Progress in the discovery of isopods (Crustacea: Peracarida)—is the description rate slowing down?

Species diversity

Between 1758 and 2023 a total of 10,687 extant isopod species have been described by a cohort of 1,144 authors (755 first authors). Of the first authors analysed here, 282 were one-time authors who described only a single isopod species. The 21 most prolific authors, each describing more than a hundred species, together described about 43% of all accepted species (see Table S1). More than half of all named species are marine species—6,151 in number. Isopods are the most species-rich crustaceans on land, with 3,840 terrestrial isopod species and 696 freshwater species. Approximately 14% of all species are obligate parasites, and 9% can be categorised as subterranean (i.e., cave-dwellers, groundwater species, inhabitants of interstitial spaces). The order Isopoda consists of 12 suborders comprising 141 families and 1,557 genera. The most species-rich genera, each containing over 100 species, are Porcellio, Armadillidium, Cirolana, Gnathia, Venezillo, Proasellus and Trichoniscus. The most species-rich isopod families are Sphaeromatidae, Armadillidae and Bopyridae (Table 1). At the other end of species richness, there are 15 monotypic families, which have only one genus containing a single species.

The first 100 years of discovery after the publication of Linnaeus’ Systema Naturae in 1758, in which the first seven still valid isopod species were described, yielded relatively few species. Until the end of the 18th century an average of only 6 species were described per decade. The following 50 years saw, on average, 43 species descriptions per decade, many of which were contributed by the three most prolific taxonomists of that time. Leach described 30 species between 1814 and 1818; J.F. Brandt contributed 37 species descriptions between 1831 and 1841; and H. Milne-Edwards added 34 new species in 1840, at which point the overall number of named isopod species had climbed to 194. For a detailed history of the discovery of marine isopods see Poore & Bruce (2012). From the 1850s to the end of the 19th century the average number of new species per decade climbed to 209. Descriptions of new isopod species started to accumulate faster, and after the 1880s the rate increased swiftly and steadily up to the 1970s, when the slope of the curve got even steeper (Fig. 1A). The terrestrial subgroup follows this overall pattern very closely (Fig. 1C), whereas for marine isopods the cumulative number of species seemed to plateau for short periods of time in the 1890s and the mid-20th century, before resuming a steep increase after the 1960s (Fig. 1B). A dip in descriptions during World War II and its aftermath is clearly visible in almost all groups (Figs. 2A–2D). Only freshwater isopods show a small peak in species descriptions during that time, largely due to Nicholls’ work, who published 36 descriptions of freshwater isopods in 1943 and 1944 (Fig. 2B). Besides having far lower species numbers than marine isopods, discoveries of freshwater species stayed low until the 1880s (Fig. 1B). The discovery of subterranean species started later, and most were discovered after the 1950s (Fig. 1D).

Isopods showed a peak of discovery in the late 20th century, with an all-time high of 200 species described in the year 1982 (Fig. 2A). Most subgroups peaked during the same period, except for freshwater isopods, which had their highest peak at the beginning of the 21st century and terrestrial species having their main peak earlier in the 1930s (Figs. 2B–2D). In the past three decades the number of species described per year has decreased notably in overall species descriptions and specifically marine isopods. Yearly descriptions of freshwater isopods are generally low, although 2020 was a record year that saw 34 freshwater species described. This was more than 10-times the average of the previous 10 years. On average one third of yearly descriptions over the past 10 years were parasitic and subterranean species.

Predictions of yet to be named species

The NHRP model predicted another 470 isopod species to be described by the year 2050 with a 95% confidence interval of 390 to 560 (Fig. 3A). Until 2100 a total of 660 (540–810) species were predicted to await scientific description, assuming the pace of description continues at its current rate. This would bring the cumulative number of isopod species up to 11,347 in 2100 (Fig. 3A). When split into subgroups, estimates from the model show that most of the future discoveries could be expected in marine and terrestrial environments, and only a small part will be from freshwaters (Fig. 3B).

Taxonomic effort

Since the first scientific description of an isopod species by Linnaeus, 755 first authors have described the species known today. Over time the number of first authors per year has increased. Since the 1950s there were more than three times as many authors involved in isopod taxonomy as during the first half of the 20th century (Fig. 4). This pattern can be seen in almost all subgroups (Fig. S1). However, the average number of species described per author has been declining over the last century (Fig. 4). Nevertheless, the overall trend sees many more taxonomists describing fewer species. A piecewise regression analysis found the breakpoint in the data series to be in 1916, whether zero values were excluded or not (Fig. 5). Since then, the average number of species described per authors active in the same year has declined.

The average publication lifetime of an author was found to be 8.4 years, with 30% of authors ranking above the average. Although a linear regression shows a weak decreasing trend (R² = 0.006, P < 0.05) in publication lifetime over the years (Fig. S2A), this change was not significant (R² = 0.00004, P = 0.88) when data of authors who started publishing after 2010 were excluded (Fig. S2C) because these authors may still be publishing in the future. Again, a weak decreasing trend of publication lifetime (R² = 0.01, P < 0.05) could be detected when all one-time authors were excluded from the linear regression analysis (Fig. S2B), but this trend was again not significant (R² = 0.0002, P = 0.77) when data for authors who started publishing after 2010 were also excluded (Fig. S2D). Furthermore, there was no significant evidence (P > 0.05) for a change in productivity over time, whether Vanhöffen was included (Fig. S3A) or excluded (Fig. S3B).

Multi-authored descriptions became more abundant during the late 19th century but stayed relatively low until the late 1960s (Fig. 6). Since the beginning of the 21st century multi-authored descriptions outnumbered the number of species described by a sole author (Fig. 7), peaking at a proportion of about 70% of new species descriptions during the 2010s and slightly over 90% within the first three years of the current decade (Fig. S4B). In contrast, the number of descriptions published by one-time authors is negligible (Fig. 6). Their proportions were high in the early history of isopod discovery (Fig. S4A) when the overall number of descriptions was low. However, since the late 19th century, the contribution of one-time authors to isopod taxonomy has been small. During this time span, the highest proportion of one-time authors was found in the current decade with close to 7% (Fig. S4A). In the last “full” decade , the 2010s, the proportion of descriptions by one-time authors was about 5%.

Named and unnamed species diversity

A decrease in the annual number of species described started more than three decades ago for all isopod species. Because this trend is not a short-term one, it cannot be explained by a time lag in data entry into the database. Estimates for future descriptions of species new to science from the non-homogeneous renewal process model predict approximately 660 additional species to be described until 2100. This suggests that 94% of isopod species that are predicted to be named by the end of this century already have been described. For other groups it has been estimated that about two thirds of all species are described, including stoneflies (DeWalt & Ower, 2019), scale insects (Deng et al., 2016), polychaete worms (Pamungkas et al., 2019), amphipods (Arfianti, Wilson & Costello, 2018) and the world’s marine species in general (Costello, Wilson & Houlding, 2012). Bryozoans have been labeled “one of the better-known taxa on Earth” due to the fact that about 80% of species predicted to be named by 2100 already had been described (Pagès-Escolà et al., 2020). Therefore, isopods represent a very well-known taxon. Of course, as more data will become available in the future these predictions may change. Bebber et al. (2007) showed that unless a taxon’s species inventory is at least 90% complete extrapolations based on existing data may be associated with large margins of error.

Some more conspicuous taxa showed a decline in new species descriptions many decades ago, i.e., mammals globally (Fisher et al., 2018) and birds and flowering plants in the UK (Bebber et al., 2007). An asymptote in description rates was reached about one hundred years ago in well-studied regions, notably in Europe for mammals, birds, black corals, echiurans and euphasiid crustaceans (Wilson & Costello, 2005), as well as for fish, gastropods, sponges, cnidarians, echinoderms, bryozoans and tunicates in Britain and Ireland (Costello, Emblow & Picton, 1996). Although an asymptote has not yet been shown for taxa globally, our data suggest that it may be emerging for isopods. Time will provide the confirmation needed. Similar analyses to those presented here for other taxa may show them to be reaching an asymptote as well.

This study did not take into account the number of already discovered but not yet formally described isopod species deposited in museum and research collections. Fontaine, Perrard & Bouchet (2012) noted an average shelf life of 21 years between the discovery and the taxonomic description of a new species. However, they also found that aquatic species have a shorter shelf life than terrestrial ones and that the shelf life of newly discovered invertebrate species is shorter than for plant or vertebrate species. For recently described isopod species, shelf life varied between 0 years (Monticelli Cardoso, Bastos-Pereira & Lopes Ferreira, 2022) and 54 years (Williams, Boyko & Marin, 2020) with an overall tendency toward the lower range of the spectrum. For example, Malek-Hosseini et al. (2022) described a new groundwater species from Iran within three years of sampling, additionally using molecular data to corroborate its species status. In contrast, the material from which the first bopyrid isopod species from hydrothermal vents was described was collected 21 to 10 years before its taxonomic description (Kato et al., 2022). Naturally, field sampling continues to unearth new species. Depending on the sampling location, the proportions of reported unnamed isopod species in field samples may vary from none (in historically well-studied areas like Europe) to about 18% (López-Orozco et al., 2022 identified three new terrestrial species) and up to as much as 93% (Poore et al., 2015 found that only 9 of 127 marine species from western Australia were previously known to science). From the latter study, none of the sampled species were identifiable with any of the 359 isopod species collected on the continental slope of south-eastern Australia of which 90% were undescribed at the time of sampling (Poore, Just & Cohen, 1994), making Australia a rich source of new isopod species. Similarly, Brandt et al. (2007) found that only 13% of the discriminated 674 deep-sea isopod species from Southern Ocean samples were known to science. Thus, the Southern Ocean as well as the waters around Australia may account for a high proportion of the yet undescribed species globally. However, when these species will be described remains unknown. A list of 21 studies which reported undescribed species (Table S2) contains 1,225 possible new isopod species, of which most were sampled in the deep sea and in and around Australia. Given the average description rate of 75 descriptions/year from the past 20 years, it would take about 16 years to formally describe all these species. It has to be noted that those species were undescribed at the time of publication of the respective study. It has not been checked whether any of the reported species have been formally described since and might now already be part of our dataset of globally described isopod species. However, it is encouraging that a significant proportion of yet-to-be described species may already be collected and awaiting description.

Although scientists are continuously adding new names to the isopod inventory, not all of those names will prove to be valid. Several newly described species might be placed into synonymy over the years. Bouchet (2006) suggested that 10–20% of new species described each year will turn out to be synonyms. Likewise, Appeltans et al. (2012) note that it takes time to discover synonyms and estimated that for every five newly named species, at least two had already been described. Most synonymies will likely be identified and resolved during comprehensive revisions of isopod genera or families (e.g., Stransky et al., 2020; Taiti & Monticelli Cardoso, 2020). Examination of museum specimens may reveal synonyms (Hughes, Bruce & Osborn, 2020), as well as lead to the recognition of species new to science (Garcia, 2020). Thus, taxonomic revisions can decrease the number of accepted species, as well as discover new species.

Cryptic diversity

Another issue that adds to uncertainty about the number of existing isopod species is cryptic diversity whereby species can only be distinguished using molecular methods. However, isopod and other crustacean taxonomists stated they could always find morphological differences on close examination and thus true cryptic diversity in isopods is negligible (Appeltans et al., 2012, Supplemental Information). Recent years have seen an increase in species delimitation studies using molecular data as well as integrative taxonomic approaches (Pante, Schoelinck & Puillandre, 2015), with some of them discovering putative new species. Species under scrutiny in such cryptic diversity studies tend to be geographically widespread species either in the deep sea (Raupach et al., 2007) or coastal habitats (Hurtado et al., 2016) or recognised species complexes already thought to harbour hidden diversity (Schnurr et al., 2018). Held (2003), for instance, tested the single-widespread-species-hypothesis of a morphologically variable Antarctic serolid isopod and identified two strongly distinct genetic clades uncovering an overlooked species. Likewise, a molecular analysis by Schnurr et al. (2018) disentangled two widely distributed munnopsid species complexes in Icelandic waters. Their data suggested that the Eurycope producta species complex consists of eight separate species, and the Eurycope inermis complex harbours four distinct species. Some of the discovered genetic clades could be linked to other already described species, leaving a total of seven species new to science. Even more putative new species have been uncovered during a genetic study of Haloniscus species from groundwater, springs, caves and salt lakes in Australia (Guzik et al., 2019). Each of the 26 new species was found to be restricted to a small geographical range. However, almost none of the previously unknown species detected by genetic sampling were truly cryptic species. Morphological characters could be found in just about every case, separating the new species from similar ones. Circling back to the problem of collected but unnamed species, few of the newly delimited species from molecular studies were formally described following their detection (Schlick-Steiner et al., 2007; Pante, Schoelinck & Puillandre, 2015). Most studies note that additional taxonomic work is required to fully support a species hypothesis with a combination of DNA data and morphological characters (e.g., Guzik et al., 2019; Jennings, Golovan & Brix, 2020). While molecular methods can be helpful in indicating specimens which may represent new species, and have been used since the 1980s for isopods and other taxa, there is no indication that they significantly increase description rates overall (Appeltans et al., 2012).

Taxonomic effort

The number of taxonomists describing new species of isopods has increased markedly over time, as it has for all taxa globally. Over the past fifty years, more authors have described isopod species than ever before (Fig. 4). Only for authors describing freshwater isopods has there been a steep decline within the past two decades (Fig. S1B), and this substantial decline is also evident in species numbers. Although it seems that freshwaters may not yield many more new species, it has been suggested that nonsaline environments harbour high cryptic diversity (Wilson, 2008). Indeed, a meta-analysis of cryptic diversity studies found that more posited cryptic species have been discovered in freshwater than in terrestrial or marine environments (Poulin & Pérez-Ponce de León, 2017). However, whether this genetic diversity translates into high species diversity is uncertain. Another interpretation of the decline in new freshwater species could be less taxonomic interest, but there seems no reason to assume why this may be the case.

Increasing numbers of people describing new species have been found in all similar studies (e.g., Joppa, Roberts & Pimm, 2011a; Appeltans et al., 2012; Tancoigne & Dubois, 2013; Costello, Vanhoorne & Appeltans, 2015; Arfianti, Wilson & Costello, 2018; Pamungkas et al., 2019; Pagès-Escolà et al., 2020), at least partly contradicting a not uncommon view that the field of taxonomy is in crisis (Godfray, 2002; Hopkins & Freckleton, 2002; Bacher, 2012). There is no doubt that taxonomy will benefit from more funding and renewed prestige (Agnarsson & Kuntner, 2007; Christenhusz & Byng, 2016; Higgs, 2016), but a lack of people describing new species is not evident from the data. The field of taxonomy is not in decline but changing. It modernised itself from a primarily morphological discipline towards a multi-disciplinary field including genetics and phylogeny. Integration of these different skill sets could explain the now higher number of multi-authored descriptions. To avoid this trend of increasing proportions of multi-authored descriptions from affecting the trend in numbers of active taxonomists over time, only the first author of a species description was considered in our analysis. Therefore, the given numbers of authors contributing to isopod taxonomy are an underestimate of the taxonomic force. Also, the proportion of authors who described only a single isopod species has not increased for more than a century. Nor have taxonomists’ publication lifetimes significantly decreased over this time. This further indicates that the increased number of taxonomic authors is an increase in effort, as concluded by others on other taxa (Joppa, Roberts & Pimm, 2011b; Appeltans et al., 2012; Essl et al., 2013), and not reduced by having proportionally more part-time taxonomists or more people who stop publishing descriptions after only a few years.

The present analysis did not consider the level of expertise of every author because this could not be determined from the available data. Some are well-established taxonomists who have spent a lifetime building up their extensive knowledge of a taxon and can therefore be considered true experts. Others are at the start of their career and still working towards expert status. Again, others contribute an essential amount of their work in other research fields, nevertheless adding valuable information with every published species description. Some people do not think it appropriate to call everyone who describes a species a taxonomist (Wheeler, 2014) and most likely, not everyone who does describe a species now and then would characterise themselves as such. However, regardless of which labels one puts on the authors of species descriptions, the fact remains that all of them contribute to the scientific inventory of the planet’s biodiversity and draft testable hypotheses. Our data show that the percentage of people who publish only a single species description is tiny and has not increased for over a century. For more information on the perceived and detectable loss of expertise and the state of taxonomy in different countries, see Lovejoy et al. (2010), Boxshall & Self (2011) and Coleman (2015), and the Australian Academy of Science (Taxonomy Decadal Plan Working Group, 2018). These assessments of taxonomy in the UK, Canada and Australia and New Zealand all considered people who described new species as a sub-set of all those working in taxonomy.

Although there have never been so many taxonomic authors than in recent decades, the average annual number of isopod species described per taxonomist has declined strongly over the last century. Such a decline in species per taxonomist has also been found for the closely related Amphipoda (Arfianti, Wilson & Costello, 2018) and for other taxa, such as scale insects (Deng et al., 2016), flowering plants (Joppa, Roberts & Pimm, 2011a), as well as spiders, amphibians, birds and mammals (Joppa, Roberts & Pimm, 2011b), marine and terrestrial parasites (Costello, 2016), fossil and extant marine bryozoans (Pagès-Escolà et al., 2020) and overall marine and non-marine species (Costello, Wilson & Houlding, 2013). The reduction in the description rate of isopod species observed here, despite peak numbers of taxonomists, suggests that most species have already been named, as concluded for other taxa (Joppa, Roberts & Pimm, 2011b; Arfianti, Wilson & Costello, 2018; Pamungkas et al., 2019). Contradicting this interpretation, Sangster & Luksenburg (2015) proposed that the lower number of species described per taxonomist is rather a consequence of the improved quality of species descriptions than a slowdown of progress in species discovery. They found that the number of pages of taxonomic descriptions has increased compared to the 1930s. So has the number of specimens on which the description of a new species is based, the number of characters to differentiate it from its most closely related species and the number of illustrations in a publication. With this increased effort put into the scientific description of a species, it may take more time from the initial discovery of a species until the publication of its formal description. However, other studies point to greater efficiencies in taxonomy due to greater access to field samples and literature, and improved museum collections, laboratory methods, publication efficiency, and communication between people (Eschmeyer et al., 2010; Costello, Houlding & Wilson, 2014). We found a similar productivity of taxonomists over their isopod-description careers, indicating that modern efficiencies and co-authorships may indeed balance out the richer species descriptions.

At the upper end of productivity, 21 taxonomists (only 3% of the taxonomic workforce over time) have described approximately 43% of all known isopod species. The three most prolific authors described almost exclusively terrestrial isopod species, which are more easily accessible and can be sampled without the deployment of advanced sampling equipment by comparison with marine isopods. Accordingly, our model estimates suggest that a considerable proportion of future discoveries might be made in the less accessible marine environment. Also, because large and geographically widespread species tend to be named first (Costello et al., 2015; Higgs & Attrill, 2015), many of the yet-undiscovered isopod species are likely to be small and/or geographically restricted species (Scheffers et al., 2012; Liu et al., 2022). There is speculation on whether most of the yet-undescribed species will be found in collections (Scheffers et al., 2012; Coleman, 2015) or will be newly discovered during fieldwork (Grieneisen et al., 2014). However, both named and unnamed species, especially freshwater and endemic species, are at risk of extinction due to human impacts (Costello, 2015; Liu et al., 2022). Because many new species tend to be discovered in biodiversity rich-spots, which already face many threats like extensive habitat loss, they will be more vulnerable (Scheffers et al., 2012; Manes et al., 2021) and are at risk of going extinct before they are even discovered (Costello, May & Stork, 2013). It is therefore important that taxonomists continue to describe new species. Only named, and as such well delimited species, can be included in threat reports and conservation plans.