Influence of Socio-Historical Events and Macroecological Patterns on the Endemic Plant Descriptions in the Iberian Peninsula.
It was not until Linnaeus' Species Plantarum in 1753 that the binomial names were consistently applied. Thus, this date was established as the starting point for the naming of plants (art. 13.1 of the ICN: Turland et al., 2018), and consequently a large share of the currently European accepted species is credited to Linnaeus (Jarvis, 1992). Because the system was established, most of the common plants in Europe were described in the 50 following years, either by Linnaeus himself or by other prolific naturalists such as Lamarck, Willdenow or Miller (e.g. Aeschimann et al., 2011).
On the whole, however, the endemic species whose geographical ranges are locally restricted have been described in later time periods. The authors who have studied the process of species discovery tend to agree that narrowly distributed and smaller size taxa are later described, not only for vertebrates (Blackburn & Gaston, 1995; DinizFilho et al., 2005) and invertebrates animals (Gaston et al., 1995a), but also for vascular plants (Bebber et al., 2007a; Cavallin et al., 2016). Likewise, the temporal pattern of known species description has also been used to estimate the number of species that remain undiscovered (Gaston et al., 1995b; Dolphin & Quicke, 2001; Ulloa et al., 2017), although methods such as interpolation from species discovery curves have high margins of error (Bebber et al., 2007b).
The protologue of a newly published name generally includes a geographical location from which it was described; this is known as locus classicus or type locality. As single data points, the type localities provide scarce information about the plant distributions and the species richness. However, at a broad scale, a spatial correlation can be expected between the pattern of aggregation of the type localities and the observed richness of narrow endemics (e.g. Peruzzi et al., 2015). Beyond that, similarly to the biodiversity data collection, the type localities are very likely biased by several socio-economic and geographic drivers, especially at finer spatial resolution (Meyer et al., 2015). Some factors including varying accessibility of areas, heterogeneous academic activity, and focus on regions with certain appeal like endemism or species richness (Boakes et al., 2010; Meyer et al, 2015; Brundu et al., 2017) condition the type localities choice.
According to the latest updated work (Buira et al., 2017) the Iberian flora -including the Balearic Islands- consists of 1328 endemic species (1825 taxa), which represents near 24% of the total native flora. The description of these plants has been uneven over the last 265 years, ranging from the first Iberian plants included in Linnaeus' Species Plantarum in the mid-eighteenth century, through the extensive works of Boissier or Willkomm in the 19th, to the specific taxonomic revisions for Flora iberica (Castroviejo, 1986-2019) in the late 20th and early 21st centuries. In fact, several Iberian narrow endemics are still being discovered and described in recent times. Although the specific contribution of some of the most prolific authors has been documented (e.g. Devesa and Viera, 2001; Gonzalez Bueno, 2010), so far any comprehensive account in terms of described taxa over time has been done for the whole Iberian endemic flora. Knowing the historical events that have boosted the discovery effort is important to determine whether the process of species description is predictable.
Buira et al. (2017) also pointed out that the Iberian plant richness is unevenly distributed, comprising the Balearic Islands, the southwestern Atlantic coast, the Pyrenees, Cantabrian and Baetic mountains as the main centers of endemism. In particular, the Baetic System, whose endemic flora is characterized by the great richness of narrow endemics and the high species turnover rate, is by far the richest region of the territory. While the distribution of endemic species richness has been studied by several authors (e.g. Sainz Ollero & Moreno Saiz 2002; Aedo et al., 2017a), little is known about the spatial pattern of the species description. The extent to which the type locality distribution reflects the real pattern of the endemic plant richness at a medium- and large-scale can be assessed by exploring the spatial relationship between both variables.
A large share of the Iberian Peninsula and the Balearic Islands are included within the Mediterranean Basin Hotspot (Myers et al., 2000). Given its high level of plant endemism, relative isolation and long botanical history, this territory is considered a suitable model to explore how the historical events and the macroecological and human presence variables influence the process of species discovery. Likewise, understanding the pattern of such variation is important to assess whether the task of description is nearly accomplished. Here, we make
the first chronological and spatial overview of the Iberian endemic flora description. Particularly our aims are (1) to provide an account of the described taxa over time and compare the temporal pattern between native and endemic taxa, (2) to draw up a brief survey of the most important historical events and prolific authors, (3) to explore how the species geographical range size and life-form (indicator of body size and seasonality) have influenced the description dates (4) examine the spatial and temporal pattern of the species plant description and its relation with the observed plant richness and the human influence on the territory.
Material and Methods
We used as a basis of this study an updated list of the Iberian and Balearic endemic taxa. The taxonomic criterion was based mainly on Flora iberica, though the list also included new taxa published after that work that are accepted in Anthos project (www. anthos.es). The endemic taxa were defined as the species and subspecies occurring exclusively in the Iberian Peninsula and the Balearic Islands, as well as those whose distributions slightly exceed the Iberian limits into the northern side of the Pyrenees (here after referred as Iberian endemics). The highly polymorphic groups in which hybridation, apomixis or reticulation phenomena are common and whose distribution ranges are generally poorly known, particularly Limonium, Alchemilla and Taraxacum, were separated for certain comparative analysis.
We compiled for each accepted taxon all the nomenclatural information, including the basionym, the describing and combining authors, and the date and place of publication. The list of the accepted taxa includes a part of non-combined names applied as they were originally described-, and another part of names that have been combined under alternative genera or at a different taxonomic rank of the basionym. In the case of combined names, the authors and dates of description are always referred to the basionyms or, in the few cases of replacement names, to the replaced synonyms (Art. 6.11 and 7.4 of the ICN: Turland et al, 2018).
The type localities were procured for each species from the text of its protologues. The geographical accuracy of the locations varied from the great precision of the GPS points to vague references to broad regions or even simply the country. The basic spatial unit for our analysis was the Universal Transverse Mercator (UTM) 50 x 50 km grid system. We therefore discarded the type localities whose geographical position could not be located, at the very least, to this geographic accuracy. By doing so, we included 1050 species, which account for 80% of the total endemic species. Most species (94%) consisted of just one type locality (sometimes more but included in a single grid cell); only 6% of species consisted of multiple type localities covering two or more grid cells.
The endemic species distribution data were obtained from Anthos and Flora-On (www.flora-on.pt) projects, which compiled about 70,000 unique records on 10 x 10 km cells. The range size of the species was measured by the total number of 10 km grid cells in which the species occurs (ranging from 1 to 1050). All the species were classified into five range size categories by dividing the total sample into five equalsized groups; these are very narrow (VN), narrow (N), medium (M), wide (W) and very wide (VW). The life-form categories were adapted from the Raunkiaer (1934) system, i.e. phanerophytes (P), chamaephytes (C), hemicryptophytes (H), therophytes (T) and geophytes (G).
Taxa Described over Time and Authors' Contribution Account
The number of total non-endemic native taxa and endemic taxa described per year from 1753 to 2017 were plotted in bar graphs and cumulative frequency polygons in order to compare the historical trends between both groups. Dates of publication were also grouped in 20 year periods for different taxa groups to make interpretation easier. The contribution of the most prolific authors and their main botanical works in terms of endemic taxa were accounted and depicted in time-line graphs. Tables including the first name initials and the dates of birth and death of all botanists cited in the text as well as the main authors of both endemic taxa and non-endemic taxa are provided in Appendix 1 (Tables 4 and 5). Maps with the type localities of the endemic species described by the most prolific authors are presented in the Appendix 2 (Fig. 7).
One-way analysis of variance (ANOVA) were conducted to explore whether significant differences existed on the mean description date for the different categories of range size and life-form.
A correlation analysis was carried out to examine the relationship among several spatial variables, i.e. the number of species described per area, observed richness and human influence. In particular, we used the number of total endemics type localities (TETL) per grid cell and two separated subsets of the narrow endemics (NETL) and widespread endemics (WETL). Two different metrics of richness were chosen. The absolute endemism (AE) is the simple count of the total endemic species observed (from distribution data) in each 50 x 50 km cell. The weighted endemism (WE) differs by taking into account the range restriction (e.g. Linder, 2001; Laffan & Crisp, 2003), and it was calculated by the sum of the inverse range size (total presence records on grid cells) of each species occurring within the 50 km grid cell. The Global Human Influence Index (Wildlife Conservation Society, 2005) was used to calculate the average degree of human influence (HII) in each 50 km grid cell. This index incorporates population density, land use and human accessibility variables, and even though it measures the current human influence we assumed that its spatial pattern has varied proportionally over the period of study. All the cells of the Iberian Peninsula (excluding the Balearic Islands) containing at least one type locality (180) were included in the analysis.
Additionally, the number of type localities (TETL) was fitted by a generalized linear model (GLM) with Poisson error distribution and log-link function using the HII and the WE variables as predictors. Finally, we tested whether the description date of the species was related with the spatial human influence and observed richness through linear regression. We performed all analyses using the R statistical software version (https://www.R-project.org) with the associate package 'car'.
Number of Iberian Taxa Described over the Time
An account of the described taxa per year from 1753 to 2017 was made for the 4534 non-endemic native taxa and for the 1825 endemic taxa. Large differences can be observed between the trends of both graphs (Fig. 1 a and b). The cumulative species number of the non-endemic native taxa follows an approximate negative exponential distribution ([R.sup.2] = 0.94), while that of the endemics is better fitted to a linear model ([R.sup.2] = 0.98). About a 40% of the non-endemic native taxa were initially described in Linnaeus' Species Plantarum in 1753 -considered the starting point-. Most of the nonendemic native species were, in fact, described during the first stage, from the latter half of the eighteenth century to the early 19th. At the end of the nineteenth century more than 90% of the non-endemic native taxa were already described.
On the contrary, only 37 Iberian endemic taxa were included in Linnaeus' work of 1753 and very few were described in the following 50 years. It was not until the beginning of the nineteenth century that the Iberian endemic species description gained momentum. Two important peaks occurred in 1838 and in 1852 and more than 15% of endemic taxa were described during this 15 year period. Likewise, these two dates were also meaningful in the description of the non-endemic native species. In the following years there was a relative steady increase and at the beginning of the twentieth century, over half of the endemics had already been described. However, the botanical activity in Iberia from 1936 to 1967 was very low and the rise of described taxa during that period was meager. In 1968 the endemics description started to increase progressively, reaching a maximum in the 1980s, but as early as the twenty-first century the activity descended again. In any event, the contribution has been significant in the last 35 years, in which 20% of the endemic taxa have been described.
In regard to the taxonomic rank, few taxa currently accepted as subspecies were described in the early years; the ratio between taxa and species is in general greater for the periods as of 1876 (Fig. 2 a). Similarly, the portion of described taxa belonging to 'apomictic' groups is much greater in later periods (from 1976 to 2017). They are mostly Limonium and Alchemilla species described at the end of the twentieth century. As with the species, the greatest number of endemic genera was described in the period between 1816 and 1855 (Fig. 2 b). Over 60% of cases, the species contained therein were initially described under other larger genera, and were later split into different monospecific or paucispecific genera. Remarkably, four endemic genera (14% of the total) have been discovered in the last period.
Main Authors and Botanical Works
A total of 365 authors have contributed to the description of the currently accepted endemic taxa. However 42% have been described only by the 12 most prolific authors (Fig. 3 a), most of which lived in the nineteenth century. Whereas Linnaeus is the author of almost half (47%) of the non-endemic native species (see Table 5 in Appendix 1), only 4.4% of the endemics are credited to him. The loci classici of the Linnean endemics were in general references to the country (Hispania or Lusitania) or at best to a broad region (e.g. Habitat in Pyrenaeis). In several cases they were wrongly indicated, like in most of the 11 Balearic endemics described by him. Thus, any type locality of the Linnean plants could be geolocated.
During this first period Linnaeus was the only relevant contributor, but in the late 18th and early nineteenth century the Spanish Cavanilles and Lagasca described several new endemic species (1.6% and 2.4% of the total respectively; Table 1), mainly from central and eastern Spain (see Fig. 7 in Appendix 2). In Portugal, the Portuguese Brotero published his Flora lusitanica (1804), while the German Hoffmannsegg and Link prepared the Flore portugaise (1809-1840). The endemics described by these three authors accounted for 3% of the total. By that time, the Genevan De Candolle prepared the third edition of the Flore franqaise (1803-1815), which included several new Pyrenean endemics -accounted in the present work as Iberian endemics and representing c. 2%.
The 20-year period from 1836 to 1855 was the most fruitful of all times, when several prolific authors coincided in time (Fig. 3 b). The Genevan Boissier was by far the most productive, who throughout his career described about 13 % (170 species, 213 taxa) of the total Iberian endemics (Table 1). Reuter was co-author of 36% of these taxa. Boissier was particularly focused on the Baetic System (south of Spain); from 1838 to 1852 he described in several works 100 Baetic endemic species (c. 40% of the total endemics of that region), 55 of which were only from Sierra Nevada. In the second half of the nineteenth century the Willkomm and Lange's Prodromus (1861-1880) had a relevant weight. In total, Willkomm described about 5% of total endemic taxa and Lange almost 3%.
By the end of the century, the number of authors increased slightly being the Spanish Pau the largest contributor. During more than 40 years he published numerous species (4.6% of the total) in several papers. Font Quer also described several endemics (c. 2%) during the first third of the twentieth century. After the works of these authors, it was not only towards the end of the twentieth century that Iberian botany was back on track again. In the last two periods (1975 to 2017), the endemics were described by a much larger number of authors (Fig. 3 a), generally specialized in certain groups, and published in periodical journals. Furthermore, in this last phase any single author described more than 20 endemic species, except those who worked with 'apomictic' groups.
About 50% percent of the types used by the top 12 authors to describe their taxa were on average collected by them. However there are great differences between authors (Table 1), whereas Linnaeus or Cosson did not collect any of their type species, others like Brotero allegedly collected all the plants he described. There were also some frequent collectors such as Bourgeau, who is mentioned in nearly 50 types (9% of the total).
Description Date for Range Size and Life-Form Categories
No statistically significant differences were found on the means description dates of the life-form categories (Fig. 4 a). In contrast, significant differences were found between all the five range-size categories (Fig. 4 b). In the same way, the description date was negatively correlated with the range size as continuous variable ([R.sup.2] = -0.52, p < 2e16), indicating that recently described species are in general geographically restricted. Near 70% of the total widespread species (W and VW) were described during the first century. The few endemics described in recent times within the categories W and VW generally came from the split of widely distributed species due to thorough taxonomic or phylogenetic studies or in few cases, from the assignment of new names to those badly applied. For instance, the name Prolongoa pectinata (L.) Boiss. was used for a widespread plant endemic to central Iberia, but Lopez Gonzalez & Jarvis (1984) demonstrated that its type material was another plant and consequently proposed a new name, P. hispanica G. Lopez & C.E. Jarvis.
Life-form Range size
Spatial Distribution of Type Localities and Spatial Average Description Date
The type localities were highly aggregated since 50% were gathered on just 26 grid cells (10% of the total grid cells). The highest number of loci classici was found in Sierra Nevada (Fig. 5, the circle exceeds the size of the cell), where nearly 90 endemic species were described, followed by western Majorca with 33 type localities recorded. All the rest of Iberian grid cells contained less than 30 type localities.
As might be expected, there was a high correlation (Table 2) between the two metrics of observed richness (AE and WE). However, the WE showed higher correlations with the number of type localities per area (TETL and NETL) and it was independent of the HII. Remarkably, the human influence (HII) was negatively correlated with the absolute endemism (AE) but positively with the number of species described per area (TETL and particularly the subset of WETL). The proportion of deviance accounted by the generalized linear model was 67% (Table 3). Although less relevant than richness (measured by WE), the HII had a positive significant contribution to the model, indicating that many endemic species were described from populated and accessible areas. Results of models fitted for the subsets are not presented here, but the HII had higher influence on the description of the widespread endemics (WETL) while it had a weak but significant effect on the description of the narrow endemics (NETL).
In regard to the spatial pattern of description date, any relation was found with the human influence (HII) and richness (AE and WE) variables, either with the average description date per grid cell or at the single level of species description date. Only the number of species described over the last 50 years was positively correlated with the richness (WE) ([R.sup.2] = 0.63, p < 5e-12), indicating that the species recently described were generally discovered in narrow endemic-rich areas. In any event, the map of the average description date is presented for informative purposes (Fig. 6). The grid cells containing at least 9 type localities showed differences of up to 100 years on the average description date, being the most recent mainly located in the minor Balearic Islands and the Prebaetic Mountains, and the earliest in Coimbra (Central Portugal) and Val d'Aran (Pyrenees).
Linnaeus only described 4.4% of the total Iberian endemic species, and most of them are widespread plants. As noted by Lopez Gonzalez (1990), some of these were exclusively bibliographic references taken from some previous authors, particularly Clusius. In other cases, they were described based on herbarium specimens (e.g. Burser's herbarium) or cultivated plants (Botanical Gardens of Hartekamp and
Uppsala) coming from material harvested during the voyages of Tournefort and Jussieu (conducted in 1687 and 1716 respectively) and labeled with the adjectives lusitanica or hispanica. Another important source of Iberian material was that provided by his botanical correspondents and the disciples who travelled to the Iberian Peninsula, Lofling being the most relevant.
The Iberian botany at Linnaeus' time was barely developed. In addition, the few Spanish botanists still followed Tournefort's system, as the creator of the Royal Botanic Garden of Madrid in 1755, Joseph Quer (Gomez Ortega, 1784; Aedo et al., 2017b). Cavanilles was the first director of such institution (from 1801) with a modern scientific training (Lopez Pinero, 2004). He and his apprentice and successor Lagasca were the first Spanish authors to describe numerous Iberian endemic taxa (c. 4% of the total), mainly from the outskirts of Madrid and eastern Iberia. However, their contribution was rather modest, partly because the former was more focused on describing plants from the New World and the latter was frustrated by the prevailing political situation (Gonzalez Bueno & Rodriguez, 1996). By that time in Portugal, the botanical knowledge was considerably higher than in Spain since there were two precursory national floras (Brotero's and Hoffmannsegg and Link's). Similarly, the Pyrenean endemic flora was reasonably well known in the early nineteenth century due to the botanical activity carried out in France, particularly that of Lapeyrouse and De Candolle. This explains the earlier average description date of some of the grid cells included in this area.
The scientific activity remained very low in Spain for much of the nineteenth century largely due to the social and political upheavals of the time. However, the taxonomy enjoyed a golden age in Europe (Enderby, 2010) and many European botanists were attracted by the highly diverse Iberian flora. There is no doubt that the most important was Boissier, whose explorations in the Baetic Mountains are the main cause of the first and most important peak (period from 1836 to 1855) in the Iberian endemic flora descriptions (20% of the total). Boissier travelled for the first time to southern Spain in 1837 and only one year after published a large share of the new plants collected there (including c. 70 endemic taxa). Additionally, that same year De Candolle included in his Prodromus (vol. 7) several Baetic Compositae collected by Boissier. As a result, 90 endemic taxa (c. 5% of the total) were published in 1838, beating the discovered taxa per year record (Fig. 1). Boissier and Reuter made two more extensive field trips to Spain and over the 25 subsequent years published numerous species from the Cantabrian Mountains, the Central System and especially the south of Iberia (see Fig. 7 in Appendix 2).
The Baetic Mountains are considered a hotspot within the Mediterranean Basin (Medail & Quezel, 1997) and particularly Sierra Nevada, where the highest number of endemics occurs (c. 250 species; Buira et al. 2017). Boissier, encouraged by his mentor De Candolle (Gonzalez Bueno, 2010), wisely chose the Baetic region when it broadly remained botanically unexplored; only Thalacker (Lagasca & Rodriguez, 1802), Bory and perhaps Alstromer (Lopez Gonzalez & Jarvis, 1984) had collected a few plants in Sierra Nevada before. The great richness of a relatively uncharted territory, together with the excellent botanical training of Boissier, his relations with local botanists and his frequent collaboration with Reuter, enabled the Genevan to get the most of his explorations and become the most prolific author of the Iberian endemic flora (13% of the total). Likewise, Boissier made a great contribution describing non-endemic species (c. 3% of the total; see Table 5 in Appendix 1), mostly Ibero-African plants that only occur in southeastern Iberia within Europe.
In the second half of the nineteenth century Willkomm and Lange explored a large share of the territory (see Fig. 7 in Appendix 2) and prepared the Prodromus florae Hispanicae, the first extensive flora of mainland Spain (incl. 5092 species) which also incorporated chorological notes for the Balearic Islands and Portugal. This was a great breakthrough that allowed the botanists at the time to identify and detect undescribed species (Aedo et al. 2017a). After that (1880), over half of the endemic species were already known, and by that time the number of authors slightly increased, including some relevant as Rouy, Sampaio or Pau. This latter author conducted a major work in collecting and describing cryptic and rare species with the aim to fill the gaps of knowledge of the Iberian endemic flora. Font Quer followed Pau's task but regrettably his botanical projects were truncated as a result of the Spanish Civil War (Ibanez Cortina, 2003). From 1936 to 1967 botanical activity was very low due to the economic and political situation and the rise of described taxa during that period was meager.
At the beginning of 1960s Flora Europaea (Tutin et al., 1964-1980) started to be published. Similar to the Willkomm and Lange's prodromus, this updated work allowed the easy identification of the plants but also the comparison with those that grew in neighboring countries. On the other hand the shortcomings inherent in all the syntheses showed up, which led to the publication of numerous new cryptic species. This was also a motivation to start the Flora iberica project in the early 1980s. Thus, in the late twentieth century period (from 1975 to 1995) there was the second major peak in the Iberian endemic flora description (15% of the total). During this last stage, the endemic taxa were described by a much larger number of authors generally specialized in certain groups and derived in many cases from large taxonomic revisions.
The description process of the Iberian endemic flora has been spatially and temporally uneven and mainly described by a few authors. About a half of the species was described by 16 authors (less than 5% of the total authors) and, in turn, almost half of the type specimens were on average self-collected by them. This shows that prolific authors were in general extensive plant collectors and it agrees with the disproportionate pattern observed by Bebber et al. (2012) in which more than half type specimens of a large dataset were collected by a reduced number of collectors (< 2%). On the contrary, a large share of the new Iberian taxa emerged from monographic studies was based on older herbarium specimens, which is also in line with the pattern observed by Bebber et al. (2010) and Cavallin et al. (2016).
Spatial Pattern of the Type Localities and Average Description Date
As found in previous studies (e.g. Gaston et al., 1995a; Blackburn & Gaston, 1995; Diniz-Filho et al., 2005; Cavallin et al., 2016), the description date for Iberian endemics is negatively correlated with the geographical range size. All these authors also found that smaller size taxa tend to be later described (both for animals and plants); however no significant differences were found on the average year of description between any of the life-form categories for the Iberian endemics. This result may suggest that detectability determined by the height and seasonality of the plants is not a constraining factor on plant discovery in the Mediterranean region, especially compared to tropical and subtropical ecosystems.
As in other papers (Peruzzi et al., 2015; Brundu et al., 2017) we detected a high correlation between the number of type localities and the observed richness, particularly with the weighted endemism (WE) metric, which takes into account the range restriction of the species. Spatial range size rarity accurately predicts the distribution of type localities since the likelihood of describing a species within an area containing several narrow endemics is certainly high. Our results show that hotspots like Sierra Nevada have focused the attraction of botanists, and plants described there are not only range-restricted, but also widespread endemics like Hypericum caprifolium Boiss. or Satureja intricate Lange. On the other hand, we also found a positive significant contribution of the human influence variable, indicating that many endemic species were described in populated and accessible areas. The surroundings of Madrid, Lisboa or Coimbra (location of the first Portuguese Botanical Garden) were the loci classici of a large share of endemics (mostly widespread), some of them (e.g. Cynara tournefortii Boiss. & Reut. and Onobrychis matritensis Boiss. & Reut.) are even currently extinct in their type localities. Furthermore, the endemics restricted to a particular mountain range like the Pyrenees were generally first described from the most accessible part of it.
It is common that regressing known species richness on variables representing human presence result in apparent positive relationships (Ficetola et al., 2014; Meyer et al., 2015). However, unlike the type localities distribution, we found no positive correlation between the observed richness and the human influence, indicating that the available species distribution data for the Iberian territory has no ostensible bias and may be representative of the real richness at the spatial resolution used here. By contrast, the type localities distribution in Iberia was significantly biased by human influence, so caution is advised when using this variable to assess the richness of a poor data region -even at large scales- or as a tool for conservation planning and resource management (e.g. Brundu et al., 2017).
Unlike other previous related studies in which the average description date was spatially correlated with human population and biodiversity knowledge (e.g. Diniz-Filho et al., 2005), we found that neither human influence nor the observed richness were related to the date of description. As already stated, the description process in the Iberian territory has been spatially and temporally uneven, influenced in many cases by historical events. Hence, while some parts of the Portuguese territory and the Pyrenees were comparatively well explored by the early nineteenth century, others like the Prebaetic Mountains were not deeply explored until much later.
Species Accumulation and Completion
Large differences have been observed between the cumulative frequency of the non-endemic native taxa and the endemic taxa. Whereas the cumulative number of species for the first group follows a typical species discovery curve (Fisher et al., 1943) with a clear tendency to stabilize, that of the endemics is better adjusted to a linear function (Fig. 1). However, it is also true that the curve in the last 25 years trends to slow down, suggesting that new species are becoming more and more difficult to find and the species description is reaching completion. Our study reveals that changes in discovery effort have been arbitrary and unpredictable and generally governed by historical events and botanical trends. Thus, as shown in previous works (Bebber et al., 2007b), the discovery curves are not reliable to estimate the total number of undescribed species or indicate the final approach to completeness in the case of the endemic Iberian flora.
A large share of the later described taxa consisted of cryptic taxa, i.e. distinct related taxa that were initially classified under one species name (Bickford et al., 2007). Thus, it is important to consider that the high number of discovered taxa in the late twentieth century (second peak) could be partly attributable to a taxonomic inflation process (i.e. change in the species concept rather than new discoveries; Isaac et al. 2004). The implementation of analytical taxonomic treatments make the species continue to be described (Bebber et al. 2007b). The taxonomic criterion used in the present work is based mostly on Flora iberica (1983-2017), so a large share of taxa described by the authors of that time is accepted. It is particularly remarkable the high number of Limonium and Alchemilla microspecies recognized at species level in Flora iberica.
Nonetheless, if the asexual and taxonomically problematic species are excluded, the portion of endemic species recently described is still much higher than in other rich European areas, such as the Alps (Aeschimann et al., 2011). More than 200 endemic species ('apomictic' excluded) have been described from 1975 to present and many of them were not related or similar to any other. Remarkably, five rare species discovered in this last period (three within the last three years) could not be assigned to any known genus and were described as new endemic genera (i.e., Gadoria Giiemes & Mota, Rivasmartinezia Fern. Prieto & Cires, Pseudomisopates Guemes and Gyrocarium Valdes). All this, suggests that there certainly remain several undiscovered species in the Iberian territory, as well as it highlights the importance of field work even in a well-explored territory like this. As in other studies (Schatz, 2002; Cavallin et al. (2016) we have found that the number of recently discovered species is spatially correlated with the richness of narrow endemics. In this sense, the mountains where numerous and later described narrow endemics occur, such as some parts of the Baetic or Cantabrian mountains, are particularly likely to harbor undescribed rare species.
Acknowledgements We would like to thank all the authors of the taxonomic treatments of Flora iberica, used to compile the list of the Iberian and Balearic endemic vascular flora. We thank to M. Porto and X. Font to provide us the data from Flora-on and BDBC respectively. This work has been supported and funded by the Spanish Government through the Flora iberica project (CGL2014-52787-C31-P).
Aedo, C. Buira, A., Medina, L. & Fernandez-Albert, M. 2017a. The Iberian vascular flora: Richness, endemicity and distribution patterns. Pp. 101-130. In: J. Loidi (cd.), The vegetation of the Iberian Peninsula, plant and vegetation 12. Springer Netherlands.
Aedo, C., M. Fernandez-Albert, P. Barbera. A. Buira, A. Quintanar, L. Medina & R. Morales. 2017b. A botanical survey of Joseph Quer's Flora espafiola. Willdcnowia 47: 243-258.
Aeschimann, D., N. Rasolofo & J.-P. Thcurillat. 2011. Analyse de la flore des Alpes. 1: historique et biodiversite. Candollea 66: 27-55.
Bebber, D. P., M. A. Carinc, G. Davidse, D. J. Harris, E. M. Haston. M. G. Penn. S. Cafferty, J. R. I. Wood, R. W. Scotland 2012 . Big hitting collectors make massive and disproportionate contribution to the discovery of plant species. Proceedings of the Royal Society, B. Biological Scicnccs 279: 2269-2274.
Bebber, D. P.. M. A. Carine, J. R. I. Wood, A. H. Wortley, D. J. Harris, G. T. Prance, G. Davidse, J. Paige, T. D. Pennington, N. K. B. Robson. R. W. Scotland 2010. Herbaria are a major frontier for species discovery. Proceedings of the National Academy of Sciences 107: 22169-22171.
Bebber, D. P., S. A. Harris, K. J. Gaston & R. W. Scotland. 2007a. Ethnobotany and the first printed records of British flowering plants. Global Ecology and Biogeography 16: 103-108.
Bebber, D. P., F. H. C. Marriott. K. J. Gaston, S. A. Harris & R. W. Scotland. 2007b. Predicting unknown species numbers using discovery curves. Proceedings of the Royal Society. B, Biological Sciences 274: 1651-1658.
Bickford D. D. J. Lohman, N. S. Sodhi. P. K. L. Ng. R. Meier. K.. Winker, K. K. Ingram & I. Das (2007). Cryptic species as a window on diversity and conservation. Trends in Ecology & Evolution. 22 (3): 148-155.
Blackburn. T. M. & K. J. Gaston. 1995. What determines the probability of discovering a species--A study of south-American oscine passerine birds. Journal of Biogeography 22: 7-14.
Boakes, E. H., P. J. K. McGowan, R.A. Fuller, D. Chang-qing, N.E. Clark, K. O'Connor, ct al. 2010. Distorted views of biodiversity: Spatial and temporal bias in species occurrence data. PLOS biology. 8, el000385 (2010).
Brundu, G., L. Peruzzi, G. Domina, F. Bartolucci, G. Galasso, S. Pcccenini, F. M. Raimondo, ct al. 2017. At the intersection of cultural and natural heritage: Distribution and conservation of the type localities of Italian endemic vascular plants. Biological Conservation 214: 109-118.
Buira, A., C. Aedo & L. Medina. 2017. Spatial patterns of the Iberian and Balearic endemic vascular flora. Biodiversity and Conservation 26: 479-508.
Castroviejo, S. (ed.), 1986-2019. Flora iberica, vols. 1-18, 21. Real Jardin Botanico-CSIC, Madrid [detailed list of volume editors and genera authors at www.floraibcrica.org].
Cavallin, E. K... C. B. Munhoz, S. A. Harris. D. Villarroel, C. E. Proenga. 2016. Influence of biological and social-historical variables on the time taken to describe an angiosperm. American Journal of Botany 103(11): 2000-2012.
Devesa, J. A. & M. C. Viera. 2001. Viajes de un botanico sajon por la Peninsula Iberica. Heinrich Moritz Willkomm (1821-1895). Universidad dc Extremadura.
Diniz-Filho, J. A. F., R. P. Bastos, T. F. L. V. B. Rangel. L. M. Bini, P. Carvalho & R. J. Silva. 2005. Macroecological correlates and spatial patterns of anuran description dates in the Brazilian Cerrado. Global Ecology and Biogeography 14: 469-477.
Dolphin, H. & D. L. J. Quicke. 2001. Estimating the global species richness of an incompletely described taxon: An example using parasitoid wasps (Hymenoptera: Braconidac). Biological Journal of the Linnean Society 73: 279-286.
Enderby, J. 2010. Imperial nature. Joseph Hooker and the practices of Victorian science. University of Chicago Press. Chicago.
Ficetola, G., M. Cagnetta, E. Padoa-Schioppa. A. Quas, E. Razzetti, R. Sindaco & A. Bonardi. 2014. Sampling bias inverts ecogcographical relationships in island reptiles. Global Ecology & Biogeography 23(11): 1303-1313.
Fisher, R. A., A.S. Corbet & C. B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology 12: 4258.
Gaston, K. J., T. M. Blackburn & N. Loder. 1995a. Which species are described first?: The case of north-American butterflies. Biodiversity and Conservation 4: 119-127.
Gaston, K. J., M. J. Scoble & A. Crook. 1995b. Patterns in species description: A case study using the Geomctridae (Lepidoptera). Biological Journal of the Linnean Society 55: 225-237.
Gomez Ortega. C. 1784. Elogio historico de Don Joseph Qucr. Impresor J. Ibarra. Madrid.
Gonzalez Bueno, A. 2010. El viaje botanico de Pierre Edmond Boissier por el sur de Espaiia (1837). Acta botanica malacitana 35: 5-21.
Gonzalez Bueno, A. & R. Rodriguez. 1996. En tomo a la 'Flora espanola': dos proyectos fallidos en los aflos centrales del siglo XEX. Anales Jardin Botanico de Madrid 54: 622-626.
Ibanez Cortina, N. 2003. Two unpublished documents of Pius font i Quer on the project of Flora Hispanica. Collectanea Botanica (Barcelona) 26: 163-180.
Isaac, N. J. B.. J. Mallet & G. M. Mace. 2004. Taxonomic inflation: Its influence on macroecology and conservation. Trends in Ecology and Evolution 19: 464-469.
Jarvis, C. 1992. The Linnaean plant name Typification project. Botanical Journal of the Linnean Society 109: 503-513.
Laffan, S. W., M. Crisp. 2003. Assessing endemism at multiple spatial scales, with an example from the Australian vascular flora. Journal of biogeography 30: 511-520.
Lagasca, M. & J. Rodriguez. 1802. Description de algunas plantas que colecto D. Guillermo Thalacker en Sierra Nevada. Anales de Ciencias Naturales 5: 263-288.
Linder, H. P. 2001. On areas of endemism, with an example from the African Restionaceae. Systematic Biology 50(6): 892-912.
Lopez Gonzalez, G. 1990. La obra botanica de Lofling en Espaiia. Pp. 33-49. In: F. Pelayo (ed.), Pehr Loefling y la expedition al Orinoco (1754-1761). Colcecion Encuentros. Serie Catalogos. Madrid.
Lopez Gonzalez, G. & C. Jarvis. 1984. De Linnaci Plantis Hispanicis Novitates Nonnullae. Anales del Jardin Botanico de Madrid 40: 341-344.
Lopez Pinero, J. M. 2004. La obra botanica de Cavanilles. Pp. 11-146. In: Segundo centenario de la muerte de un gran botanico. Sociedad Economica de Amigos del Pals, Valencia.
Medail F, Quezel P (1997) Hot-spots analysis for conservation of plant biodiversity in the Mediterranean basin. Annals of the Missouri Botanical Garden 84: 112-127.
Meyer, C., H. Kreft, R. Guralnick & W. Jetz. 2015. Global priorities for an effective information basis of biodiversity distributions. Nature Communications 6: 8221.
Myers, N., R. A. Mittermeicr, C. G. Mittermeicr, G. A. B. da Fonscca & J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853-858.
Peruzzi, L., G. Domina, F. Bartolucci, G. Galasso, S. Peccenini, F. M. Raimondo, A. Albano, et al. 2015. An inventory of the names of vascular plants endemic to Italy, their loci classici and types. Phytotaxa 196: 1217.
Raunkiaer, C. 1934. The life forms of plants and statistical plant geography. Oxford University Press, London.
Sainz Ollero, H., J. C. Moreno Saiz. 2002. Flora vascular endemica espanola. Pp. 175-195. In: F. D. Pineda de J. M. Miguel & M. Casado (eds.), La Diversidad Biologica de Espafia. Pearson Educacion, S.A., Madrid.
Schatz, G. E. 2002. Taxonomy and herbaria in service of plant conservation: Lessons from Madagascar's endemic families. Annals of the Missouri Botanical Garden 89: 145-152.
Turland, N. J., J. H. Wicrsema, F. R. Barrie, W. Greuter, D. L. Hawksworth, P. S. Hcrcndcen, S. Knapp, et al. (eds.) 2018. International code of nomenclature for algae, fungi, and plants (Shenzhen code) adopted by the nineteenth international botanical congress Shenzhen, China, July 2017. Rcgnum Vegetabilc 159. Glashiitten: Koeltz Botanical Books.
Tutin, T. G., V. H. Hcywood, N. A. Burges, D. H. Valentine, S. M. Walters & D. A. Webb (eds.). 1964-1980. Flora Europaca, vols. 1-5. Cambridge University Press, Cambridge.
Ulloa, C., P. Acevcdo-Rodriguez, S. Beck. M. J. Belgrano. R. Bemal, P. E. Berry, L. Brako ct al. 2017. An integrated assessment of the vascular plant species of the Americas. Science 358: 1614-1617.
Wildlife Conservation Society--WCS, and Center for International Earth Science Information Network CIESIN--Columbia University. 2005. Last of the wild project. Version 2, 2005 (LWP-2): Global Human Influence Index (HII) Datasct (Geographic). Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC).
Appendix 1 Table 4 Dates of birth and death of all botanists and plant collectors cited in the text in alphabetical order Botanist Birth-Death Alstromer, C. 1736-1794 Boissier, P.E. 1810-1885 Bory. J.-B. 1778-1846 Bourgeau. E. 1813-1877 Brotero, F.A. 1744-1828 Burser, J. 1583-1639 Cavanilles. A.J. 1745-1804 Clusius, C. 1526-1609 Cosson, E. S.-Ch. 1819-1889 De Candolle, A.P. 1778-1841 Font Quer, P. 1888-1964 Hoffmannsegg, J.C. 1888-1964 Jussieu, A. 1686-1758 Lagasca, M. 1776-1839 Lamarck, J.-B. 1744-1829 Lange, J.M.C. 1818-1898 Lapeyrouse, P.P. 1744-1818 Link. J.H.F. 1767-1851 Linnaeus, C. 1707-1778 Lofling. P. 1729-1756 Miller, P. 1691-1771 Pau. C. 1857-1937 Quer, J. 1695-1764 Reuter, G.F. 1805-1872 Rouy, G. 1851-1924 Sampaio, G.A. 1865-1937 Tournfort. J.P. 1656-1708 Webb. P.B. 1793-1854 Willdenow, C.L. 1765-1812 Willkomm, H.M 1821-1895 Table 5 Main authors of the non-endemic native Iberian taxa on the left and main authors of the endemic taxa on the right. The names in parentheses are usual coauthors. Numbers in italics are solely 'apomictic' taxa Authors (non-endemic) Species (taxa) Linnaeus, C. 1962 (1971) Boissier, P.E. 114(140) De Candolle, A.P. 92(111) Dcsfontaines. R.L. 81 (86) Lamarck. J.-B. 69 (76) Willdenow, C.L. 51 (59) Villars, D. 42 (46) Miller, P. 40 (42) Allioni, C. 38 (40) Cavanilles. A.J. 35 (40) Poiret, J.L.M. 35 (37) Brotero, F.A. 33 (36) Gouan, A. 33 (37) Hudson, W. 30 (30) Gussone, G. 29 (34) Pourret, P. A. 29 (29) Jacquin, N.J. 28 (30) Lagasca. M. 25 (29) Cosson, E.S.-Ch. 24 (28) Jordan, A. 23 (35) Lapcyrouse, P.P. 23 (25) Smith, J.E. 22 (26) Willkomm, H.M. 21 (35) Lange, J.M.C. 21 (24) Boissier, P.E. (Reuter, G.F.) 170(213) Pau, C. 61 (87) Linnaeus. C. 58 (76) Willkomm. H.M. 54 (89) Erben, M. 44 (44) Lange, J.M.C. 40 (52) De Candolle. A.P. 39 (42) Lagasca, M. 31 (35) Frohner, S.E. 25 (25) Font Quer, P. 23 (36) Brotero, F.A. 23 (26) Cavanilles, A.J. 20 (29) Hoffmannsegg, J.C. (Link, J.H.F.) 18 (24) Cosson. E.S.-Ch. 18 (22) Talavera, S. 16 (22) Sennen. Fr. 15 (26) Rothmaler, W.H.P. 14 (17) Lapcyrouse, P.P. 12 (13) Rouy, G. 11 (20) Sampaio, G.A. 9 (19) Lamarck, J.-B. 9 (14) Webb, P.B. 9 (12) Coiney. A.H. 9 (10) Lopez Gonzalez, G. 8 (11)
Caption: Fig. 7 Type localities of the endemic species described by the most prolific authors and main mountain ranges and other geographic locations cited in the text. Authors are represented by symbols and ages by colors
Caption: Fig. 1 Number of taxa described per year from 1753 to 2017 (red bars), cumulative percentage (blue line) and trend line (dashed line), a) Iberian non-endemic native taxa (4534 taxa referred), the first column (year 1753) represents nearly 40% of the total and juts out from the graph; the best-fitting curve is a negative exponential distribution ([R.sup.2] = 0.94). b) Iberian endemic taxa (1825 taxa referred); the best-fitting line is a linear model ([R.sup.2] = 0.98)
Caption: Fig. 2 Number of a) endemic species and endemic taxa (specics and subspecies) and b) endemic genera described for 20 year periods from 1753 to 2017: total: 1328 species, 1825 taxa and 29 genera referred. The shaded part in the upper graph (a) is the portion corresponding to 'apomictic' groups
Caption: Fig. 3 a) Percentage of the total endemics described per 20-year periods in pale blue bars, portion described by the 12 most prolific authors in dark blue bars (authors of only 'apomictic' species not included), and number of total authors per period (red points), b) Timeline showing the active period of the top 12 authors and their main works (abbreviated names in Table 1). The top 12 authors account for 42% of the total endemics
Caption: Fig. 4 a) Boxplot of description dates for life-forms categories (phanerophytcs (P), chamaephytcs (C), hemicryptophytcs (H), therophytes (T) and geophytes (G)). The width of the box is proportional to the sample size. ANOVA test: p = 0.87, p = 0.48. b) Boxplot of description dates for range size categories (very narrow (VN), narrow (N), medium (M), wide (W) and very wide (VW)). ANOVA test: p = 104, p/2c-16; Tukey multiple comparisons were all significant (p < 0.05)
Caption: Fig. 5 Number of type localities per grid cell (black circles: total endemics; red inner circles: narrow endemics) and weighted endemism (WE) calculated by the sum of inverse range sizes (color scale). 1130 type localities are included (80% of total endemic species). The circles exceed the size of the cell in Sierra Nevada (SE Iberia)
(Caption: Fig. 6 Average description date of endemic species in the grid cells containing [greater than or equal to]9 type localities (color scale) and number of endemic species described in the last 50 years (circles). 'Apomictic' species are not included
Antoni Buira (1,2,3) * Carlos Aedo (1)
(1) Real Jardin Botanico-CSIC, Plaza de Murillo 2, 28014 Madrid, Spain
(2) Escuela Intemacional de Doctorado dc la Universidad Rcy Juan Carlos, Calle Tulipan s/n, 28933 Mostoles, Madrid. Spain
(3) Author for Correspondence; c-mail: email@example.com
Table 1 Endemic species and taxa described by the 12 most prolific authors (usual coauthors in parentheses), portion of narrow endemics, percentage of the types self-collected and list of the main works (order numbers used in the Fig. 2; abbreviation follows IPN1 (www.ipni.org)). The main works do not include all the taxa described by the authors Author Species Narrow Types (taxa) (%) (%) Linnaeus 58 (76) 5 0 Cavanilles 20 (29) 5 75 Brotero 23 (26) 4 100 Lagasca 31 (35) 13 75 Hoffmannsegg 18 (24) 16 100 (Link) De Candolle 39 (42) 19 25 Boissier (Reuter) 170 (213) 34 70 Cosson 18 (22) 28 0 Willkomm 54 (89) 39 45 Lange 40 (52) 27 46 Pau 61 (87) 43 41 Font Quer 23 (36) 56 60 Author Main works Linnaeus 1. L., Sp. Pl. (1753) Cavanilles 2. Cav., Icon. (1791-1801) Brotero 3. Brot., Fl. Lusit. (1804) Lagasca 4. Lag., Elench. Pl. (1816) Hoffmannsegg 5. Hoffmanns. & Link. Fl. Portug. 1809-1840) (Link) De Candolle 6. Lam. & DC., Fl. Franc, ed. 3 (1805-1815) 7. DC., Prodr. (1824-1839) Boissier (Reuter) 8. Boiss., Elench. PL Nov. (1838) 9. Boiss., Voy. Bot. Espagne (1839-1845) 10. Boiss. & Rout., Diagn. Pl. Nov. Hisp. (1842) 11. Boiss. & Reut., Pugill. Pl. Afr. Bor. Hispan. (1852) 12. Boiss., Diagn. Pl. Orient, scr. 2 (1854-1859) Cosson 13. Coss., Notes Pl. Crit. (1849-1852) Willkomm 14. Willk. & Lange, Prodr. Fl. Hispan. (1861-1880) Lange 14. Willk. & Lange, Prodr. Fl. Hispan. (1861-1880) Pau 15. Pau. Not. Bot. Fl. Espan. (1887-1895) Font Quer Table 2 Correlation matrix of the spatial variables: human influence index (HII); weighted endemism (WE); absolute endemism (AE); total endemics type localities (TETL); narrow endemics type localities (NETL); widespread endemics type localities (WETL). Significance in bold number; level of significance * p < .05 ** p <. 01 *** p <-001 HII WE AE TETL NETL WETL HII 1 WE -0.02 1 AE -0.24 *** 0.76 *** 1 TETL 0.14 * 0.67 *** 0.55 *** 1 NETL 0.09 0.66 *** 0.5 *** 0.89 *** 1 WETL 0.23 *** 0.38 *** 0.45 *** 0,79 *** 0.48 *** 1 Table 3 Results of the generalized linear model (GLM) fitted to the total number of type localities (TETL). Adjusted [D.sup.2] value = 0.67. Level of significance * p < .05 ** p < .01 *** p < .001 Estimate Std. Error z value Pr(>[absolute value of z]) (Intercept) -0.05772 0.09441 -0.611 0.541 Weighted endemism 0.68728 0.02566 26.783 <2e-16 *** (WE) Human influence 0.19631 0.02724 7.207 5.71e-13 *** (HII)
|Printer friendly Cite/link Email Feedback|
|Author:||Buira, Antoni; Aedo, Carlos|
|Publication:||The Botanical Review|
|Date:||Dec 1, 2019|
|Previous Article:||Biogeographical Review of Asteraceae in the Espinhaco Mountain Range, Brazil.|
|Next Article:||Edaphic Endemism in the Amazon: Vascular Plants of the canga of Carajas, Brazil.|