innovation time the oceans had already been drastically altered by human activity Award, and the UW Royalty Research Fund. (Lotze et al., 2006). This limitation prevents meta-analyses from char- acterizing ocean ecosystems as they were before they were impacted DATA AVAILABILITY STATEMENT by fishing, pollution, and climate change (Jackson, 2001; Jackson et al., The data that support the findings of this study are openly avail- 2001). This lack of an appropriate “baseline” should be acknowledged able in the Dryad Digital Repository at https://doi.org/10.5061/ alongside the results presented here. While we did detect an increase dryad.kwh70rz0z. The code used to produce the statistical results in Anisakis spp. over the past several decades, does this increase rep- reported herein is openly available via GitHub at www.github.com/ resent a rise in infection, or a recovery of anisakids to some pre-impact wood-lab/Fiorenza_et_al_2020_GCB.​ baseline? In other words, are anisakid abundances increasing in re- sponse to human impacts on the environment (e.g., fishing, pollution, ORCID climate change), or are they recovering alongside their exploited marine Chelsea L. Wood https://orcid.org/0000-0003-2738-3139 mammal hosts? Our data cannot discriminate among these possibili- ties, but parasitological analysis of natural history collections (Harmon REFERENCES et al., 2019; Howard et al., 2019) could reveal trajectories of anisakid Acha, P. N., & Szyfres, B. (2003). Anisakiasis. In M. R. Periago (Ed.), changes over the past century or more, illuminating patterns across Zoonoses and communicable diseases common to man and animals (pp. 231–236). Washington, DC: Pan American Health Organization. a broader span of time. Finally, another weakness of meta-analytic 2864 | FIORENZA et al.

Alonso, A., Daschner, A., & Moreno-Ancillo, A. (1997). Anaphylaxis Bush, A. O., Lafferty, K. D., Lotz, J. M., & Shostak, A. W. (1997). Parasitology with Anisakis simplex in the gastric mucosa. New England Journal of meets ecology on its own terms: Margolis et al revisited. The Journal Medicine, 337, 350–351. of Parasitology, 83(4), 575–583. https://doi.org/10.2307/3284227 Anderson, R. (2000). Nematode parasites of vertebrates: Their development Chai, J.-Y., Murrell, K. D., & Lymbery, A. J. (2005). Fish-borne parasitic and transmission (2nd ed.). Wallingford, UK: CAB International. zoonoses: Status and issues. International Journal for Parasitology, 35, Armentia, A., Lombardero, M., Callejo, A., Martin-Santos, J., Martin 1233–1254. https://doi.org/10.1016/j.ijpara.2005.07.013 Gil, F., Vega, J., … Martinez, C. (1998). Occupational asthma by Choudhury, G., & Bublitz, C. (1994). Electromagnetic method for detec- Anisakis simplex. Journal of Allergy and Clinical Immunology, 102, tion of parasites in fish. Journal of Aquatic Food Product Technology, 3, 831–834. 49–63. https://doi.org/10.1300/J030v​03n01_06 Arneberg, P., Skorping, A., Grenfell, B., & Read, A. F. (1998). Host densities Christensen, V., Coll, M., Piroddi, C., Steenbeek, J., Buszowski, J., & Pauly, as determinants of abundance in parasite communities. Proceedings D. (2014). A century of fish biomass decline in the ocean. Marine of the Royal Society of London. Series B: Biological Sciences, 265(1403), Ecology Progress Series, 512, 155–166. https://doi.org/10.3354/ 1283–1289. https://doi.org/10.1098/rspb.1998.0431 meps10946​ Audicana, M. T., Ansotegui, I. J., de Corres, L. F., & Kennedy, M. W. Clapham, P. J., & Link, J. S. (2006). Whales, whaling, and ecosystems in (2002). Anisakis simplex: Dangerous – Dead and alive? Trends in the North Atlantic Ocean. In J. A. Estes, D. P. Demaster, D. F. Doak, Parasitology, 18, 20–25. https://doi.org/10.1016/S1471​-4922(01) T. M. Williams, & R. L. Brownell Jr. (Eds.), Whales, whaling, and ocean 02152​-3 ecosystems (pp. 314–323). Berkeley, CA: University of California Audicana, M. T., & Kennedy, M. W. (2008). Anisakis simplex: From ob- Press. scure infectious worm to inducer of immune hypersensitivity. Clinical Cuo, L., Lettenmaier, D. P., Alberti, M., & Richey, J. E. (2009). Effects of Microbiology Reviews, 21, 360–379. https://doi.org/10.1128/CMR. a century of land cover and climate change on the hydrology of the 00012​-07 Puget Sound basin. Hydrological Processes, 23, 907–933. https://doi. Ballachey, B., Gorbics, C., & Doroff, A. (2002). Sea otter mortality in Orca org/10.1002/hyp.7228 Inlet, Prince William Sound, Alaska, winter 1995–1996. Alaska Fish Dailey, M., & Stroud, R. (1978). Parasites and associated pathology ob- and Wildlife Service: Marine Mammals Management. served in cetaceans stranded along the Oregon coast. Journal of Wildlife Bao, M., Pierce, G. J., Pascual, S., Gonzalez-Munoz, M., Mattiucci, S., Diseases, 14, 503–511. https://doi.org/10.7589/0090-3558-14.4.503 Mladineo, I., … Strachan, N. J. (2017). Assessing the risk of an emerg- Daschner, A., Alonso-Gomez, A., Cabanas, R., Suarez-de-Parga, J.-M., & ing zoonosis of worldwide concern: Anisakiasis. Scientific Reports, 7, Lopez-Serrano, M.-C. (2000). Gastroallergic anisakiasis: Borderline 43699. https://doi.org/10.1038/srep4​3699 between food allergy and parasitic disease – Clinical and allergologic Bao, M., Pierce, G. J., Strachan, N. J., Martinez, C., Fernandez, R., & evaluation of 20 patients with confirmed acute parasitism by Anisakis Theodossiou, I. (2017). Consumers’ attitudes and willingness to simplex. Journal of Allergy and Clinical Immunology, 105, 176–181. pay for Anisakis-free fish in Spain. Fisheries Research, 202, 149–160. https://doi.org/10.1016/S0091​-6749(00)90194​-5 https://doi.org/10.1016/j.fishres.2017.06.018​ Deardorff, T., Kayes, S., & Fukumura, T. (1991). Human anisakiasis trans- Beck, M., Evans, R., Feist, S., Stebbing, P., Longshaw, M., & Harris, E. mitted by marine food products. Hawaii Medical Journal, 50, 9–16. (2008). Anisakis simplex sensu lato associated with red vent syn- del Pozo, M. D., Moneo, I., de Corres, L. F., Audicana, M. T., Munoz, D., drome in wild adult Atlantic salmon Salmo salar in England and Wales. Fernandez, E., … Garcia, M. (1996). Laboratory determinations in Diseases of Aquatic Organisms, 82, 61–65. https://doi.org/10.3354/ Anisakis simplex allergy. Journal of Allergy and Clinical Immunology, 97, dao01979 977–984. https://doi.org/10.1016/S0091​-6749(96)80073​-X Beman, J., Arrigo, K., & Matson, P. (2005). Agricultural runoff fuels large ESRI. (2011). ArcGIS desktop: Release 10. Redlands, CA: Author. phytoplankton blooms in vulnerable areas of the ocean. Nature, 434, European Commission. (Ed.). (2011). Commission regulation (EU) No 211–214. https://doi.org/10.1038/natur​e03370 1276/2011 of 8 December 2011 amending Annex III to Regulation Bilska-Zajac, E., Rozycki, M., Chmurzynska, E., Karamon, J., Sroka, J., (EC) No 853/2004 of the European Parliament and of the Council as Kochanowski, M., … Cencek, T. (2015). Parasites of Anisakidae family – regards the treatment to kill viable parasites in fishery products for Geographical distribution and threat to human health. Journal human consumption. 1276/2011. of Agricultural Science and Technology A, 5, 146–152. https://doi. FAO. (2008). The state of world fisheries and aquaculture. Retrieved org/10.17265/2161-6256/2015.02A.010​ from http://www.fao.org/3/a-i0250e.pdf Boily, F., & Marcogliese, D. J. (1995). Geographical variations in abundance of FDA. (2009). Food code. College Park, MD: US Food and Drug Admin­ larval anisakine nematodes in Atlantic cod (Gadus morhua) and American istration, Public Health Service, US Department of Health and Human plaice (Hippoglossoides platessoides) from the Gulf of St. Lawrence. Services. Canadian Journal of Fisheries and Aquatic Sciences, 52, 105–115. Fey, S., Siepielski, A., Nussle, S., Cervantes-Yoshida, K., Hwan, J., Huber, Borenstein, M., Hedges, L., Higgins, J., & Rothstein, H. (2009). Introduction E., … Carlson, S. (2015). Recent shifts in the occurrence, cause, to meta-analysis. West Sussex, UK: John Wiley & Sons Ltd. and magnitude of animal mass mortality events. Proceedings of the Bouree, P., Paugam, A., & Petithory, J.-C. (1995). Anisakidosis: Report National Academy of Sciences of the United States of America, 112, of 25 cases and review of the literature. Comparative Immunology, 1083–1088. https://doi.org/10.1073/pnas.14148​94112 Microbiology, and Infectious Diseases, 18, 75–84. https://doi.org/ Froese, R., & Pauly, D. (Eds.). (2000). FishBase 2000: Concepts, design and 10.1016/0147-9571(95)98848-C​ data sources. Los Baños, Laguna, Philippines: ICLARM, 344 pp. Buchmann, K., & Mehrdana, F. (2016). Effects of anisakid nema- Garcia-Palacios, L., Gonzalez, M., Esteban, M., Mirabent, E., Perteguer, todes Anisakis simplex (sl), Pseudoterranova decipiens (sl) and M., & Cuellar, C. (1996). Enzyme-linked immunosorbent assay, immu- Contracaecum osculatum (sl) on fish and consumer health. Food noblot analysis and RAST fluoroimmunoassay analysis of serum re- and Waterborne Parasitology, 4, 13–22. https://doi.org/10.1016/ sponses against crude larval antigens of Anisakis simplex in a Spanish j.fawpar.2016.07.003 random population. Journal of Helminthology, 70, 281–289. Burge, C. A., Mark Eakin, C., Friedman, C. S., Froelich, B., Hershberger, Geraci, J. R., & St. Aubin, D. J. (1987). Effects of parasites on marine mam- P. K., Hofmann, E. E., … Harvell, C. D. (2014). Climate change influ- mals. International Journal for Parasitology, 17, 407–414. https://doi. ences on marine infectious diseases: Implications for management org/10.1016/0020-7519(87)90116-0​ and society. Annual Review of Marine Science, 6, 249–277. https://doi. Gibson, D., Harris, E., Bray, R., Jepson, P., Kuiken, T., Baker, J., & Simpson, org/10.1146/annur​ev-marin​e-01021​3-135029 V. (1998). A survey of the helminth parasites of cetaceans stranded FIORENZA et al. | 2865

on the coast of England and Wales during the period 1990–1994. Lotze, H. K., Lenihan, H. S., Bourque, B. J., Bradbury, R. H., Cooke, R. Journal of the Zoological Society of London, 244, 563–574. https://doi. G., Kay, M. C., … Jackson, J. B. C. (2006). Depletion, degradation, org/10.1111/j.1469-7998.1998.tb00061.x​ and recovery potential of estuaries and coastal seas. Science, 312, Hammond, P. S., Macleod, K., Berggren, P., Borchers, D. L., Burt, L., 1806–1809. https://doi.org/10.1126/scien​ce.1128035 Canadas, A., … Vazquez, J. A. (2013). Cetacean abundance and dis- Magera, A., Flemming, J., Kachner, K., Christensen, L., & Lotze, H. (2013). tribution in European Atlantic shelf waters to inform conservation Recovery trends in marine mammal populations. PLoS ONE, 8, and management. Biological Conservation, 164, 107–122. https://doi. e77908. https://doi.org/10.1371/journ​al.pone.0077908 org/10.1016/j.biocon.2013.04.010 Marcogliese, D. (2001a). Distribution and abundance of sealworm Harmon, A., Littlewood, D., & Wood, C. (2019). Parasites lost: Using nat- (Pseudoterranova decipiens) and other anisakid nematodes in fish and ural history collections to track disease change across deep time. seals in the Gulf of St. Lawrence: Potential importance of climatic Frontiers in Ecology and the Environment, 17, 157–166. https://doi. conditions. NAMMCO Scientific Publications, 3, 113–128. org/10.1002/fee.2017 Marcogliese, D. (2001b). Implications of climate change for parasitism of Harvell, C. D., Mitchell, C. E., Ward, J. R., Altizer, S., Dobson, A. P., Ostfeld, animals in the aquatic environment. Canadian Journal of Zoology, 79, R. S., & Samuel, M. D. (2002). Climate warming and disease risks for 1331–1352. https://doi.org/10.1139/z01-067 terrestrial and marine biota. Science, 296, 2158–2162. https://doi. Marcogliese, D. J., & McClelland, G. (1991). Corynosoma wegeneri org/10.1126/scien​ce.1063699 (Acanthocephala: Polymorphida) and Pseudoterranova decipi- Hays, R., Measures, L. N., & Huot, J. (1998). Euphausiids as intermedi- ens (Nematoda: Ascaridoidea) larvae in Scotian Shelf groundfish. ate hosts of Anisakis simplex in the St. Lawrence Estuary. Canadian Canadian Journal of Fisheries and Aquatic Sciences, 49, 2062–2069. Journal of Zoology, 76, 1226–1235. Mattiucci, S., Cipriani, P., Webb, S. C., Paoletti, M., Marcer, F., Bellisario, Hochberg, N. S., Hamer, D. H., Hughes, J. M., & Wilson, M. E. (2010). B., … Nascetti, G. (2014). Genetic and morphological approaches Anisakidosis: Perils of the deep. Clinical Infectious Diseases, 51, 806– distinguish the three sibling species of the Anisakis simplex species 812. https://doi.org/10.1086/656238 complex, with a species designation as Anisakis berlandi n. sp. for A. Howard, I. A., Davis, E., Lippert, G., Quinn, T. P., & Wood, C. L. (2019). simplex sp. C (Nematoda: Anisakidae). Journal of Parasitology, 100, Abundance of an economically important nematode parasite in- 199–214. creased in Puget Sound between 1930 and 2016: Evidence from Mattiucci, S., Giuletti, L., Paoletti, M., Cipriani, P., Gay, M., Levsen, A., museum specimens confirms historical data. Journal of Applied … Nascetti, G. (2018). Population genetic structure of the parasite Ecology, 56, 190–200. https://doi.org/10.1111/1365-2664.13264 Anisakis simplex (s. s.) collected in Clupea harengus L. from North East Hurst, R. (1984). Identification and description of larval Anisakis sim- Atlantic fishing grounds. Fisheries Research, 202, 103–111. plex and Pseudoterranova decipiens (Anisakidae: Nematoda) from Mattiucci, S., Nascetti, G., Cianchi, R., Paggi, L., Arduino, P., Margolis, L., … New Zealand waters. New Zealand Journal of Marine and Freshwater Bullini, L. (1997). Genetic and ecological data on the Anisakis simplex Research, 18, 177–186. complex, with evidence for a new species (Nematoda, Ascaroidea, Jackson, J. B. C. (2001). What was natural in the coastal oceans? Anisakidae). Journal of Parasitology, 83, 401–416. Proceedings of the National Academy of Sciences of the United States McClelland, G. (2002). The trouble with sealworms (Pseudoterranova of America, 98(10), 5411–5418. https://doi.org/10.1073/pnas.09109​ decipiens species complex, Nematoda): A review. Parasitology, 124, 2898 S183–S203. Jackson, J. B. C., Kirby, M. X., Berger, W. H., Bjorndal, K. A., Botsford, L. McClelland, G., & Martell, D. J. (2001). Surveys of larval sealworm W., Bourque, B. J., … Warner, R. R. (2001). Historical overfishing and (Pseudoterranova decipiens) infection in various fish species sampled from the recent collapse of coastal ecosystems. Science, 293, 629–637. Nova Scotian waters between 1988 and 1996, with an assessment of https://doi.org/10.1126/science.1059199​ examination procedures. NAMMCO Scientific Publications, 3, 57–76. Jensen, T., & Idas, K. (1992). Infection with Pseudoterranova sp. (Krabbe, Measures, L. N. (2014). Anisakiosis and pseudoterranovosis. USGS National 1878) larvae in cod (Gadus morhua) relative to proximity of seal colo- Wildlife Health Center Circular 1393. Reston, VA: US Geological Survey. nies. Sarsia, 76, 227–230. Murphy, S., Pinn, E. H., & Jepson, P. D. (2013). The short-beaked common Karl, H. (2008). Nematode larvae in fish on the German market 20 years dolphin (Delphinus delphis) in the North-east Atlantic: Distribution, of consumer related research. Archiv für lebensmittelhygiene, 59. ecology, management, and conservation status. Oceanography and Khan, M. Q., & Williams, J. (2016). Anisakidosis: A fortuitous mim- Marine Biology, 51, 193–280. icker of gastrointestinal malignancy. BMJ Case Reports, https://doi. Nawa, Y., Hatz, C., & Blum, J. (2005). Sushi delights and parasites: The risk of org/10.1136/bcr-2016-216164 fishborne and foodborne parasitic zoonoses in Asia. Clinical Infectious Kimura, S. (2001). Positive ratio of allergen specific IgE antibodies in Diseases, 41, 1297–1303. https://doi.org/10.1086/496920 serum, from a large scale study. Rinsho Byori, 49, 376–380. Nieuwenhuizen, N., Lopata, A. L., Jeebhay, M. F., Herbert, D. B. R., Krahn, M., Wade, P., Kalinowski, S., Dahleim, M., Taylor, B., Hanson, M., Robins, T. G., & Brombacher, F. (2006). Exposure to the fish para- …Waples, R. (2002). Status review of southern resident killer whales site Anisakis causes allergic airway hyperreactivity and dermatitis. (Orcinus orca) under the Endangered Species Act. US Department Journal of Allergy and Clinical Immunology, 117, 1098–1105. https:// of Commerce, National Oceanic and Atmospheric Administration, doi.org/10.1016/j.jaci.2005.12.1357 National Marine Fisheries Service. Noguera, P., Collins, C., Bruno, D., Pert, C., Turnbull, A., McIntosh, A., … Lafferty, K. (2012). Biodiversity loss decreases parasite diversity: Theory Cook, P. (2009). Red vent syndrome in wild Atlantic salmon Salmo and patterns. Philosophical Transactions of the Royal Society B Biological salar in Scotland is associated with Anisakis simplex sensu stricto Sciences, 367, 2814–2827. https://doi.org/10.1098/rstb.2012.0110 (Nematoda: Anisakidae). Diseases of Aquatic Organisms, 87, 199–215. Levsen, A., & Berland, B. (2012). Anisakis species. In P. Woo & K. https://doi.org/10.3354/dao02141 Buchmann (Eds.), Fish parasites, pathobiology and protection (pp. 298– Paggi, L., Nascetti, G., Cianchi, R., Orecchia, P., Mattiucci, S., D’Amelio, 309). London, UK: CAB International. S., … Bullini, L. (1991). Genetic evidence for three species within Lohi, S., Mustalahti, K., Kaukinen, K., Laurila, K., Collin, P., Rissanen, H., … Pseudoterranova decipiens (Nematoda, Ascaridida, Ascaridoidea) in Mäki, M. (2007). Increasing prevalence of coeliac disease over time. the North Atlantic and Norwegian and Barents Seas. International Alimentary Pharmacology and Therapeutics, 26, 1217–1225. https:// Journal for Parasitology, 21, 195–212. https://doi.org/10.1016/0020- doi.org/10.1111/j.1365-2036.2007.03502.x 7519(91)90010-5​ 2866 | FIORENZA et al.

Persson, L. (1991). Behavioral response to predators reverses the out- Uga, S., Ono, K., Kataoka, N., & Hasan, H. (1996). Seroepidemiology of come of competition between prey species. Behavioral Ecology and five major zoonotic parasite infections in inhabitants of Sidoarjo, East Sociobiology, 28, 101–105. https://doi.org/10.1007/BF001​80987 Java, Indonesia. Southeast Asian Journal of Tropical Medicine and Public Pierce, G. J., Bao, M., MacKenzie, K., Dunser, A., Giuletti, L., Cipriani, P., Health, 27, 556–561. … Hastie, L. C. (2018). Ascaridoid nematode infection in haddock Uriarte, I., & Villate, F. (2004). Effects of pollution on zooplankton (Melanogrammus aeglefinus) and whiting (Merlangius merlangus) in abundance and distribution in two estuaries of the Basque coast Northeast Atlantic waters. Fisheries Research, 202, 122–133. https:// (Bay of Biscay). Marine Pollution Bulletin, 49, 220–228. https://doi. doi.org/10.1016/j.fishres.2017.09.008​ org/10.1016/j.marpolbul.2004.02.010​ Purello-D’Ambrosio, F., Pastorello, E., Gangemi, S., Lombardo, G., Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor Ricciardi, L., Fogliani, O., & Merendino, R. (2000). Incidence of sen- package. Journal of Statistical Software, 36. sitivity to Anisakis simplex in a risk population of fishermen/fish- Ward, J., & Lafferty, K. (2004). The elusive baseline of marine disease: mongers. Annals of Allergy, Asthma, and Immunology, 84, 439–444. Are diseases in ocean ecosystems increasing? PLoS Biology, 2, 542– https://doi.org/10.1016/S1081-1206(10)62278​ -8​ 547. https://doi.org/10.1371/journ​al.pbio.0020120 Rudel, T. (2008). Meta-analyses of case studies: A method for studying re- Wood, C. L., Baum, J., Reddy, S. M. W., Trebilco, R., Sandin, S., Zgliczynski, gional and global environmental change. Global Environmental Change, B., … Micheli, F. (2015). Productivity and fishing pressure drive vari- 18, 18–25. https://doi.org/10.1016/j.gloen​vcha.2007.06.001 ability in fish parasite assemblages of the Line Islands, equatorial Scala, E., Giani, M., Pirrotta, L., Guerra, E., Cadoni, S., Girardelli, C., … Pacific. Ecology, 96, 1383–1398. https://doi.org/10.1890/13-2154.1 Puddu, P. (2001). Occupational generalised urticaria and aller- Wood, C. L., Lafferty, K. D., & Micheli, F. (2010). Fishing out marine parasites? gic airborne asthma due to Anisakis simplex. European Journal of Impacts of fishing on rates of parasitism in the ocean. Ecology Letters, Dermatology, 11, 249–250. 13, 761–775. https://doi.org/10.1111/j.1461-0248.2010.01467.x Shaw, D. J., Grenfell, B. T., & Dobson, A. P. (1998). Patterns of macropar- Wood, C. L., Sandin, S., Zgliczynski, B., Guerra, A. S., & Micheli, F. (2014). asite aggregation in wildlife host populations. Parasitology, 117, 597– Fishing drives declines in fish parasite diversity and has variable ef- 610. https://doi.org/10.1017/S0031​18209​8003448 fects on parasite abundance. Ecology, 95, 1929–1946. https://doi.org/ Shibata, E., Ueda, T., Akaike, G., & Saida, Y. (2014). CT findings of gastric 10.1890/13-1270.1 and intestinal anisakiasis. Abdominal Imaging, 39, 257–261. https:// Woolhouse, M. E. J., & Gowtage-Sequeria, S. (2005). Host range and doi.org/10.1007/s0026​1-014-0075-3 emerging and reemerging pathogens. Emerging Infectious Diseases, Shrestha, S., Kisino, A., Watanabe, M., Itsukaichi, H., Hamasuna, K., Ohno, 11, 1842–1847. https://doi.org/10.3201/eid11​12.050997 G., & Tsugu, A. (2014). Intestinal anisakiasis treated successfully with Young, P., & Lowe, D. (1969). Larval nematodes from fish of the subfam- conservative therapy: Importance of clinical diagnosis. World Journal ily Anisakinae and gastro-intestinal lesions in mammals. Journal of of Gastroenterology, 20, 598–602. https://doi.org/10.3748/wjg.v20. Comparative Pathology, 79, 301–312. https://doi.org/10.1016/0021- i2.598 9975(69)90043​-7 Siokou-Frangou, I., & Papathanassiou, E. (1991). Differentiation of zoo- plankton populations in a polluted area. Marine Ecology Progress Series, 76, 41–51. https://doi.org/10.3354/meps0​76041 Skaug, H. J., Oien, N., & Bothun, G. (2004). Abundance of minke whales SUPPORTING INFORMATION (Balaenoptera acutorostrata) in the Northeast Atlantic: Variability in Additional supporting information may be found online in the time and space. Canadian Journal of Fisheries and Aquatic Sciences, 61, Supporting Information section. 870–886. Smith, J. W., & Wootten, R. (1978). Anisakis and anisakiasis. Advances in Parasitology, 16, 93–163. Fiorenza EA, Wendt CA, Dobkowski Spraker, T., Lyons, E., Tolliver, S., & Bair, H. (2003). Ascaridoid nema- How to cite this article: todes and associated lesions in stomachs of subadult male northern KA, et al. It’s a wormy world: Meta-analysis reveals several fur seals (Callorhinus ursinus) on St. Paul Island, Alaska (1987–1999). decades of change in the global abundance of the parasitic Journal of Veterinary Diagnostic Investigations, 14, 432–437. nematodes Anisakis spp. and Pseudoterranova spp. in marine Sprengal, G., & Luchtenberg, H. (1991). Infection by endoparasites re- fishes and invertebrates. Glob Change Biol. 2020;26:2854– duces maximum swimming speed of European smelt Osmerus eperla- nus and European eel Anguilla anguilla. Diseases of Aquatic Organisms, 2866. https://doi.org/10.1111/gcb.15048 11, 31–35. https://doi.org/10.3354/dao01​1031