Material and methods

Study area. The MBBR is located in the central-western part of Mexico, on the boundaries of the states of Michoacán and Estado de México, between the coordinates 19°18’-19°59’ N and 100°06’-100°22’ W. The reserve encompasses an area of 56,259 ha (Figure 1) and contains two buffer zones and three core zones (Altamirano, Cerro Pelón and Chincua-Campanario-Chivati; hereafter last core zone will be referred only as Chincua), which altogether comprise a total of 13,551 ha (^{Diario Oficial de la Federación 2000}). The reserve presents mountains and intercalated ridges between small inter-montane valleys and grasslands, with an altitudinal interval that ranges from 2,400 to 3,600 m a.s.l. (^{SEMARNAT 2001}). Physiographically, the reserve is situated in the morphotectonic province known as Trans-Mexican Volcanic Belt (^{Ferrusquía-Villafranca 1993}), which is characterized by the presence of extrusive igneous rocks of andesites, basalts, granites, rhyolites and tuffs. The main soils found in the region are Andosols and with a lower frequency there are Cambisols, Regosols and Vertisols. The regional climate is warm temperate with dry winter (Cw in the Köppen-Geiger climate classification system sensu ^{Kottek et al. 2006}), with summer rains. The total annual precipitation ranges from 700 to 1,250 mm with average annual temperatures that range from 8 to 22 °C. The vegetation types described for the reserve are: Abies forests, Pinus forests, Pinus-Quercus forests, Quercus forests, tropical montane cloud forests, scrublands and grasslands (^{Espejo-Serna et al. 1992}, ^{Soto-Núñez & Vázquez-García 1993}, ^{Mejía-Mendoza 1996}, ^{SEMARNAT 2001}, ^{Cornejo-Tenorio et al. 2003}, ^{Giménez-de Azcárate et al. 2003}).

Figure 1 Location of the Monarch Butterfly Biosphere Reserve. A) Location of the States of Michoacán and Estado de México in Mexico (grey areas), B) The rectangle indicates the area of the Reserve within the States of Michoacán and Estado de México, C) Municipalities within the Reserve, light grey corresponds to the buffer zones and darker grey to the core zone of the Reserve (modified of ^{Ramírez-Ramírez et al. 2006}); core zones: bottom (Cerro Pelón), middle (Chincua-Campanario-Chivati) and top (Cerro Altamirano). 

Data analysis. The floristic composition of the MBBR was produced from the species list of ^{Cornejo-Tenorio et al. (2003)} and a revision of material deposited in the herbarium of the Instituto de Ecología, AC, of the Centro Regional del Bajío (IEB). Additionally, from 2004 to 2006, a total of 49 field trips were conducted in order to collect other botanical material, exploring only the core zones of the reserve. All botanical samples were obtained and processed according to conventional techniques (^{Lawrence 1951}, ^{Lot & Chiang 1986}). Several families published in the series Flora del Bajío y de Regiones Adyacentes (e.g. ^{Rzedowski & Calderón de Rzedowski 2003}, ²⁰⁰⁵, ²⁰⁰⁸) and the Flora Fanerogámica del Valle de México (^{Calderón de Rzedowski & Rzedowski 2001}) were used for the identification of specimens, as well as the comparison with specimens deposited in the herbarium IEB. Specialists were required to identify some species. The first and second duplicates of the 3,684 vouchers collected are deposited in the herbaria of the Instituto de Biología de la Universidad Nacional Autónoma de México (MEXU) and in the IEB, respectively. Duplicates of these vouchers were sent to other herbaria. Some species reported by ^{Cornejo-Tenorio et al. (2003)} associated with roads and secondary vegetation were excluded from the present study: Chenopodium ambrosioides L. (Chenopodiaceae), Dalea thouinii Schrank (Fabaceae), Lepidium virginicum L. (Brassicaceae), Malva parviflora L. (Malvaceae), Polygonum lapathifolium L. (Polygonaceae), Reseda luteola L. (Resedaceae), Solanum cardiophyllum Lindl. (Solanaceae) and Verbena litoralis Kunth (Verbenaceae).

The floristic list is ordered alphabetically by family, genus, species and sub-specific categories, following the classification system proposed by ^{Christenhusz et al. (2011a)} for the ferns and related groups, ^{Christenhusz et al. (2011b)} for the gymnosperms and the APG III (2009) for the flowering plants. The species name was taken from the database ^{The Plant List (2013)}; when the binary name was indicated as “not resolved”, the name accepted by the Missouri Botanical Garden (^{Tropicos, 2014}) was included. In order to determine the conservation status or risk category of the species, we consulted the NOM-059-SEMARNAT-2010 (^{Diario Oficial de la Federación 2010}).

Floristic affinities were determined based on the current distribution of the species, were assigning to the following categories: i) Endemic (present only within Mexico), ii) North American (Mexico, reaching United States of America and Canada), iii) Central American (Mexico to Panama and the Antilles), iv) South American (Mexico to South America), and v) broad distribution (species found in the second, third and/or fourth of the above categories, or even in other continents). Different floras, taxonomic and monographic reviews, as well as the ^{Tropicos (2014)} database, were consulted in order to obtain the geographic distribution of the species (e.g. ^{Villarreal 1998}, ^{Rzedowski & Calderón de Rzedowski 2003}, ²⁰⁰⁵, ²⁰⁰⁸, ²⁰¹¹, ^{Espejo-Serna et al. 2010a}, ^b, ^{Rzedowski et al. 2011}, ^{González-Elizondo & González-Elizondo 2014}).

Beta diversity among the core zone was estimated by calculating the values of the Jaccard (I_J) index for total and endemic species. Selection of this algorithm was based on the recommendation of ^{Jost et al. (2004)}, using the formulas:

where a is the number of species shared among the samples under comparison, while b and c represent the number of species registered to sample 1 and 2, respectively.

Results

Floristic composition, growth form and vegetation types. The flora of the core zones of the reserve includes 97 families, 337 genera and 694 species, of which 20 correspond to infra-specific taxa (Appendix). The list consists mainly of the angiosperm group, of which 77 % correspond to the eudicotyledons and 14.7 % the monocotyledons, while the ferns and gymnosperms account for 7.3 and 1 %, respectively (Table 1). The family Asteraceae is the most diverse, with around 20 % of the total of genera and species documented (Table 2). The 10 families with the higher number of species contained 46.6 % of the genera and 51.3 % of the species. In terms of the genera, Salvia (Lamiaceae) is the most diverse, with 23 species. The genera with more species, mainly belong to the most diverse families (Table 2), except for Sisyrinchium (Iridaceae) and Tillandsia (Bromeliaceae).

The most important components of the MBBR flora are the endemic species (38.3 %), followed by those of broad distribution with nearly 23 % (Table 3). A total of 64 families (66 %) and 147 genera (43.6 %) contain endemic species. The family Asteraceae and the genus Salvia have the highest number of species restricted to Mexico (Table 2). When the entire flora is taken into account the North America affinities presented the lower frequency.

The herbaceous plants represent the most abundant growth form in the reserve, with 515 species, which constitutes 74.2 % of the total flora, followed by the shrubs (73 species, 10.5 %), trees (49, 7.1 %), climbers (37, 5.3 %), epiphytes (10, 1.4 %), hemiparasites (6, 0.9 %) and parasites (4, 0.6 %). Included in the climbers are those plants that present herbaceous and/or woody stems, since this characteristic is often difficult to determine because this can be age or size-dependent. The species reported as woody climbers are Archibaccharis hirtella (Asteraceae), Celastrus pringlei (Celastraceae), Clematis dioica (Ranunculaceae), Cyclanthera integrifoliola (Cucurbitaceae) as well as Solandra guttata and Solanum appendiculatum (both Solanaceae). However, C. pringlei sometimes occurs as a shrub or small tree with hanging branches. One other species that may become a climbing plant of up to 10 m in height is Salvia gesneriiflora (Lamiaceae), although in this study it is included in the shrubs, since the climbing condition was only observed in Cerro Pelón.

When we compared vegetation types that host the highest number of species (total, endemic and exclusive (species collected only within one vegetation type)), we observed that these attributes follow a similar pattern, in which the Quercus and Abies forests respectively, had the highest number of species, while those with the lowest numbers of species were the Pinus forest and the Juniperus scrubland (Table 4). The exclusive species accounted for 56.4 % of the total number, while those recorded in only two vegetation types accounted for almost 23.6 %. The remaining species were found in three to five vegetation types. It is noteworthy that no single species was present across all of the vegetation types.

Floristic similitude among the core zones. The three core zones host an almost identical number of families, but there are marked differences in terms of the number of species. The Altamirano zone presented higher numbers of species, followed by Chincua and Cerro Pelón; the last core zone had the least number of genera (Table 5). The core zones of the reserve appear to be similar in terms of the vegetation types represented (Table 5); however, the richness and composition of the plants are dissimilar. We found statistically significant differences in the number of species presented in each core zone (χ² = 216.25, d.f. = 2, P < 0.001). As a rule, all core zones showed higher numbers of endemic species in comparison with other categories of geographic distribution (Table 3); in general, the other geographic affinities categories showed the same position previously obtained for entire flora.

Of the 694 species, 460 (66.3 %) are found in one core zone only, 140 (20.2 %) in two and 94 (13.5 %) in all three zones. When species exclusive to one only one core zone are considered, Altamirano has 215 species, and Chincua and Cerro Pelón has 151 and 91, respectively. Likewise, endemic species share among three core zones is scarce (only 27 species). Therefore, beta diversity among these areas is high (0.03-0.11), regardless of whether this attribute is estimated with for entire flora or endemic taxa (Table 6).

Discussion

Floristic composition, growth form and vegetation types. When we compare our results to those of ^{Cornejo-Tenorio et al. (2003)}, the updated list of the MBBR is increased by nearly 40 % in terms of number of species. The three most important families with the highest number of species are still Asteraceae, Fabaceae and Lamiaceae, but have increased in species numbers by 30, 54 and 34 %, respectively. These families are among the ten most diverse of the flora of Mexico (^{Villaseñor 2003}), with Asteraceae in first place. This result was to be expected, since Mexico is considered the main center of diversification of this family (^{Turner & Nesom 1993}). Another aspect to consider is that the MBBR presents vegetation types in which Asteraceae usually reaches high numbers of species, especially in temperate forests (^{Turner & Nesom 1993}). The low representation of the Fabaceae in the study area is due, in part, to the fact that this family presents its greatest species diversity in tropical forests (^{Sousa and Delgado 1993}).

In terms of the genera, Salvia and Stevia remain in first and second place, with an increase of 43.5 and 33.3 % relative to the data presented by ^{Cornejo-Tenorio et al. (2003)}. ^{Rzedowski (1991a)} indicates that in the montane regions with a cool and semi-humid climate, a species rich flora has evolved in which various genera, such as Eupatorium, Muhlenbergia, Quercus, Salvia, Sedum, Senecio and Stevia, are notable. Regarding these taxa, Salvia is the second most diverse at the national level (^{Villaseñor 2004}). Recently, ^{Martínez-Gordillo et al. (2013)} estimated 307 species (nearly 76 % endemics) for this genus. In accordance with those figures, the MBBR hosts almost 7.5 % of the Mexican salvias and 36 % of 64 species listed for Michoacán (^{Cornejo-Tenorio & Ibarra-Manríquez 2011}). A similar figure exists for the genus Quercus since the reserve hosts almost 7 % of the 161 species of the genus recorded in Mexico, and 38 % of those recorded for Michoacán (^{Valencia-Á. 2004}). A notable issue is that Altamirano core zone is the only known site for the recollection of Q. greggii in Michoacán (^{Romero-Rangel et al. 2014}).

In relation to growth forms, we found the same pattern described by ^{Cornejo-Tenorio et al. (2003)}, with a marked dominance of herbaceous plants (74.8 %), followed by shrubs (9.6 %). This coincides with information on the flora of Mexico; where nearly 61 % of the components corresponds to herbaceous plants (^{Villaseñor & Ortiz 2014}).

The Mexican flora showed an important percent of endemic species, which vary from 48 to nearly 60 % (^{Rzedowski 1991b}, ^{Villaseñor & Ortiz 2014}). In this respect, the percentage (36 %) of endemic herbs plants found in the MBBR is lower. However, notable percentages of endemism are recorded within the group of epiphytes (70 %) and hemiparasites (83 %) in the reserve. In the case of the epiphytes, this percentage may be explained by the contribution of Tillandsia (Table 2) since nearly 70 % of the species of this genus recorded in Mexico are endemics (^{Espejo-Serna et al. 2004}).

The vegetation types with the highest number of species in the MBBR are the forests of Quercus, of Abies and the tropical montane cloud forest. The percentage of exclusive species found in these forests is variable, with the Quercus forest notable in this aspect since 50 % of the 194 species associated with this forest type are not found in the any other vegetation type. ^{Challenger (1998)} indicated that the floristic composition of the pine (Pinus) and oak (Quercus) forests of Mexico is highly variable from one site to another, even within the same region. This is due to the environmental heterogeneity presented by these forests and to their complex biogeographic history. With regard to the floristic composition of Abies forests, ^{Sánchez-González et al. (2005)} compared exclusivity across 12 sites on the Trans-Mexican Volcanic Belt and found that only 8 % of the species were shared with the majority of the sites, a value that should not be minimized since some of these are abundant species which define the forest structure. In the MBBR, grasslands are the vegetation type with the third highest number of exclusive species. It should be noted that these communities host interesting species, some of very restricted distribution, such as Sisyrinchium conzattii, known only in the highest mountains of Distrito Federal, Estado de México and Michoacán (^{Calderón de Rzedowski & Rzedowski 1985}). For the latter state, this species presents a conservation challenge since it has only been recorded in the grasslands of Sierra Chincua, where populations are very scarce (^{Espejo et al. 2010b}).

Floristic similitude among the core zones. The main vegetation types considering their number of species in the MBBR (Quercus, of Abies and the tropical montane cloud forests) showed the same rank position for the endemic component (Table 3), reflecting that documented at national level by ^{Rzedowski (1991b)} and ^{Villaseñor & Ortiz (2014)}. The floristic affinities in the flora of the MBBR are not easily compared, since the categories of geographic affinities often differ among biogeographic studies. In this sense, ^{Sánchez-González et al. (2005)} found that 32 % of the species of the Abies forest of the mountain Tláloc (Estado de México) are endemic to Mexico, and 31 % extend to Central America. The Abies forests in our study site has a similar pattern, as these distribution categories represent 39.2 and 21.9 %, values that are also similar to those found for the entire flora (Table 4).

The high values of beta diversity between core zones cannot be explained by their differences in area (Table 4). One factor to consider is that they share various vegetation types (fir forest, pine-oak forest, Juniperus scrubland and grassland) but differ in some, particularly in the presence of tropical montane cloud forest. Moreover, it is very important to consider the different roles of each vegetation type with a detailed appraisal of the areas they occupy within each core zone and the differences in environmental factors. This type of appraisal should be implemented in the near future. Likewise, a more profound floristic exploration of Cerro Pelón is required, because during the study period the conditions were unfavorable for this task, especially in the tropical montane cloud forest. To illustrate this fact, we point out that the micro-endemic species Perymenium ibarrarum Rzed. & Calderón, which is found in this core zone, has recently been described based only in the specimens of Ibarra-Manríquez (5,160) and Cornejo Tenorio (^{Rzedowski & Calderón de Rzedowski 2011}).

This degree of floristic exclusivity that can be observed of the core zones is really important (Table 6). Another approach to corroborate the floristic differences among core zones is the presence of 12 species placed under some category of protection in terms of conservation status or risk by the Mexican government environmental policy (NOM-059-ECOL-2010), seven of these are found in only one core zone (Cerro Pelón, Carpinus caroliniana, Populus simaroa, Rhynchostele cervantesii and Zinowiewia concinna; Chincua, Comarostaphylis discolor, Juniperus monticola and Trifolium wormskioldii). Four species are shared by Altamirano and Chincua zones (Dahlia scapigera, Gentiana spathacea, Monotropa hypopitys and Furcraea parmentieri), whit only Cupressus lusitanica has been recorded in all three core zones.

Finally, it is important to emphasize the fact that the core zones of the MBBR include more than just the Abies forests that serve as winter refuges for the Monarch butterfly; but are also important reservoirs of plant biodiversity, either for Mexican endemics or for other species under some protective category of conservation risk. Our study also demonstrates that is extremely important to consider that the core zones are complementary in terms of plant conservation and require realistic and urgent efforts in order to secure their long-term maintenance.