Planktonic associations between medusae (classes Scyphozoa and Hydrozoa) and epifaunal crustaceans

Jellyfish are known to carry various epibionts, including many of the subphylum Crustacea. However, the associations between gelatinous zooplankton and other invertebrates have been chronically overlooked. Crustacea, a massive clade of economically, ecologically, and culturally important species, includes many taxa that utilize gelatinous zooplankton for food, transport, and protection as both adults and juveniles. Here we compile 211 instances of epifaunal crustaceans recorded on Hydromedusae and Scyphomedusae from a century of literature. These include 78 identified crustacean species in 65 genera across nine orders found upon 37 Hydromedusa species and 48 Scyphomedusae. The crustacean life stage, location, nature of the association with the medusa, years, months, and depths are compiled to form a comprehensive view of the current state of the literature. Additionally, this review highlights areas where the current literature is lacking, particularly noting our poor understanding of the relationships between juvenile crabs of commercially valuable species and medusae.

While research has expanded around services jellyfish provide (Riascos et al., 2018), much of this research focuses on benefit and harm to vertebrates (Brodeur, 1998;Cardona et al., 2012;Mir-Arguimbau, Sabatés & Tilves, 2019). However, the relationships between scyphomedusae, hydromedusae and other invertebrates are currently poorly characterized. A prime invertebrate group to analyze through this lens is Crustacea. Crustaceans are some of the most visible and well-studied marine invertebrates. They are present in every region and are integral components of food webs, including species of high commercial value and known ecological significance (Boudreau & Worm, 2012). Ecological processes that impact them are thus relevant to humans. However, studies focusing on epifaunal crustaceans and jellyfish interactions have been scarce, incomplete, and taxonomically imprecise. Moreover, such studies are often narrowly focused accounts of interactions with single individuals (Weymouth, 1910;Reddiah, 1968;Yusa et al., 2015). Some early communications discuss these interactions as common knowledge that has, however, failed to be recorded in the scientific literature (Jachowski, 1963). This review provides a list of documented crustacean epibionts on medusae of the orders Scyphozoa and Hydrozoa. This work aims to assess the breadth and depth of jellyfish-crustacean interaction and develop a resource for further studies.

METHODOLOGY
Four independent sets of searches were conducted in Google Scholar using keywords, as described in Fig. 1. All four searches were conducted in early November 2019 and were revisited in January 2021 to include all results through the end of 2019. Searches were performed in English, and as such, only papers published in or with an available translation to English were included. The number of papers yielded by each of the four searches is shown in Fig. 1, ranges from 4,840 articles (for keywords Crustacea, Scyphozoa) to 13,300 (for keywords Crustacea, Jellyfish) (See Fig. 1 for details). Only papers in which the primary focus was associations between medusae (Hydrozoa and Scyphozoa) and crustaceans were further selected.
The four searches performed returned many invariable results. All titles and abstracts were checked for relevance. Results from 161 papers were obtained initially and then narrowed to 81, after excluding repeat papers mistakenly included multiple times and papers on cubomedusae, ctenophores, ascidians, and non-crustacean epibionts. Also, results from six relevant literature reviews were included (Vader, 1972;Pagès, 2000;Towanda & Thuesen, 2006;Ohtsuka et al., 2011;Schiariti et al., 2012;Wakabayashi, Tanaka & Phillips, 2019). These reviews account for 40 interactions from 29 sources ( Table 1). The inclusion of the literature reviews was deemed essential to include results from earlier sources and non-English sources not available on Google Scholar. Results from literature reviews that had no information on the nature of the interaction between the medusa and crustaceans (such as taxa identification, location, etc.) were eliminated. Records were also analyzed for taxon validity using the World Register of Marine Species (WoRMS). Seven papers within the database that referred to invalid taxa with no valid synonymized name in WoRMS were removed. Results from 97 unique sources (68 articles from the Google Scholar search and 29 from literature reviews) were kept. From these 97 sources, 211 distinct interactions were extracted. Details provided by each paper were recorded in Table 1.

RESULTS AND DISCUSSION
The final table produced by this review process includes 211 recorded interactions between hydrozoan or scyphozoan medusae and crustaceans, extracted from 97 papers (Table 1). For both cnidarians and crustaceans, order, family, genus, and species are included in Supplementary Materials. Results that lacked taxonomic identification (at least Family level) were not included. The final table (Table 1) provides sampling information, such as year and month of sampling, sampling method, and region of sampling. For crustaceans, records include the life stage involved in the interaction, sex of the epibiont, location on the hosts, and additional notes, if available. In most studies, fewer data were available on the cnidarian hosts, reducing the degree to which these interactions could be analyzed in terms of hydromedusan or scyphomedusan life stage. In the next paragraphs, we discuss the jellyfish-crustacea interactions through all of the categories included.

Diversity of scyphozoan hosts
A supermajority of records (70%, or 148/211) involves Scyphomedusae, with 53 records involving just the five most common scyphozoan species: Lychnorhiza lucerna (Haeckel, 1880), Catostylus mosaicus (Quoy & Gaimard, 1824), Stomolophus meleagris Table 1 Associations reported organized by host. Every association in all reviewed papers with details on species and higher order classification of host, species of associate, sex and life stage of associate, notes on association, location on host, location association was recorded, date of record, depth of association and literature source.     1984-1985? Dittrich, 1988 (Continued )         1888). These records are heavily concentrated in the upper water column. Deeper water collections (ROV/HOV) were dominated by hydromedusae (69%, or 27/39), while records involving the upper water column (0-30 m) were more common and dominated by scyphomedusae (78%, or 83/106). Sixty-seven records included no specific sampling depth. These records were generally more than 50 years old. Although they are likely near-surface sampling records and mainly report known shallow-water species, they cannot be verified as such because of the lack of explicit information. Most of these (87%, or 58/67) are records of scyphomedusae. Overall, the diversity of scyphomedusae was low, with only 39 species from 27 genera represented in records ( Fig. 2A). The genus Chrysaora had the largest contingent of accounts, with 21 individual records of associations across at least seven Chrysaora species. This genus has been reported to interact with 16 different epifaunal crustaceans. The genera Chrysaora, Lychnorhiza, and Catostylus accounted for a third of scyphozoan records. These records originate mainly from the upper water levels of various locations (i.e., the east coast of the United States, the southeast of Brazil, the southern Australian coast, and the western Philippines, Japan and Pakistan).

Diversity of hydrozoan hosts
Twenty-six genera, and six Hydrozoan orders were reported interacting with Crustacea in 63 records (Fig. 2B). The order Leptothecata included the greatest number of records (18), with 17 records of Siphonophorae and 12 of Narcomedusae. The diversity of Hydrozoa was significantly limited by region, with 45 of the 63 records (71%) from the Gulf of California. Additionally, those from the Gulf were acquired from primarily deep water ROV missions. The medusae recorded belonged to 28 known species, with twelve records unable to provide higher resolution than genus and a single Prayid siphonophore only identified to the family level. Rosecea cymbiformis (Delle Chiaje, 1830) (4), Aegina citrea (Eschscholtz, 1829) (5), and Aequorea coerulescens (Brandt, 1835) (6) were the three most common species.

Diversity of crustacean epibionts
The crustaceans included Hexanauplia (reported in 37 discrete observations), Malacostraca (173), and a single representative of Branchiopoda (Evadne sp.) (Fig. 3). Recorded Hexanauplia consisted of mainly specialist groups known to be obligate epibionts and had overall low species resolution, with 13 of the 23 documented associations lacking a species name. The Macrochironidae, a group of known scyphozoan parasites, makes up 12 of the copepod epibiont records. Outside of this family, no additional Hexanauplia epibiont was recorded more than twice. The single reported case of a medusa with Evadne sp. occurred in a broad analysis of items found on a Catostylus medusae (Browne & Kingsford, 2005). As this was not replicated throughout medusae within the study, or in other studies, it is unlikely this is a common or genuine association. The bulk of the associations involve crustaceans of the class Malacostraca. These 173 records include amphipods and decapods in equal proportion (47%, or 81/173 each), isopods (5%, or 9/173), and mysids (1%, or 2/173). The amphipods are dominated by the parasitic family Hyperidae, recorded in 32 separate encounters. Members of the family of Hyperidae are present across 22 identified scyphozoan and hydrozoan species, making them the most widely distributed family. Hyperia galba (Montagu, 1813) is present in nine records from both surface and deep-water samples, making it the single most plentiful within the amphipods. Outside of the family Hyperidae, Tryphana malmii (Boeck, 1871) is recorded six times in association with deep-sea jellyfish. Most amphipod species recorded were recorded on multiple host species.
Three species of cirripeds were recorded 15 times in association with jellyfish, Alepas pacifica (Pilsbry, 1907) accounting for twelve of such records, Conchoderma virgatum (Spengler, 1789) accounting for two, and a single report of an unidentified Anelasma epibiont on a Pelagia noctiluca (Forsskål, 1775) from 1902. Alepas pacifica has been found on seven separate host species, all scyphozoans. The vast majority of these records came from a single literature review included within an extensive paper from Vader (1972). None of these species were found in deep-sea records.

Field collections
Only 58 papers included some explicit method of capture of the jellyfish and its epibiont (Fig. 4). Between 1862 and 1962, only seven of the twenty records reported a method of capture. From 1963 to 1989, this increased to 64%, with 25 of 39 records including the collection method. Since 1990, there have been only seven failures to report collection methods out of 140 accounts. The most common method of collection, used in 31 of the papers, is "by hand", defined as using handheld dip nets, buckets, plastic bags, and, in limited cases, collection of carcasses from beaches. Trawling was first used in 1968 and has remained in use until recently, reported in 17 of the 33 associations after 2010. Although 38 records were obtained through deep water methods (HOV and ROV), these were used scarcely before 1999. Some studies employed multiple methods, with divers and ROV, or dip net and trawl capture, such that it was unclear which associations were found by each collection method. These were listed as "multi-method" and include four papers.
The larger proportion of scyphozoan hosts to hydrozoan hosts may be a sampling artifact. The vast majority of the papers discussed here were only analyzing interactions in the top 30 m of the water column. A fair number, especially earlier texts, involve serendipitous encounters at the water's edge or within sight of the surface (Bowman, Meyers & Hicks, 1963;Jachowski, 1963;Vader, 1972;Martin & Kuck, 1991). The larger, more visible nature of surface water scyphozoans of the rhizostomes and semaeostomes makes them an easier collection target than deep water species. Note that only a single scyphozoan of the order Coronatae, which has no large shallow representatives, was recorded as well. Many elements of the sampling methods impact the scope of this data, and the preeminence of hand collection and papers written on chance occurrences, as opposed to prolonged study, result in a picture that heavily weights organisms more frequently seen or interacted with by humans.
The oldest records of jellyfish-crustacean interaction involved hand collection with buckets and nets, often from shore. These include first accounts of hyperiid amphipod-jellyfish associations from the Chesapeake Bay (Bowman, Meyers & Hicks, 1963). Buckets and nets have remained mainstays, with hand collection accounting for 34 of the 108 post-2000 records and 32 of the 55 pre-2000 records. Buckets and plastic bags are likely preferable to nets, as they may reduce chances of epibiont detachment and medusa damage.
Trawling (by ring nets, otter nets, and bottom trawls), while reported in twelve papers, has been a prominent capture method in South America for the last two decades. However, trawling provides an additional threat, as epibionts may detach, get caught in the bell of a medusa, or move to a different location within the carcass. Given the damage sustained by gelatinous bodies during trawls, and the inability to capture more delicate associations, this is the methodology that seems most likely to provide low-quality relationship information. A focus on a lower number of medusae examined in more detail, may provide more useful information on the ecology of the interaction between jellyfish and their epibionts. Notably, Greer et al. (2017) uses a combination of in situ imaging (with an automatic ISIIS imaging system) and trawls. Trawls were used to verify the identity of organisms seen in the captured images. Such a protocol should be considered for future quantitative and qualitative work.
A total of 66% of the records (136/211) are from known surface encounters. 18% of the records (38/211) involve deep water accounts using either an ROV/HOV. These records are distributed unevenly across depths with few records below the mesopelagic zone (Fig. 5). Most of these records fail to provide epibiont location on the jellyfish but provide Figure 4 Collections information for both number of papers using a collection method and number of associations reported from this collection type. Types are blue water diving (BWD), collection by hand (HC), multiple methods (MULTI), ring net (RN), scuba diving (SC), trawling (TR), in situ observation (OBS) or unknown (Unknown). Associations from papers in which multiple methods were used, but specific methods are known for each association are categorized under the known method. Many papers are comprised of multiple associations, as such, the "Individual" columns include each association separately, "Paper" columns report by paper. Full-size  DOI: 10.7717/peerj.11281/ fig-4 the only available information on deep water scyphomedusa and hydromedusa hosts. Most of the deep water records are from the Gulf of California. While this sampling method is useful, the high cost and difficulty of use of ROV and HOV equipment make it unrealistic for the vast majority of researchers. The limited number of deep-water accounts and the novelty of many of the findings on each dive can be attributed mainly to these limitations (Gasca & Haddock, 2004;Gasca, Suárez-Morales & Haddock, 2007;Gasca, Hoover & Haddock, 2015).
Given the fragility of scyphozoan and hydrozoan medusae, as well as the delicacy of the interaction with their epibionts, the most precise picture of the jellyfish-crustacean associations has been achieved from dip net, plastic bag, bucket, or other by-hand collection methods. These are not only a cost-effective strategy requiring little additional equipment, they also maintain maximum integrity of the organisms. Hand collection, however, is restricted to analyzing associations that are close to the surface. Trawl sampling provides a reliable way to collect many medusae offshore but sacrifices sample integrity. ROV is an imperfect sampling method, often failing to record epibiont positioning, but allows for the only viewing, documentation, and collection of deep water associations, thereby being uniquely important, especially for hydromedusa research. Moreover, the majority of the records document all symbionts on the target host species, often with little data beyond a name or tentative classification for the epibiont. This lack of closer examination leads to an inability to correctly categorize the nature of the relationship, including positioning, feeding behaviors, and duration of the interaction.
In conclusion, the overall best sampling results come from observation-first methodologies such as collection by-hand while snorkeling and diving, as in Mazda et al. (2019), ROV/HOV in situ underwater photography, as employed by Gasca, Hoover & Haddock (2015), or imaging and supplemental trawling as in Greer et al. (2017). Obtaining underwater pictures of medusae and epibiont is crucial to the understanding of the associate placement in relation to host and its behavior. It is also more informative than post hoc in-lab examinations and analysis of trawl contents, because the stress of collection and sampling may impact the epibiont position within the host (Hayashi, Sakagami & Toyoda, 2004). As waterproof video equipment becomes less expensive, options like a simple GoPro may provide clear enough imaging to allow novel in situ observations. Adding an underwater imaging component to sampling may also enable collectors to revisit the ecological context of the association.

Life stages
Age classes and sex, where available, are reported in Table 1. 63% of all records (133/211) reported an age class for the crustacean. 65% of the interactions with a listed age class (65%, or 86/133) reported crustacean juveniles, eggs, larval stages, copepodites, megalopae, or other immature forms. For a minority of records (37%, or 73/211), no information on the crustaceans' age class and sex was available. When individuals were described as "male" or "female" without any qualifier attached, they were catalogued and treated as adult specimens (Table 1). Megalopae were noted only nine times out of the 106 records that reported an age class for the crustacean associate (8%). In these nine records, the megalopae belonged to the genera Callinectes, Periclimenes, Metacarcinus, Cancer, and Charybdis, and were all in association with Scyphomedusae (Orders: Rhizostomeae and Semaeostomeae). In addition to megalopae, phyllosoma larvae of the families Scyllaridae and Scyllarinae were reported 12 times. The occurrence of larvae of this type associated with medusae and, more generally, with gelatinous zooplankton is well known, especially along the Japanese coast (Wakabayashi, Tanaka & Phillips, 2019). Within and upon the host, juvenile crustaceans were often coexisting with adult forms. Eighty-one of the associations include juveniles (excluding megalopae, eggs, and copepodites), sometimes embedded in host tissue (Towanda & Thuesen, 2006;Browne, 2015;Yusa et al., 2015;Browne, Pitt & Norman, 2017;Mazda et al., 2019). The presence of eggs and ovigerous females was reported in 39 cases from 23 different species. In at least three papers, females and ovigerous females were present in exceptionally high proportions relative to adult males (Filho et al., 2008;Oliva, Maffet & Laudien, 2010;Mazda et al., 2019). Records of megalopae of the commercial crab, Charybdis feriata were reported in substantial numbers on two separate hosts (Kondo et al., 2014;Boco & Metillo, 2018). In other reports, associations between juvenile Metacarcinus gracilis (Dana, 1852) and medusae are hypothesized to be beneficial to the crab as the medusae supply means of transport and food acquisition, which may be similar across juvenile decapod-scyphozoan associations (Towanda & Thuesen, 2006).

Nature of associations between medusae and crustaceans
There is no agreement between authors on the degree to which medusae and crustaceans' interactions are parasitic, commensal, or otherwise. In the case of the scyphozoan Phacellophora camtschatica and the decapod Metacarcinus gracilis (Dana, 1852), the interaction may involve a mutualistic cleaning relationship as M. gracilis graduates into adulthood (Towanda & Thuesen, 2006). Other reports of megolopae do not suggest any parasitization of the medusae. Weymouth (1910) also indicates that this is a commensal relationship important to M. gracilis megalopae until they reach~20mm.
In other cases, such as the shrimp Perimincles paivai, the commensals seemed to be feeding on the mucus, not the host tissue (Browne & Kingsford, 2005;Filho et al., 2008). Dittrich (1988) demonstrates an aggressive parasitoidism by Hyperia galba in which a large subset of host medusae was so reduced by predation as to lose almost all morphological features. While the ultimate death of these hosts is not recorded within the text, the loss of all tentacular structure and non-mesoglear tissue would make survival nearly impossible. The numbers in which Hyperia can be found on some of the recorded medusae, occasionally upwards of 100 amphipods engaging in host consumption, may lend credence to the parasitoid rather than classically parasitic nature of this relationship in many hosts (Vader, 1972;Dittrich, 1988;Towanda & Thuesen, 2006). However, additional reports on the same species and other hyperiids reported that this group engages in cradle positioning, facing outwards from the medusa, into the water column with no reported predation, or engage in only limited predation of the gonadal tissue or mesogleal tissue (Bowman, Meyers & Hicks, 1963;Browne, 2015). Based on this information it seems likely that the family Hyperidae includes a variety of strategies, and the family Hyperia itself may also encompass non-aggressive parasitism, aggressive parasitism, and parasitoidism. In part, this may be due to temporal behavioral differences within species, with more extreme predation in summer and autumn and limited parasitism in spring as populations raise and fall (Bowman, Meyers & Hicks, 1963;Dittrich, 1988). "Inverted cradle" positioning is a recurring feature of amphipod associates (Bowman, Meyers & Hicks, 1963;Condon & Norman, 1999). While some of the crustaceans fed on the medusae themselves, Towanda & Thuesen (2006) primarily recorded crustaceans engaging in theft of prey collected by medusae. Many crustaceans that were reported feeding on the medusae were feeding entirely or in part on the highly regenerative gonadal tissue (Pagès, 2000;Towanda & Thuesen, 2006;Ohtsuka et al., 2009) or engaging in the excavation of small pits in the host mesoglea (Humes, 1953;Jachowski, 1963;Browne, 2015). Reports of Libinia dubia (H. Milne Edwards, 1834) have the greatest agreement on the parasitic nature of the species' interactions with their medusa host (Jachowski, 1963;Phillips, Burke & Keener, 1969;Schiariti et al., 2012).
The largest exception to the above patterns of limited consumption or longer term residence is the scholarship surrounding phyllosoma larvae on gelatinous zooplankton. These larvae have been reported to stab a pair of pereiopods through the exumbrella or exterior of a nectophore and use the medusa as propulsion and food source. This is a common occurrence both in the northern Gulf of Mexico and at various locations along the Japanese coast (Greer et al., 2017;Wakabayashi, Tanaka & Phillips, 2019). In the review on the subject by Wakabayashi, Tanaka & Phillips (2019), it is hypothesized that the flattened body and ventral mouth of these phyllosoma larvae is ideal for consumption of gelatinous zooplankton while attached. The exact length of this parasitoid association is unknown, though it is likely generally ended by the medusa's eventual death as the larva eats its way through.
The degree to which crustaceans engage in host consumption may be in part obscured by the speed with which medusae regenerate tissues, especially gonadal and oral arm tissues (Towanda & Thuesen, 2006). The number of associates (at least eight crustacean species) found residing within the bell and around the gonads, suggests that gonadal tissue may be common nourishment even when bell and arm tissue is not consumed. Overall, the relationships of crustaceans with their medusa hosts remain largely uncharacterized and require additional study. Few papers have analyzed the gut contents of the epibionts, which would be a helpful tool in determining whether inverted positioning on hosts was actually a signal of lack of consumption, or simply a break from such (Vader, 1972;Pagès, 2000;Towanda & Thuesen, 2006;Oliva, Maffet & Laudien, 2010). Detailed records of the diets of such organisms are difficult to reconstruct. However, specific searches for nematocysts in digestive tract and excretions or stable isotope analysis have proven successful at identifying cnidomedusae as possible food sources (Schiariti et al., 2012;Fleming et al., 2014). Expanding future works to include both these practices, photographs of the host medusae, and notes on swimming strength, tentacular loss and other signs of deterioration would improve our understanding of how detrimental these relationships actually are. This sort of documentation of host condition is impossible when specimens are collected via trawl.
In addition to consumption, the issue of host choice and host specificity has been analyzed only sparsely. There is evidence in multiple studies that while some individual jellyfish host symbionts, others in the same area lack them due to their size or species (Towanda & Thuesen, 2006;Ohtsuka et al., 2011;Boco & Metillo, 2018). While exotic species often have lower amounts of parasitization in their introduced range (Torchin et al., 2003), the degree to which epibionts in medusae are affected by host or epibiont endemicity is unknown. The high number of cryptic species, a history of misidentification, and poor understandings of historical ranges compound issues with sparse research on the topic (Dawson, 2005;Graham & Bayha, 2007;Morandini et al., 2017;De Souza & Dawson, 2018).
Only one study provides an indication of how nuanced the relationship between gelatinous zooplankton hosts and epibionts may be; 6 years of monthly observation showed that single adult females of the amphipod Oxycephallus clausi (Bovallius, 1887) had a broad range of gelatinous hosts, but shifted to primarily Ocyropsis fusca (Rang, 1827), a lobate ctenophore, during brood release (Mazda et al., 2019). While ctenophores are not the focus of this review, it shows that the nature of interactions may change during the crustacean lifecycle. These sorts of long-term analyses are hard to pursue, but provide a fascinating look at the range of information that can be collected with observational methods. Uneven sex ratios, such as those seen in the case of Oxycephallusclausi (97% female), are present across many associations (Condon & Norman, 1999;Filho et al., 2008;Oliva, Maffet & Laudien, 2010;Mazda et al., 2019). The most common explanation for this higher ratio of females and often ovigerous females is use of scyphozoan and hydrozoan hosts primarily as nursery habitat for movement and protection of juveniles (Gonçalves et al., 2016;Gonçalves et al., 2017;Mazda et al., 2019). Potential territoriality in some females, like those of P. paivai, may help ensure more resources for their brood, and is in line with other symbiont crustaceans (Baeza et al., 2017). For deep sea crustaceans, such as Pseudolubbockia dilatata (Sars, 1909), more even sex ratios would be expected, as there is evidence of long-term resident brooding pairs, and mate scarcity is a feature of deep sea life. Evidence for long-term association and pairing has not been found for other deep water crustaceans, although understanding these deep sea interactions is generally hampered but small sample sizes and difficulty of observation (Gasca, Suárez-Morales & Haddock, 2007;Baeza et al., 2017;Gasca & Browne, 2018).

Years and locations
The oldest records examined were only available from earlier literature reviews (Pagès, 2000;Towanda & Thuesen, 2006;Schiariti et al., 2012). The first record is the Bate (1862) account of the amphipod Iphimedia eblanae on the scyphozoan Rhizostoma pulmo (Macri, 1778) from 1862, also reported in the Vader (1972) review on amphipod associations with medusae. Thiel (1976) refers to older records from as far back as 1791.
Overall, the number of records detailing interactions has risen over time but has not exceeded ten papers during any 5 years. While these numbers are increasing modestly, the number of distinct interactions that any given paper reports have increased. Pre-1990s articles, on average put forward information on 1.24 associations per paper. In contrast, the average number of associations reported in papers published from 1990 to 2018 increased more than twofold (an average of 2.83 records per paper). These surveys provide useful records of separate associations found in one area or on one organism and are informative of ecosystem features on a regional level. Still, given the studies' breadth, they often lack depth, not characterizing relationships between individual host species and their associates.
Records were unevenly distributed globally, with Africa and Europe completely devoid of records from the past 30 years with the exception of a single note on an accidental observation from Gran Canaria, Spain. The eastern coast of North America (one record since 1984 (Tunberg & Reed, 2004) and China (no direct records)), as well as West Africa (one record from 1972 (Bruce, 1972)) and the Mediterranean Sea (last collections 1985 (Dittrich, 1988)) also lack records from the last 30 years. The areas consistently covered by recent papers are Australia ( ), the Philippines (2014, the eastern coast of South America , and the western United States . Japanese records represent the longest continuity over time, with 33 records between 1902 and 2019. The association that consistently appears throughout time is that of Alepas pacifica (Thoracica, Lepadiformes) with Nomura's Jellyfish (Nemopilema nomurai) (Pagès, 2000;Yusa et al., 2015). The first record of this association was in 1902 (Pagès, 2000), and the most recent in 2015 (Yusa et al., 2015). Phyllosoma larvae of multiple species, Chlorotocella gracilis (Balss, 1914), and Latreutes spp. also have records spanning multiple decades and papers.
It is worth mentioning that the uneven geographic distribution of associations reported herein may be an artifact of lack of readily available English translations of works from some areas. Reports from Japan and China of crustacean and gelatinous zooplankton associations are mentioned by Hayashi, Sakagami & Toyoda (2004) and Wakabayashi, Tanaka & Phillips (2019), but were not available in English and therefore are not accounted for in this review. Similarly, European records may be underestimated, as non-English records are absent. Other locations' lack of records may be a more accurate representation of a gap in academic knowledge. Africa's west and eastern coasts are known to be understudied ecosystems, and so the missing research here is likely not just untranslated (Berkström et al., 2019). As in other ecological inquiries, the expansion of Local Ecological Knowledge into the study of gelatinous zooplankton should be considered, as fishermen and coastal communities often have a deep knowledge of organisms and their associations (Berkström et al., 2019). Fishermen are often well acquainted with specific gelatinous zooplankton species and know their harms, and may have knowledge of symbionts living upon or within them (Al-Rubiay et al., 2009).
Slipper lobster larvae of the genera Scyllarus and Ibacus have been reported many times across various hosts (Wakabayashi, Tanaka & Phillips, 2019). Some slipper lobsters are commercially fished for consumption, and a large number of these larvae (40% in the Gulf of Mexico) have been shown to live attached to gelatinous zooplankton (Greer et al., 2017).
The consumption of some Scyphozoan hosts, such as Catostylus mosaicus and Rhopilema spp., makes their records valuable as well. The fishing pressures on the jellyfish populations may significantly impact the crustaceans that rely on their oral arms and bells for transport and nourishment of their juvenile stages. Further understanding of these relationships may be especially important in cases where both the medusae (e.g., Rhopilema spp., Lobonemoides robustus (Stiasny, 1920) and Catostylus spp.) and crustacean (Charybdis feriata) are subject to fishing (Boco, Metillo & Papa, 2014;Boco & Metillo, 2018, Kondo et al., 2014. Finally, current information on Callinectes sapidus and its relationship to and frequency of interaction with host jellyfish is needed, as the blue crab represents a commercially valuable fishery in the Gulf of Mexico and along the Atlantic Coast of the USA. Understanding the nature of the relationships between economically valuable species of Crustacea and common scyphozoans and hydrozoans can improve fisheries practices and regulation, as already acknowledged for economically important fish and their jellyfish hosts (Tilves et al., 2018). The importance of maintaining juvenile communities for commercially sized adult populations to recruit from is well established and a frequent impetus for marine protection areas. The fishing of medusae is different from most modern vertebrate fishing. It is temporally highly variable, and blooms, when found, are fished as intensely as possible by local fishermen. It is also comparatively new as an export industry, especially in Southeast Asia (Omori & Nakano, 2001). Additional regulation and management should be considered for jellyfish species known to harbor juveniles of commercially viable crustaceans. It is clear that many crustaceans, fish, and other organisms live in, upon and around medusae, thus indiscriminate efforts to remove or destroy blooms of endemic species are likely unwise (Tilves et al., 2018;Riascos et al., 2018).

CONCLUSION
Many of the interactions we reviewed are fragmented and not comprehensive. Studies covering timing and breadth of infection of commercially valuable crustaceans on marine scyphozoans are scarce, but may be valuable information to fully understand the complexity of their life cycle, and thus the species' vulnerability at each life cycle stage. The general picture of the commensal relationships that arise from this review is complex and emphasizes the diversity of jellyfish and crustaceans' relationships. Any attempt to paint them as uniformly parasitic fails to acknowledge the diversity of crustacean host-use strategies. While some seem to be parasitic or parasitoid, others are life-stage dependent commensals reliant on medusae for transportation. Some deep water crustaceans may be lifelong commensals (Gasca, Suárez-Morales & Haddock, 2007). In each of these cases, the work thus far is far from exhaustive. Additional research on seasonality, maternal care, territoriality, impact on host and other such matters should be further pursued.
The scyphozoans and hydrozoans studied here represent only a small proportion of the globally recognized species. Even shallow water coastal species are poorly covered. This research has been restricted to a small selection of near-shore sites over the past 50 years, leaving inadequate coverage even in regions with a significant scyphozoan research presence (i.e., the Mediterranean, western Europe, China, northeastern North America). Because much of the published research focused on single occurrences, this paper's overall results do not necessarily capture the broader ecology of the species involved (Bowman, Meyers & Hicks, 1963;Jachowski, 1963;Suzuki, 1965;Ohtsuka et al., 2011). Similarly, species descriptions that mention an association without details on the conditions in which it was found offer little insight on the frequency and ecological role of such interactions (Humes, 1953;Reddiah, 1968;Bruce, 1972;Criales, 1984;Bruce, 1988;Bruce, 1995;Bruce, 2008).
Best practices moving forward should include some of the following elements: in situ imaging pre-collection, observations on medusa health, analysis of epibiont gut contents when possible, preferential use of non-destructive collection methods, observations on symbiont placement within or upon the medusa, and frequency, geographical and temporal variation of the association.
With this review, we hope to highlight a significant knowledge gap and a lack of formal study on the ecology of the crustaceans residing on and around jellyfish, as well as a glimpse of the ecological complexity of these interactions. We provide easy access to a century of ecological research and a framework for analyzing and contextualizing future research on this topic.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
The authors received no funding for this work.