Current and potential distribution

Taxus globosa is a native species of Mexico and of northern Central America (^{Zamudio, 1992}; Spjut, 2007). Its distribution area covers from 25°23’ N in the state of Nuevo León, Mexico, to 14°19’ N in the department of Chalatenango, in El Salvador (^{Contreras-Medina et al., 2011}) (^{Shemluck et al., 2003}). In Mexico, its distribution is sporadic along the Eastern Sierra Madre in Nuevo León, Tamaulipas, Querétaro, Hidalgo, Puebla and Veracruz (^{Contreras and Luna, 2001}); San Luis Potosí (^{García and Castillo, 2000}); in Sierra Juárez, Oaxaca (^{Zamudio, 1992}), and in the Highlands of Chiapas (^{Contreras and Luna, 2001}).

It is also found in the temperate forests of Guatemala, located in the Sierra de las Minas Biosphere Reserve, in the Sierra of Cuchumatanes, and in the Volcanic Chain (^{Standley and Steyermark, 1958}; ^{Parker, 2008}); in Southern El Salvador, on the top of the Pital mountain (^{Moldenke and Moldenke, 1984}), and in Western Honduras, in the Santa Bárbara, Celaque and Tilarán national parks, the last of which is considered as the southernmost boundary of its distribution area (^{Zamudio, 1992}; ^{Contreras-Medina et al., 2010},²⁰¹¹; ^{López and García, 2015}) (Figure 1).

Taken from ^{López and García (2015)}.

Figure 1 Geographic distribution of Taxus globosa Schltdl. 

Taxus globosa requires very particular microclimate conditions for its establishment, including specific levels of light, water, temperature and humidity determined by the microtopography of the site (^{García and Castillo, 2000}; ^{Contreras and Luna, 2001}). In this regard, ^{Sánchez (2012)} determined that in two localities of Veracruz, the populations of Taxus were at sites with average temperatures of 14.5 and 15.4 °C, with relative humidity above 80 % and with levels of light of 2.5 to 3.2 mmol m^-2 s^-1; furthermore, García and Castillo recorded their presence in ravines with north-northeast exposures.

In particular, Mexican populations are geographically restricted and do not form large stands (^{López and García, 2015}); seemingly, their distribution is partly influenced by historical factors of how conifers were introduced to Mexico and Central America by the northern part of the hemisphere in the second half of the Tertiary period (^{Mirov, 1967}; ^{Hils, 1993}), and therefore, for ^{Rzedowski (2006)} it is one of the species with a discontinuous distribution that establish a floristic relationship with the east of North America (the United States of America and Canada) and it is part of the deciduous forest in the United States of America (Taxus brevifolia Nutt.) and of the montane cloud forest of Mexico and Central America (Taxus globosa) (^{Zavala, 2001}) in the process of its migration toward the south, as has been suggested for other taxa and vegetal associations (^{Pearson and Dawson, 2003}).

There are studies that have expanded the knowledge of the geographic distribution and the potential of the species, through the estimation of the ecological niche based on the use of algorithms such as Maxent and Bioclim (^{Contreras-Medina et al., 2010}, ²⁰¹¹; ^{García-Aranda et al., 2012a}, ^2012b; ^{López and García, 2015}). On this particular, ^{Contreras-Medina et al. (2010)} generated potential distribution maps using three distribution scenarios per change in land use, at different times and with Maxent-generated models of the remaining potential distribution of the species in federal natural protected areas (NPA) and Prioritary Terrestrial Regions (PTR) for conservation. In relation to this work, ^{Contreras-Medina et al. (2011)} indicate that the species has been registered in only seven NPAs of all Mesoamerica, four of which belong to Mexico, with a 19.67 % cover (Table 1). In addition, six points of occurrence of T. globosa were located at less than 5 km of distance from the boundaries of a NPA, although reportedly the species hardly occurs in two of them due to the tropical and dry climate prevalent in both: Valley of Tehuacán-Cuicatlán and the Metztitlán Ravine (Table 2). ^{López and García (2015)} located 75 towns distributed in nine states (Table 3).

The potential distribution of T. globosa in Mexico suggests the existence of 35 259 km², without considering the impact caused by the change of land use (^{Contreras-Medina et al., 2010}). ^{García-Aranda et al. (2012b)} determined that the species has a very limited potential distribution area, of 9 650 km² (Bioclim) to 10 385 km² (Bioclim+Topography being the most significant model), which amount to 0.49-0.52 % of the national territory.

The potential distribution models show a major reduction of the area due to the change of land use (^{Contreras-Medina et al., 2010}); furthermore, the variables slope and precipitation of the driest month are the ones that exert the greatest influence (^{García-Aranda et al., 2012a)}. According to ^{Contreras-Medina et al. (2010)}, in 1976 T. globosa covered a surface area of 26 620 km² (a 24.5 % reduction of the original vegetation without human intervention; for the year 1996 they estimated 6 847 km² (i.e. an 80.58 % reduction, and for the year 2000 they estimated 6 297 km² (an 82.14 % reduction).

For the year 2000, ^{Contreras-Medina et al. (2010)} estimated the potential distribution area of the species within the NPAs in 744 km², which amounted to 11.8 % of the potential surface area distributed through 10 NPAs along the Eastern Sierra Madre and the Sierra Madre of Oaxaca (the Monterrey Peaks, Sierra Gorda of Guanajuato, Sierra Gorda, Los Mármoles, the Metztitlán Gully, the Necaxa River Hydrographic Basin, Cofre de Perote, Orizaba Peak, the White River Canyon (Cañón de Río Blanco), and the Valley of Tehuacán-Cuicatlán). Notably, the potential surface area located in Chiapas does not coincide with any federal NPA. The potential surface area of T. globosa located in one of the Priority Terrestrial Regions (PTR) was 3 052 km², the equivalent of 48.46 % of the potential area for the year 2000, and it coincides with 15 PTR along its distribution area (^{Contreras-Medina et al., 2010}).

A potential habitat with less probabilities of occurring, but which meets the environmental requirements of the species, has also been identified in the states of Querétaro, Oaxaca, Guanajuato, Chiapas, San Luis Potosí, Coahuila, State of Mexico, Tlaxcala, Morelos, Jalisco and Colima (^{Contreras-Medina et al., 2011}). The presence of T. globosa in the State of Mexico was recorded by ^{Standley (1926)} and, although there are no herbarium specimens to support this information, the model of the ecological niche predicts its existence in this state (^{Contreras-Medina et al., 2010}).

Vegetal associations and dendrometry

Taxus globosa grows in the middle and lower canopy, beneath the shade of other, taller trees, in environments with a high relative humidity, near intermittent or permanent water runoffs (^{Soto et al., 2011}); in north-northeast exposures (^{García and Castillo, 2000}; ^{García-Aranda et al., 2012a)}, with varied slopes; for example, on the San Isidro plateaus, in Río Verde, San Luis Potosí (García and Castillo, 2000), the slope is 40 %, and at the El Chico National Park in Hidalgo, it ranges between 10 and 60 % (^{Zavala, 2001}). The species exhibits a strong preference for growing at the bottom of humid, shaded ravines, due to rain precipitations or to frequent mists; it grows on acidic soils with a good drainage and rich in organic matter (^{Cortés et al., 2000}; ^{Charco, 2007}).

In pine-oak forests, the species coexists with Pinus pseudostrobus Lindl., P. ayacahuite Ehrenb., P. patula Schltdl. et Cham., Arbutus xalapensis Kunth, Quercus crassifolia Bonpl., Q. germana Schltdl. et Cham., Q. rysophylla Weath., Q. laurina Bonpl., Tilia mexicana Schlecht., Meliosma dentata Urban., Abies religiosa (Kunth) Schltdl. et Cham., and A. vejarii subsp. mexicana (Martínez) Farjon (^{García and Castillo, 2000}; ^{López and García, 2015}); in the montane cloud forest, it is associated with Ostrya virginiana (Mill.) K. Koch, Liquidambar styraciflua L., Cornus disciflora DC. (^{Luna et al., 1994}), Fraxinus sp., Magnolia tamaulipana A. Vázquez, Ternstroemia sylvatica Schltdl. et Cham. (^{García-Aranda et al., 2012}a), Persea americana Mill., and Quercus affinis Scheidw (^{García and Castillo, 2000}) (Table 4).

Few studies document the structure, composition, richness and diversity of the vegetation of the T. globosa stands, which in general are described as mere field observations, and, therefore, the age structure, growth, development, and regeneration mechanisms of the Mexican populations are unknown. The populations of the Southernmost sector of the Northeastern region of Mexico have a larger basal area, a greater height and a smaller herbaceous cover than the northern populations, which are located at more disturbed sites, and, likewise, the southern sites are located at higher altitudes, with a greater slope than that prevalent in the north. Thus, their structure is influenced by the environmental heterogeneity, the characteristics of the localities, and the degree of disturbance of the populations (^{García-Aranda et al., 2012a)}.

T. globosa stands usually have a limited density of individuals and surface area; its dasometric characteristics are very variable between populations (Table 5). In more isolated areas, far from human settlements and with lower degree of direct (furtive felling) or indirect (bovine and goat cattle management) anthropic impact, there is a higher density of individuals. Those areas that are fenced and protected from the cattle are not very densely populated but exhibit a higher ratio of juvenile to adult individuals (^{Zavala et al., 2001}). On the other hand, areas with a steep relief, pronounced slopes and high stoniness have a great density of seedlings, indicating stability and durability of the population, as suggested by ^{Iszculo et al. (2005)}, for the European yew.

Conservation status and reproduction

The species is rare in many of the localities where its occurrence has been recorded (^{Contreras et al., 2001}). Since little is known of the distribution, density, structure, composition and wealth of its populations, it has been classified as Subject to Special Protection (Pr) by the Mexican Official Norm NOM-059-SEMARNAT-2010 (^{Semarnat, 2010}), which means that it may be threatened by factors that exert a negative influence on its viability; therefore, it has been determined that there is a need to favor its recovery and conservation, or the recovery and conservation of the populations of the species associated to it. At the international level, T. globosa is included in the red list of threatened species of the International Union for Conservation of Nature (IUCN) as an endangered species because its area of occupancy is less than 2 000 km², probably due to past and current exploitation, as well as to the deforestation of the montane cloud forest and of other ecosystems where it grows; these factors led to the reduction of the population size, which may continue (^{González-Espinosa et al., 2011}; ^{Thomas, 2013}).

According to ^{López and García (2015)}, out of 29 assessed T. globosa populations, 15 had a good conservation status, based on their acceptable natural regeneration; two had an inadequate conservation status, due to the low density of specimens, and 12 exhibited totally unfavorable conditions and were at risk of disappearing, due to the change of land use, deforestation, and the fragmentation of their habitat.

The isolation between the populations leads to assume the existence of a limited gene flow; besides, the exploitation of this and other taxa of the upper tree canopy involves a high risk for the conservation of the species (^{Soto et al., 2011}). Therefore, it is necessary to know the biology and forms of reproduction of the natural populations in order to ensure their conservation. According to ^{Zavala (2001)}, very precise, local information has been generated about their reproductive characteristics of the populations of El Chico, Hidalgo, and about the forms of propagation, the number of male and female specimens, the proportion between juvenile and adult trees, and the aspects of their natural regeneration (^{Zavala, 2002}).

^{Zavala et al. (2001)} observed that regeneration seems to be based on seed production, although at certain sites there is a significant production of new shoots from the stem base up (^{López, 2008}) ―a characteristic that may be exploited for its ex situ propagation and conservation. Nevertheless, the seeds are the most important structures for ensuring the permanence and genetic diversity of the populations; but, unfortunately, their production is limited, and their germination is very difficult (^{Ramírez-Sánchez et al., 2011}).

Reproduction of T. globosa from seeds has been unsuccessful (^{Zavala-Chávez, 2001}; ^{Nicholson and Munn, 2003}) due to underdevelopment of the embryo, which is indicative of morphological latency (^{Ramírez-Sánchez et al., 2011}). According to ^{Ramírez-Sánchez (2018)}, their latency has various causes (including physiological ones) which render germination even more difficult; for this reason, only vegetative methods have brought success (^{Muñoz et al., 2009}).

^{Ramírez-Sánchez et al. (2011)} determined that the morphological variables and the physical characteristics of the T. globosa seeds exhibit significant differences between geographical regions, as the seeds of northern Mexico are larger than those of central Mexico; such differences were ascribed to the fragmentation of the habitat, to the small size of the populations and to the genetic drift, as the populations of central Mexico consist of 10 to 200 individuals, while in the northern region, they are estimated to comprise 1 000 to 5 000 individuals.

A low production of seeds has also been observed, which suggests that the species follows an inter-annual regeneration pattern (^{López and García, 2015}); for this purpose, the population is required to include a minimum number of adults, with an adequate proportion between the sexes, in order to achieve an effective pollination and minimum rates of seed formation. At the El Chico National Park, out of 251 individuals, 29.9 % were adults; of these, 44.0 % were female, and 56 % were male (^{Zavala, 2002}). However, ^{Soto et al. (2011)} determined that the male/female ratio is close to 50:50 in all the populations. Only 6 to 15 seeds per ripe tree are produced, as a result of the proportion between mature and juvenile trees. For example, of 150 individuals observed, only 32 % female specimens were reproductively mature, while more than half (68 %) were juvenile, with a mean estimated age of 10.6 years (± 8.0), the oldest trees being 19 years old, and the youngest, 3 years (^{Zavala et al., 2001}).

Apparently, few juvenile trees have been incorporated into the population in the last few years, which may indicate reduction of the population or of the seed production; but, the number of individuals that make up the whole population is unknown, as are their ages, their annual seed production, the number of seeds that eventually germinate, and the survival rate of the seedlings (^{Zavala et al., 2001}).