General approach and editorial standards

The backbone classification follows phylogeny-based taxonomic treatments as far as they are available; otherwise, appropriate classical or conventional taxonomic works are used as the basis. The classification with respect to the acceptance of taxa follows treatments that were published, in press or at least submitted during the compilation of the present checklist. The taxonomic concept or circumscription of the accepted taxa is indicated by the secundum (“sec.”) reference (Berendsohn 1995). Orthographical corrections may have been applied independently. In addition, several necessary combinations and validations of names are made here (see Nomenclatural novelties below). For those names that we have excluded from the core checklist (unresolved names, names of uncertain application, artificial hybrids), the sec. reference normally indicates the source for the respective placement of the name. For synonyms, the “syn. sec.” reference indicates the assignment of the synonym to the concept of either the accepted name or one of its homotypic synonyms; this may or may not be the same reference as that of the secundum of the taxon. The statement of the nomenclatural status can also be independent of the sec. or syn. sec. reference of a name and is referenced separately; the sources are given on the website  http://caryophyllales.org/cactaceae/Checklist, as well as in the online  Supplemental content (wi.51.51208_Checklist.pdf) of the present publication.

General editorial standards regarding the citation of author names and nomenclatural references in many respects follow the rules formulated by Kilian & al. (2009+) for the Cichorieae Network, another initiative using the EDIT Platform. Titles of monographs are abbreviated in conformity with Taxonomic literature, ed. 2 (Stafleu & Cowan 1976–1988; Stafleu & Mennega 1992–2000; Dorr & Nicolson 2008, 2009), but all components start with capital letters. Titles of monographs not listed in Taxonomic literature follow the version used in IPNI, with the exception that work's editor is not included in the title. Titles not mentioned in any of these sources are abbreviated as in serials. Titles of serials are abbreviated in conformity with Botanico-periodicum-huntianum (Bridson & al. 2004, online edition); titles not listed there are abbreviated according to the standards therein defined. The publication year always refers to the actual year of publication of the page cited. If a work has been published in parts and its publication dates have been established, the publication year is one of the corresponding parts and can then differ from the year(s) on the (main) title page of that work. In the case of known differences between the year given and the real publication date, the former is cited in square brackets and set in quotation marks. Series that form part of the title are normally included in the standard abbreviation for nomenclatural citations.

The tools

The EDIT Platform for Cybertaxonomy ( https://cybertaxonomy.eu/), (Ciardelli & al. 2009; Berendsohn 2010; EDIT 2019) is a suite of open-source software tools and services, which has been developed over the past 20 years with the aim of covering all aspects of an integrative taxonomic workflow (Borsch & al. 2015; Kilian & al. 2015; Henning & al. 2018). The underlying Common Data Model (CDM) has been constantly refined to mirror the intricacies of nomenclatural and taxonomic information, including all aspects of monographic and floristic work. The taxonomic editor software, online data portal and publication pipelines of the EDIT Platform are used in the Caryophyllales Network to capture, process, attribute, document, publish and maintain the data. Its role in the context of the Caryophyllales synthesis and the features that distinguish it from other tools for taxonomic compilations are more fully detailed in Borsch & al. (2015) and Berendsohn & al. (2018). It should be mentioned that the ongoing work on species-level checklists of core Caryophyllales families is carried out in a single underlying CDM database, providing synergies with respect to jointly used resources such as reference citations.

Data compilation and principal sources

Subfamilies and tribes — We have refrained from including an infrafamilial and tribal classification at this time. All genera can be assigned to subfamilies, with the majority belonging to the two major subfamilies Opuntioideae and Cactoideae. The genera Leuenbergeria Lodé, Maihuenia (Phil. ex F. A. C. Weber) K. Schum. and Pereskia Mill., which form the basal Cactaceae grade, would have to be placed in their respective monogeneric subfamilies; the names do exist. Tribes could be given for most taxa as well, yet some genera have not yet been sampled in phylogenetic analyses or poor node support so far did not allow placing them confidently at the tribal level. The names of subfamilies and tribes can certainly be included at a later stage to make the checklist more complete, but the current simple alphabetical list of genera is intended to make it easier for users to find taxa.

Accepted generic names — Accepted generic names and their synonyms were taken from the Caryophyllales genus-level checklist (Hernández-Ledesma & al. 2015), with additions from more recent phylogenetic studies that included changes to generic concepts. In the (rare) cases of conflicting treatments, the respective editor of the partial treatment simply decided on a source to follow and documented it in the notes accompanying the genus entry.

Infrageneric names (i.e. names at ranks below genus and above species) were not included, unless the currently accepted generic name was based on an earlier infrageneric name.

We have defined five categories to assess the robustness of the circumscriptions of genera, described in more detail below. Table 1 gives an overview of this assessment, as well as the literature used as circumscription references for the genera (“sec.” references). Discussion notes in the actual checklist provide more detailed information about the current state of knowledge in terms of monophyly or phylogenetic relationships.

Table 1.

Overview of accepted genera, their species numbers and an assessment of the monophyly and the phylogenetic knowledge of each genus (see text for more details). – A: genera that are monophyletic based on a complete or very comprehensive phylogenetic study, and a phylogeny-based taxonomic synopsis was published; B: genera that are monophyletic based on phylogenetic studies that support the clade based on a sufficiently dense or even complete sampling or support a monotypic genus as a distinct lineage; C: genera that are probably monophyletic, but too few of the species were sampled in phylogenetic studies to reliably establish monophyly; D: genera that are polyphyletic, paraphyletic or nested in another genus; E: genera for which no assessment of monophyly has yet been possible, because only a few species were sampled or no phylogenetic study has been conducted so far. See the Discussion for a more detailed explanation of these categories. The column “Names total” does not count the name of the genus itself.

Continued.

Continued.

Continued.

Continued.

Accepted species names and synonyms — A list of all Cactaceae names with nomenclatural references used in the World Flora Online backbone based on The Plant List 1.1 (TPL 2013) was received from the WFO Data Centre in February 2018. These 17,531 names had each received a unique WFO-identifier (a unique number) and were to be resolved completely, to be able to reimport the new backbone without disrupting the data linked to the names in the WFO database. The names were converted into a standard format so as to enable cut-and-paste data entry into the EDIT Platform's Taxonomic Editor software, which then automatically parses names and nomenclatural references into the corresponding fully atomized fields in the relational database. Additional names from literature were entered similarly.

Data entry started with those genera for which a phylogeny-based treatment, including a synopsis of species and their synonyms, was available and which therefore could be directly adopted for this backbone.

For the more common cases, where phylogenetic studies establish the circumscription of the genus but do not provide a new taxonomic treatment at the species level, a suitable recent taxonomic work or monograph was used. For genera that lack a sufficient level of phylogenetic robustness or have not been sampled in a phylogenetic study so far, available treatments as detailed below were used to compile the list of species, complemented with more recently described new taxa. Works considered suitable for this purpose were monographs of genera or subgenera or sections, checklists of tribes or genera, lexicographic family-wide treatments, floristic treatments and regional checklists. Among the most important sources in the latter three categories were The new cactus lexicon (Hunt 2006), the 3rd edition of the CITES Cactaceae checklist (Hunt 2016), the Catálogo de cactáceas Mexicanas (Guzmán & al. 2003), the Cacti of eastern Brazil (Taylor & Zappi 2004), the Flora do Brasil (Flora do Brasil 2020), the Catálogo de las plantas vasculares de Bolivia (Kiesling & al. 2014), the Cactaceae treatment in the Flora of North America (Parfitt & Gibson 2003), the Opuntioideae checklist for North and Central America (Hernández & al. 2014) and Mapping the cacti of Mexico (Hernández & Gómez-Hinostrosa 2011, 2015). Further references are cited in Table 1 and a complete list of all sec. references used for the checklist can be found in the  Supplemental content (wi.51.51208_Checklist.pdf) online (see  https://doi.org/10.3372/wi.51.51208) and at  http://caryophyllales.org/cactaceae/bibliography.

For many genera, there was more than one suitable reference that could be used, and the final choice of a secundum reference was left to the contributors for the individual genera. When deciding as to which sec. reference would be used for accepted species, priority was given to phylogenetic and taxonomic research published in peer-reviewed journals, monographs, floras, country/regional checklists and family-wide lexicographic treatments. Newer works were often preferred over older works. Taxonomic synopses in non-peer-reviewed journals and amateur journals were used if the editor responsible for the respective genera considered them of sufficient quality, no alternative was available and if the information therein was necessary from a nomenclatural standpoint. Online databases were normally not used as sec. references for accepted taxa but were commonly used for synonyms. Finally, there are several cases where the present checklist implements new generic concepts for the first time, in which cases the respective partial treatment itself is the sec. reference.

Remaining synonyms, unresolved names, and names of uncertain application — Close to 8000 names were included into the backbone after completing the working process as described above, still leaving almost 10,000 names from the WFO backbone. Of those, only 124 represented duplicates or artefacts and three were synonyms applicable to taxa in other families — these cases are reported to WFO for their elimination from the WFO backbone and are not treated any further. The remainder included names that had never been used in a recent treatment, e.g. names in segregate genera not accepted by later authors, names not validly published, names not listed in synonymies and names of hybrids. The vast majority of these names were combinations based on basionyms already covered in the list and were therefore easily resolved through IPNI (2020), Tropicos (2020) and Plants of the World Online (POWO 2019), as well as the original World Flora Online backbone dataset based on The Plant List 1.1 (TPL 2013). Many names published in the 19^th century and not listed in either Tropicos or POWO had been included in the monograph of Britton & Rose (1919–1923) and could thus be resolved. The resulting database of 18,686 names was matched with the World Checklist of Vascular Plants (WCVP) received from Royal Botanic Gardens, Kew in December 2019 (WCVP 2019), which resulted in some 4000+ further names, in their majority infraspecific names, that had not been covered in The Plant List or recent monographs, as well as numerous names of artificial hybrids. After identifying misspellings and orthographical variants, these names were also incorporated into the backbone.

For names that could not be classified as belonging to one of the taxa, so-called “pseudotaxa” were created and included in the taxonomic hierarchy immediately underneath the family. This allowed including the names in the database, finding them in searches and indicating their special status.

Four kinds of pseudotaxa were defined: (1) “names of uncertain application”, including nomina nuda; (2) names of artificial hybrids (natural hybrids are listed in the core checklist); (3) “unresolved names”, including all further names that could not be assigned to one of the preceding categories; and (4) incertae sedis, gathering species or subspecies that should probably be accepted, but their placement in a genus is not clear. Under the heading “excluded”, are names that could be positively excluded for a certain reason, which is given with the name.

Revision and community input — The initial compilation of the checklist was mainly done by the first author of this publication. After its completion, Cactaceae taxonomists, systematists and phylogeneticists were invited to revise the treatments for genera that they were free to select, and most of those invited responded positively. The same approach had also been successfully used before for the Caryophyllales genus-level checklist (Hernández-Ledesma & al. 2015).

The revisions encompassed reviewing and adjusting the status of the accepted species names and synonyms, adding potentially missing names, adjusting sec. references used in the respective partial treatment, writing taxonomic notes and, if needed, providing new combinations, validations and typifications. This contribution was the basis to be listed as a co-author of the checklist and of this publication. To facilitate the revision process, a database output was provided as a simple Word file and all revisions and comments were made therein; the changes in the database were subsequently made by the first author. The extent to which the partial treatments had to be revised differed between the different genera and was expectedly higher for genera, for which no up-to-date treatment was available, and therefore the revisions took between several weeks to about two months to complete.

Assessment of the state of phylogenetic knowledge for genera

Table 1 provides an assessment of the reliability of the generic circumscriptions. Hunt (2006) in The new cactus lexicon, attempted a scheme to indicate the reliability of the taxonomy, qualifying the information source as: (A) monograph; (B) regional or partial specimen-based revision; (C) semi-popular handbook/study-group publication and/or older regional treatment; and (D) provisional assessment, often protologue-based. Because most phylogenetic studies based on DNA sequences in Cactaceae were published only after 2006, this can now be used as a further indicator. Here we use the extent of the phylogenetic knowledge included in the work as the main indicator of taxonomic stability, using five categories:

Fig. 1 compares the numbers of genera per category, with encouraging results. Most of the genera have already been included in phylogenetic studies and fall in either category A (21 genera) or category B (84 genera). Therefore, over two-thirds of all Cactaceae genera have already been confirmed as monophyletic based on phylogenetic studies.

The recent trend has clearly been reinstating earlier-described genera or segregating small genera based on phylogenetic studies, which consistently recover small clades corresponding to genera and support the many previously published monotypic genera as isolated lineages. Thirty-six genera are monotypic and a further 40 have between two and five species.

Species numbers in Cactaceae

The main achievement of this checklist is an updated phylogeny-based, species-level taxonomic backbone. As for the reliability of the overall species number, this varies between different genera and is linked to the taxonomic stability of the genera, as explained above.

The most reliable basis for such a species-level backbone is a phylogeny-based synopsis or monograph (category A). These include 21 genera with 225 species and 37 heterotypic infraspecific taxa, which is slightly above 14% of all Cactaceae for which we have this reliable assessment of species diversity. Most genera fall into category B, and while no synopsis was published based on the phylogenetic study, there are often older classical treatments or one of the recent family-wide synopses that provide a rather reliable basis for the species to be included and accepted in these genera. This concerns 84 genera containing 1104 species plus 291 heterotypic infraspecific taxa. So, in terms of species numbers, almost 60% of the family has a solid basis for the accepted species, or 72% if all taxa, including heterotypic infraspecific taxa, are considered.

It is now interesting to compare this checklist with earlier family-wide treatments and estimates of species numbers. Table 2 shows a comparison to the two recent reference works covering the whole family, and notably our results match the treatment of Anderson (2001, 2005) very closely, although the taxa are certainly only partly congruent. Hunt (2006) on the other hand took a much more conservative approach in accepting genera and species.

Many new taxa have been described since 2006: 217 new species, 4 new hybrid species, 101 new subspecies, 35 varieties and 9 forms, yet not all of them are currently accepted here.

Many Cactaceae species are known to have a large number of synonyms, mostly resulting from their variability, popularity in collections and a long history of cultivation, and from changing generic concepts and moving species from one genus to another. Table 3 lists the three species with the most synonyms, with Aylostera pygmaea (R. E. Fr.) Mosti & Papini being the most extreme example with 256 synonyms in total, but Melocactus macracanthos (Salm-Dyck) Link & Otto having the most heterotypic synonyms: 124 homotypic groups in total, compared to only 81 homotypic groups in A. pygmaea.

We are aware that estimating species limits using current phylogenetic methods is generally harder than inferring generic limits and that the results of such approaches generally harder to interpret due to individuals of the same species or infraspecific taxon not being resolved as monophyletic. Also, many phylogenetic studies do not include enough individuals to be able to infer species limits in detail. And certainly there are species or species complexes that are not sufficiently understood and require more detailed study. An approach to flag species with the current state of research on species delimitation has been taken by Borsch & al. (2018) for the genus Iresine P. Browne (Amaranthaceae).

But disregarding the shortcomings, this checklist is not only phylogeny-based, but also expert-revised, making it a reliable source for the species diversity of the Cactaceae.

Secundum references

Currently, some ambiguity exists with respect to the secundum references as cited in the checklist. In the EDIT Platform, there are two possibilities to cite references for the context of a taxon: (1) the circumscription reference, coined as the secundum (sec.) reference by Berendsohn (1997), should cite a publication where the circumscription of the taxon or its differentiation from other (related) taxa of the same rank is treated; (2) the placement reference, which is the source for the assignment of an already defined taxon to a place in a given classification (in the core checklist: the placement in a taxonomic group of higher rank). In analogy to (2), the “syn. sec.” reference depicts the placement of a name in the synonymy of a taxon, and the “err. sec.” reference (not yet used in the Cactaceae dataset) is the source of the inclusion of a taxon that has been misnamed in its sec. reference elsewhere (which is then cited as “sensu”). However, in the present state of our data, the sec. reference in some cases represents the source of a name (particularly when the taxonomic position of that name is not yet clear), or the reference for the placement (especially if the placement resulted in a changed name), or even the reference for the source of a name or taxon under one of the categories excluded from the core checklist. Furthermore, the data entry does not at present support multiple sec. references, so the responsible editor must choose one over others that may be equally relevant. These issues are being discussed in the EDIT Platform's user and developer communities and will be solved as soon as possible.

Table 2.

Comparison of the numbers of accepted taxa in the two most widely used works and our results.

Fig. 1.

Comparison of the numbers of genera per category of monophyly and taxonomic robustness. – A: genera that are monophyletic based on a complete or very comprehensive phylogenetic study, and a phylogeny-based taxonomic synopsis was published; B: genera that are monophyletic based on phylogenetic studies that support the clade based on a sufficiently dense or even complete sampling or support a monotypic genus as a distinct lineage; C: genera that are probably monophyletic, but too few of the species were sampled in phylogenetic studies to reliably establish monophyly; D: genera that are polyphyletic, paraphyletic or nested in another genus; E: genera for which no assessment of monophyly has yet been possible, because only a few species were sampled or no phylogenetic study has been conducted so far. See the Discussion for a more detailed explanation of these categories.

Table 3.

Species with the most synonyms.

Future work, interaction with the World Flora Online and some technical aspects

As a joint community effort within the Caryophyllales Network, the taxonomic backbone for Cactaceae will be continuously updated according to newly published results of phylogenetic and taxonomic research and with respect to further names that may be found in obscure and hard-to-get historical publications once they become digitized and freely available. The addition of further nomenclatural information, such as links to protologues and to images of type specimens, will be an ongoing task. Also, further work on the currently unresolved names is needed.

The backbone serves as a very useful tool for future monographic and floristic treatments, because taxonomic and nomenclatural decisions accumulate and do not have to be entered anew, and the output from the EDIT Platform can be taken as a baseline. It is also planned to provide the output from this database to other providers of taxonomic or nomenclatural data, especially those that have directly or indirectly contributed to the data (Kew WCVP, Tropicos). Providing data to the Catalogue of Life ( https://www.catalogueoflife.org/) is planned as well and will be done through WFO.

Outputs in various formats are possible from the EDIT Platform database. One of them, the so-called CDM-light format (Luther & al. 2019), consists of a series of interrelated, comma-separated value (CSV) tables that, e.g., provided the base for the statistics drawn from the data and for the transformation of the data into the formatted checklist published here in the  Supplemental content (wi.51.51208_Checklist.pdf) online. The document-type output also greatly facilitates input from collaborators who do not interact directly with the database. Another platform output format follows the standard Darwin Core Archive format defined by the WFO consortium, so that a pipeline from the Caryophyllales Taxonomic Expert Network to the WFO database is established. The checklist data will be provided to the WFO gatekeeper as a replacement of the current, TPL-based taxonomic backbone for the WFO portal. As part of the Caryophyllales Network's responsibilities as a Taxonomic Expert Network for WFO, all previously used WFO name IDs were revised, and 4695 new IDs were assigned to names that were not previously present. This will also allow future straightforward input and machine-readable interaction with WFO. Specific outputs can be requested from the Caryophyllales secretariat at caryophyllales@bgbm.org.

The EDIT Platform also allows entering further information that can be associated with a taxon, such as morphological descriptions, specimens and specimen descriptions, molecular data, photographs, common names, etymology, chromosome counts, anatomical information, state of conservation, etc.

The first species-level taxonomic backbone treatment that was completed in the context of the Caryophyllales initiative was that for Nepenthaceae (176 spp.) (Berendsohn & al. 2018; Caryophyllales Network 2015+;  http://caryophyllales.org/nepenthaceae/home). In comparison to the present Cactaceae treatment, the Nepenthaceae treatment exemplifies the additional inclusion of descriptive and other monographic information. A possible further goal for Cactaceae would be to add distribution data, first at TDWG level ( https://www.tdwg.org/), which are currently exemplarily available in the portal for the Rhipsalideae only. The ultimate aim is to provide referenced information for distributions, either by means of area records, as exemplified by the Euro+Med PlantBase (Korotkova & Raab-Straube 2017+) or by means of voucher specimens. Distribution data for Cactaceae are available at the country level in basically all family-level synopses and on a finer scale in floras, regional treatments and synopses. Species-level distribution maps based on such literature sources have also already been published for the whole family (Barthlott & al. 2015). Hence, literature data already contain area records for the whole family, which can be added to the backbone in the future. This can be complemented by links to global aggregators of species occurrence data, such as GBIF ( https://www.gbif.org/). However, distribution information for taxa with a defined concept considering expert-revised voucher and/or publication information will continue to provide the most reliable source for, e.g., decision-making in conservation.

Such goals will also require further development of the collaborative workflows in taxonomy and their required technical bases. For example, linking and integrating flora treatments with global monographs is one of the principal challenges for the botanical community and a major research topic in biodiversity informatics.

As for the mechanisms of future updates, we intend to keep it as up-to-date as possible. There are no specific cycles of review planned, but the checklist will be updated whenever necessary and IPNI will be checked periodically for new names. The database output will also be published at irregular intervals as a citable PDF publication with a DOI through BGBM-Press ( https://www.bgbm.org/en/publikationen-publications).

Experts on Cactaceae are invited to contribute as editors in the future. It is important to emphasize that this Cactaceae backbone is, and must remain, a community work.