Introduction

The concept of biodiversity, applied to floristic richness, considers the number of taxa (categories of the taxonomic hierarchy) present in any geographical or administrative unit, such as county, state or country. With this number, it is possible to quantitatively evaluate diversity and compare it among areas. There are international agreements that prioritize the quantification of biodiversity of the signatory countries, especially those with poor or insufficient knowledge of biodiversity at the national and/or regional levels, as is the case of Mexico (^{Conabio, 2012}).

Mexico has a long and growing tradition of studying its vascular flora, reflected in the significant increase in recent decades of specimens housed in national scientific collections and abroad, backed by an immense bibliography. However, the knowledge of national floristic richness is still unsatisfactory mainly due to the difficulty of synthesizing scattered information in such publications along with the lack of well-curated databases of specimens documenting this richness. It is also clear that most genera require additional taxonomic study (revisions or monographs), and large areas of land remain unexplored to date.

The first estimates of the vascular flora of Mexico, proposed more than 2 decades ago, quoted between 17,000 and 30,000 species (reviewed in ^{Villaseñor, 2003}). A decade ago, an extensive literature review led to an estimate of about 22,351 species of vascular plants (^{Villaseñor, 2003}, ²⁰⁰⁴). Later, ^{Llorente-Bousquets and Ocegueda (2008)}, collaborating with many specialists, published the first list of species of vascular plants of Mexico, which included 22,332 species, a figure remarkably similar to that reported by ^{Villaseñor (2003)}. Their list was the first publication that documents in detail aspects of Mexican plant biodiversity (^{Conabio, 2008}). Unfortunately, the exercise carried out by the Conabio (the Mexican Biodiversity Commission) has been little used, probably because the general public has limited access to such information, and the databases still have little impact on the presentation and management of biological information. Moreover, due to the dynamism of taxonomy, published scientific names are constantly changed due to updates and corrections, or added in publication of numerous new species. Therefore, the documented information on Mexico's floristic richness should be regularly updated through the publication of floristic lists or catalogs that synthesize information on the species reported.

A catalog (floristic list or checklist) represents a more or less critical summary of the information gathered or known about the plant species (or other taxonomic designation) of a region (^{Nimis, 1996}), and it may vary in content or approach. Sometimes, as in this work, they only list the scientific names collected for the country; on other occasions they provide additional information, such as representative specimens, literature for particular taxonomic groups, synonymy, or specific comments aimed to clarify doubts or taxonomic conflicts (see, for example, ^{Dávila et al., 2006}; ^{García-Mendoza & Meave, 2011}; ^{Guzmán, Arias, & Dávila, 2003}; ^{Ibarra-Manríquez, Villaseñor, & Durán, 1995}; ^{Villaseñor, Ortiz, Beutelspacher, & Gómez-López, 2013}). Usually, a basic species list is what is first published for any region, so it always requires a critical evaluation. The reliability of the existing literature is an issue for any catalog and unreliable primary sources result in biases or difficulties in compiling lists; in addition, it is practically impossible to verify all species identifications, and the number of reviews or monographs consulted or available is relatively low. A further problem is the use of different taxonomic criteria; specialists do not always coincide in the circumscription of species, genera or even taxonomic categories of higher rank, and reconciling these different treatments is not trivial. Sometimes an inventory follows one of these criteria, while another may prefer an alternative approach. Expert opinions help clarify uncertainties, especially when the geographical distribution of species reveals errors of reference in a given regional inventory. Ultimately, it is up to catalog users to judge the reliability of names and additional information presented.

Many people are skeptical of the scientific value of catalogs, especially scholars of biodiversity that require information that catalogs do not provide directly. However, among the merits of catalogs is that they synthesize a wealth of information accumulated throughout the history of botany and exploration of any territory. Listing species names is key to accessing a world of additional information on species, including aspects of natural history and current and potential uses. For taxonomists, lists are certainly valuable in order to consider the number of species to study in a review or monograph, and facilitate the inclusion of many species that have been overlooked in previous treatments. For ecologists and phytogeographers, catalogs are the first step to forming an informed opinion on the relationships between certain floras, and attempting to explain the causes, origins and evolution of diversity.

The aim of this contribution is to provide an updated catalog of the native vascular plant species of Mexico. This catalog is expected to serve as a basis for a better understanding of the Mexican flora and to promote more comprehensive floristic and taxonomic studies of groups or regions that require more detailed inventory or systematic work. As ^{Nimis (1996)} points out, catalogs are catalysts for new research projects and questions, and their relevance is not limited to floristic or taxonomic studies. The intention of this list of species was 2-fold: first, to document the current state of floristic knowledge, and second, to provide a basic reference that specialists can use to compare their data and more efficiently perform future taxonomic reviews. By examining the information provided here, it will soon be possible to handle more precise information on the floristic richness of the country and progress toward the long-awaited goal of having a flora of Mexico.

Materials and methods

The catalog is the result of the review of over 2,500 references covering different aspects of the flora of Mexico. Among them are the numerous fascicles published by different regional taxonomic treatments for the country. They include, for example, Flora of El Bajío and adjacent regions (>190 fascicles and >30 additional fascicles), Flora of Veracruz (>150), Flora of the Tehuacán-Cuicatlán Valley (>130), Flora of Guerrero (>60), Flora of Jalisco (>20), Flora Mesoamericana (5 volumes) and Flora Novo-Galiciana (17 volumes). The protologues of many species that had never been mentioned in previous inventories or floristic treatments were also consulted. For example, in the last decade (2006-2015) 924 brand new species occurring in Mexico have been described, and 656 species have undergone name changes due to the proposal of new taxonomic combinations. Very few of these 1,580 names had been mentioned in publications of inventories or vegetation studies, and their inclusion in the catalog derives directly from the publications where they were originally described.

The compilation of this catalog has also benefited from the large body of published state and regional inventories published. In his 2 early-century reviews, ^{Villaseñor (2003}, ²⁰⁰⁴) reported the major contributions of these publications to the flora of Mexico, and pointed out the 13 states lacking statewide inventories at the time. Interestingly, updated floristic lists have been published for 5 of them since (Ciudad de México - formerly known as Distrito Federal [^{Rivera-Hernández & Flores-Hernández, 2013}]; Jalisco [^{Ramírez-Delgadillo et al., 2010}]; Nuevo León [^{Villarreal & Estrada, 2008}]; Oaxaca [^{García-Mendoza & Meave, 2011}]; and Puebla [^{Rodríguez, Villaseñor, Coombes, & Cerón, 2014}]). Moreover, 284 local and regional inventories from all over the country have been reviewed (reference list available from the author upon request). It is unfortunate that many important works are available in libraries only as “gray literature” (e.g. technical reports, unpublished undergraduate and graduate theses; ^{Corlett, 2011}). Figure 1 illustrates the geographic distribution of these reviewed inventories; the area included in each inventory is represented by a circle centered at the intersection of the long and wide axes of the surveyed area. Thus, the size of the circle is proportional to the studied surface.

Figure 1 Mexican localities with floristic inventories (N=284). The size of the circle is proportional to the studied surface in each inventory. Total surface covered by the studies is 378,887km² (19.2% of Mexico's territory). 

Undoubtedly, another invaluable source of data was the intense review of the library and specimens housed at the National Herbarium (MEXU) at the Instituto de Biología, UNAM and internet databases (IPNI, REMIB, The Plant List, Tropicos, Unibio, etc.). Finally, numerous experts also lent valuable support reviewing draft lists of their specialty and providing relevant information on species important literature that is difficult to access or not yet recorded in the consulted information sources. Many of them (Table 1) should be cited as responsible for these groups in the catalog, though any nomenclatural discrepancies and the accepted name included in the list are the sole responsibility of the author.

Results

Floristic richness nationwide

So far more than 54,000 names to compile the final list have been reviewed, which is included as Appendix 2. The number of accepted species is 23,314, distributed in 73 orders, 297 families, and 2,854 genera (Table 2). In addition, there are 1,414 subspecific categories (subspecies, varieties or forms) that were not considered in the analysis; since there is uncertainty in assigning these taxa appropriately to the states, biomes or regions where they were reported, I preferred to document diversity to the species level. Considering these subspecific taxa would have increased floristic richness to about 24,728 native taxa. The higher taxonomic categories (families and orders) agree with the most recent classification proposals. For ferns, lycophytes, and gymnosperms, the arrangement follows the proposals of ^{Christenhusz, Chun, and Scheider (2011)} and ^{Christenhusz, Reveal, et al. (2011)}, and angiosperms are arranged following the proposal of the APGIII (^{APG III, 2009}; ^{Chase & Reveal, 2009}; ^{Haston, Richardson, Stevens, Chase, & Harris, 2009}; ^{Wearn, Chase, Mabberley, & Couch, 2013}).

Appendix 1 contains the families of vascular plants in the accepted classification schemes. Families follow a numerical order, defining a linear sequence in accordance with recent classification proposals. For each family a summary of the number of genera and species accepted is presented. The genera and species are listed in Appendix 2, with species endemic to Mexico being denoted with an asterisk (*).

The list does not include naturalized exotic species, many of which escaped from cultivation. ^{Villaseñor and Espinosa-García (2004)} documented the presence in Mexico of 618 exotic species. Currently, this figure needs to be revised, because similar to the case of native species, in addition to recent taxonomic changes, some 300 additional species have been documented. Consequently, summing the richness of species, subspecific taxa, and naturalized exotics, the number of vascular plants of the flora of Mexico totals more than 25,700 taxa of infrageneric rank. Finally, the list includes 48 species and 2 genera of hybrid origin (nothospecies and nothogenera).

Determining accepted names in the catalog required the revision of many additional names considered under synonymy. Considering that 31,263 synonyms were found, the average number of synonyms per accepted species is 1.3 (mode = 1); in other words, on average each species has been described more than once. It is important to emphasize that the synonymy search was not exhaustive and is almost certainly an underestimate. The ratio is based solely on the names retrieved from the floristic-taxonomic literature for the flora of Mexico, without considering relevant information for species which also occur outside the national territory. The numbers of species accepted and considered as synonyms in different taxonomic groups is variable; however, it should be noted that few orders comprise a lower number of synonyms than accepted species. Table 3 summarizes the number of accepted species and synonyms recorded in each order to date.

Table 4 shows the 25 families and genera with the most species. Comparing these lists with those provided by ^{Villaseñor (2004)}, only Bignoniaceae, Cucurbitaceae and Scrophulariaceae are excluded, which were listed before among the families of angiosperms with the largest numbers of genera (^{Villaseñor, 2004}). Notably, this list now also includes a family of ferns (Pteridaceae). Moreover, all genera with the largest numbers of species (^{Villaseñor, 2004}) continue to be ranked highest; the only exception is the genus Opuntia, which was reduced from 90 to 72 species due to the segregation of the genus Cylindropuntia (18 species).

The 25 most diverse families include 15,306 species, representing 65.6% of Mexico's total floristic richness. The additional 34.4% is distributed unevenly among the remaining 272 families. It is worth noting that 44 families are represented in Mexico by a single species, 23 by 2 species, and 45 by 3 to 5 species; consequently, rare or poorly represented families in Mexico constitute an important fraction of the plant diversity of Mexico.

Statewide floristic richness

An important question is whether the floristic richness of Mexico is distributed evenly throughout the country or if there are areas hosting a disproportionately high richness, and if so, where are they located? Some studies have tried to answer this question, but without considering all of the flora of Mexico (for example, ^{Luna-Vega, Espinosa, Rivas, & Contreras-Medina, 2013}; ^{Rzedowski, 1991}). The identification of extremely rich areas is the recurring theme of studies focused on conserving biodiversity. This paper provides a synthesis of floristic richness at the state level in order to identify the different levels of diversity by political entity (Table 5).

Table 5 presents an update of the vascular plant richness by state provided by ^{Villaseñor and Ortiz (2014)}, supplemented with data for ferns and gymnosperms. Oaxaca, Chiapas, Veracruz, Jalisco and Guerrero respectively, continue to occupy the top 5 ranks of species richness. However, the accumulated information for other states, thanks to the collaboration with experts working with data on their floristic richness, has substantially changed compared to values recorded earlier.

Figure 2 illustrates the distribution of species, genus and family richness for each state synthesized in Table 5. The values are divided by the logarithm (log₁₀) of the state's surface area, a procedure that essentially standardizes mean species richness per 10 km². A diversity gradient from south to north is depicted, with some states of central Mexico breaking from this gradual shift of richness. Similarly, the 2 major peninsulas of the country (Baja California and Yucatán) had the lowest richness values for all 3 taxonomic levels evaluated.

Figure 2 Richness of species, genera, and families in Mexican states. Figures are standardized by area by dividing the total values by the logarithm (log10) of the state's surface (S). 

Apparently, most states showed a roughly equal proportion of species per genus (species/genus ratio; mean = 3.5, minimum = 2.3, maximum = 5.2), except Oaxaca, for which the largest ratio was recorded. The variation is more noticeable when comparing the proportion of species per family (minimum = 10.8, maximum = 38.3; average = 8.2; Fig. 3).

Figure 3 Proportion of species described by genera and families in the 32 political states of Mexico. 

Endemism

The total percentage of endemic species of Mexico among all vascular plants had not been previously reported. Most estimates are primarily for flowering plants. For this set, ^{Rzedowski (1991)} estimated their level of endemism at 52%. Subsequently, ^{Villaseñor and Ortiz (2014)} estimated the value at 50.4%. The figures included in this paper suggested that the level of endemism for all vascular plants is 49.8% (Table 5). Given the consistency between Rzedowski's estimate and these updated figures, levels of endemism are likely around 50%; in other words, about half of the flora of Mexico is endemic to its territory.

Another interesting finding from these analyses is that 3,167 species endemic to Mexico (13.6% of the total floristic richness and 27.3% nationwide endemics) are found only in 1 state. If we add to this figure those species that are not endemic to Mexico, but only found in a single Mexican state (classified here as restricted species), mostly at the northern or southern edge of the species’ geographic range, the proportion of rare species (species found in a single state) reaches 21.8% of the national floristic richness. Table 5 indicates how many of these rare species are recorded in each one of the states of Mexico.

Floristic richness in the major biomes of Mexico

In agreement with the uneven distribution of floristic richness statewide, irregular distribution patterns are also evident at larger scales, although it is feasible to recognize areas with characteristic species assemblages. For example, ^{Villaseñor and Ortiz (2014)} described 5 major biomes into which the vegetation types in Mexico are grouped, ranging in value from 5,296 species in the tropical humid forest (THUF) to 8,824 species in the temperate forest (TEMF). Here, the species richness distribution is assessed by biome, with special consideration of the species characteristic to each one. Species are classified as being restricted to (or characteristic of) a biome when distributed in only 1 biome or in 2 contiguous biomes (e.g., humid mountain forest (HUMF) and TEMF, or HUMF and THUF).

Table 6 shows the numbers of characteristic species, recorded by biome in each Mexican state. Highlighted in bold typeface are the 5 states contributing with the largest richness values to each biome. There are states where 1 or more of the biomes are not represented; for example, in Aguascalientes there is neither humid mountain forest (HUMF) nor seasonally dry tropical forest (SDTF) and there in neither HUMF nor temperate forest (TEMF) in Quintana Roo. However, the richness of these biomes is recorded anyway in these states because many species are found in more than 1 biome and are therefore distributed even in states where their principle biome is lacking. For example, most species reported for the HUMF in Aguascalientes are shared between HUMF and TEMF, the latter of which occurs in the state. Other species, for example Acaciella angustissima (Mill.) Britton & Rose (Fabaceae), Sambucus nigra L. (Adoxaceae), Tecoma stans (L.) Juss. ex Kunth (Bignoniaceae), or Zanthoxylum fagara (L.) Sarg. (Rutaceae), are widely distributed and have broad ecological tolerance, and are therefore found in almost every states and all 5 biomes in Mexico. On the other hand, of the 6,763 species reported by ^{Villaseñor and Ortiz (2014)} for the HUMF, 3,941 are characteristic of HUMF, found only in HUMF or in HUMF and only 1 more (adjacent) biome. Chiapas excells over all remaining states by having the largest number of HUMF characteristic species (2,399), and it is followed by Oaxaca (2,302) and Veracruz (1,875). In contrast, of the 6,852 species reported for the xeric scrubland (XESC), 4,614 are characteristic of the biome, highlighting Coahuila with the largest number of XESC characteristic species (1,764), followed by Chihuahua (1,461) and Sonora (1,411). Table 6 shows the values and percentages of characteristic species by biome in each of the states. In turn, Figure 4 shows the general distribution of characteristic species for each biome in Mexico as a whole. As in Figure 2, richness values are standardized per 10 km² by dividing the total values by the logarithm (log₁₀) of the state's surface.

Figure 4 Map of the 5 main biomes of Mexico and richness values in each of them. Values were standardized by area by dividing the total values by the logarithm (log10) of the state's surface. Biomes following the definition of ^{Villaseñor and Ortiz (2014)}. 

Discussion

Mexico's total floristic richness (23,314 species) places it fourth in the world, behind Brazil (32,000 species, ^{BFG, 2015}), China (29,000 species; ^{Qian & Ricklefs, 1999}) and Colombia (24,000 species; ^{Rangel, 2015}). ^{Villaseñor (2003)} positioned Mexico in fifth place, below South Africa, although the most recent account for that country reported the existence of 20,000 native species (^{Germishuizen, Meyer, Steenkamp, & Keith, 2006}). Also noteworthy is the high level of endemism of the flora of Mexico; although recent values reduce the percentage of endemism to 49.8%, it still occupies a privileged place among the world's countries, reporting a native flora that is very restricted to its political territories. Excluding archipelagos and other island territories that are characterized by large proportions of endemic species, Mexico is apparently surpassed only by South Africa among mainland countries in its proportion of endemic species (57.1%).

Reviewing the literature, especially protologues, has been instrumental for the inclusion of many species that have been previously forgotten or overlooked in our floristic knowledge, and continuing searches among journal publications allows the inclusion of the many recently described species occurring in Mexico whose inclusion in international databases (IPNI, The Plant List, Tropicos, etc.) is not immediate, and in fact sometimes it never happens. Consequently, search protocols should be established to increase the efficacy in incorporating these taxa into national databases, as this would allow the continuous updating of catalogs like the one presented in Appendix 2. Moreover, dealing with the diverging species concepts used by different specialists has not been trivial; the decision to accept or reject a name often reflects experience in a particular group or personal preferences.

It is interesting to contrast the species richness between northern and southern halves of the country (Fig. 2); this contrast more or less identifies the latitude of contact between the Nearctic and Neotropical realms. In turn, the decrease in richness in extreme northwest and southeast of Mexico, on the Baja California and Yucatán peninsulas, respectively, is likely the result of the peninsular effect, which causes peninsular territories to be less diverse than the continental nearby mainland due to their geographic isolation and the lack of important biomes (^{Gaston & Williams, 1996}). Also interesting is the existence of some low-diversity states embedded within a matrix of high-diversity sites; this is the case of Tlaxcala, which had richness values well below its neighboring states at all 3 taxonomic levels. It would be revealing to further explore the cause of these low richness values; discerning whether this is due to lower collecting effort, land use change, or other factors would be informative and could modify the results reported here.

The data in Table 6 point out the large number of species that are found in only 1 or 2 biomes; thus, relatively few species actually show a wide ecological tolerance. Surely, this widely distributed component includes weedy species and those that tolerate disturbance, commonly occurring in successional plant communities known as secondary vegetation. For example, some 2,640 species are currently documented as having weedy and/or ruderal behavior (Villaseñor, unpublished data), of which 1,544 (58.5%) are reported from more than 2 biomes. Similarly, of more than 1,000 species known to grow in secondary vegetation, over 700 (70%) are also reported as occurring in 3 or more biomes.

Like other mega-diverse countries, the results reported here suggest that a large percentage of Mexico's flora tends to have high values of rarity, as indicated either by the breadth of its distribution (species recorded in few states; Table 5) or its habitat specificity (species recorded in few biomes; Table 6). Therefore, it should not come as a surprise to conclude that many species remain to be encountered before we can fully assess vascular plant diversity in Mexico (^{Villaseñor, 2015}). With each new intensive exploration of a given territory (and sometimes by pure serendipity) anywhere in the country, new species are discovered and recorded, as a consequence of their narrow geographic and/or habitat distributions.

Concluding, any catalog, in the format of a flora, a taxonomic review, or a monograph, is constantly changing. The number of species reported here does not reflect the final number of different plants existing in Mexico. In fact, it may never be possible to explore any country in its entirety. This is particularly true for nations with orographies as complex as that of Mexico, so that each newly explored region will unavoidably reveal surprises. Even those regions that are better-researched are important sources of new floristic information (^{Ertter, 2000}). Moreover, the dynamism of taxonomy and systematics, revived in recent decades by the implementation of new techniques and methods of analysis (cladistics, evolutionary biogeography, molecular biology, phylogeography, etc.) leads to constant changes in circumscriptions and taxonomic rearrangements, which requires periodic updates to the floristic information. The compilation of information in databases and the new tools provided by the informatics of biodiversity (^{Bisby, 2000}; ^{Graham, Ferrier, Huettman, Moritz, & Peterson, 2004}; ^{Soberón & Peterson, 2004}), have made the curation and maintenance of taxonomic information more efficient, allowing for easier updates to catalogs like the one included here, and ultimately leading to the accumulation of more and better knowledge of our ever changing floristic richness.

Importantly, the sources of information for scholars of biodiversity and its conservation comes from 4 main sources: (1) floras, reviews and monographs; (2) catalogs and inventories; (3) databases, and (4) unpublished experience of experts. These sources are not, however, independent of one another; floras, taxonomic revisions and monographs rely on catalogs, which rely on experts to carry them out, and databases rely on all 3 of these sources for their reliable development, curation, and management. For this reason, countries whose biodiversity is high but poorly known, such as Mexico, should continue to support inventory studies, so that these other sources, including personal intellectual enrichment of specialists, continue to improve and facilitate the exploration and proper evaluation of the country's natural capital. Scientific collections and databases with a reliable curatorial level are the basis for taxonomic and systematic work; they constitute what ^{May (1990)} calls the bricks and mortar for building the biodiversity knowledge on which ecologists, biogeographers and analysts of biodiversity and conservation depend.

Completing the flora of Mexico is a huge undertaking that requires the accumulation of a vast amount of information, from data derived from field explorations, inventories, preparation of regional floras and taxonomic reviews, to the creation of databases of well-curated data. Floristic checklists, as one of these components, report the numbers of species recorded in the area of study and reflect regional species richness on which many studies assessing patterns of diversity, endemism (biogeographic), conservation, and macroecological patterns, among others, depend. They all rely on the quality of data, which must be reported with the utmost seriousness and a deep understanding of the floristic and taxonomic aspects essential to the enterprise. The identification and correct application of the names of plants is central to many areas of research in natural sciences and since the coining of the concept of biodiversity, in government agencies as well. Producing lists of threatened species, such as the Mexican NOM-054 (^{Semarnat, 2010}), the IUCN Red List of Threatened Species (^{Walter & Gillett, 1998}), and the international agreements regulating international trade of biological specimens (CITES), requires updated information on plants and their names.

The exercise carried out in this work, which aims to standardize the names used in the flora of Mexico, will hopefully serve as a useful framework for refining floristic information. Ultimately, this will provide benefits not only to the taxonomist community, but also to many other colleagues who use, require and routinely consult the kind of information presented here. Finally, paraphrasing the great botanist Arthur Cronquist, this catalog holds the work of our predecessors and will be the support of future floristic and taxonomic work.