Study area

The study was carried out in four natural protected areas (NPAs) of the Zoque Tropical Forest Biological Corridor, Chiapas, Mexico: a) The Ocote Tropical Forest Biosphere Reserve (REBISO), b) The Sumidero Canyon National Park (PNCS), c) the Zone Subject to Ecological Conservation, La Pera (ZSCELP) and d) The Villa de Allende Tropical Forest Area (ZPFVVA) (^{Semarnat-Conanp, 2001}; Fig.1).

Figure 1 Study area at the Zoque Tropical Forest Biological Corridor, Chiapas, Mexico. Design: H. Gutiérrez-Santiag). 

Phenological and environmental monitoring.- Throughout a year, monthly visits were carried out to each of the four NPAs for the monitoring of flowering and fructification, seed collection of C. guatemalensis and record the environmental variables (altitude, soil type, environmental temperature, monthly rainfall, and predominant vegetation in the monitoring site). The monitoring was conducted to accessible sites in an expanse of 1 hectare ensuring that significant populations of C. guatemanlensis are found in each NPAs. For the correct identification of the species, samples were collected that were herborized according to ^{Lot and Chiang (1990)} ; the taxonomic identification was made through the review of specialized flora bibliography (^{Stevens, Ulloa, Pool, & Olga, 2001}) and through comparisons with the collections of the Ministry of Environment and Natural History Herbarium (CHIP: 44713), the Eizi Matuda Herbarium (HEM: DB-23664) and the National Herbarium of Mexico (IBUNAM: MEXU-PVsn46367), as well as from the revision in the The Plant List ^{[TPL] (2013)} and Tropics database (^{Tropicos, 2017}). The trees of C. guatemalensis were selected based on the probabilistic factor that a phenotype has sufficient genetic basis to show a favorable reaction to different environments; the selection was of a directional type, five specimens were selected per plot in each NPAs; according to inclusion criteria such as: the straightness of the stem, the natural pruning, the conformation of the crown and branches, free of pests and diseases ^{(Centro Agronómico Tropical de Investigación y Enseñanza [Catie] 1995}, ^{Pola et al. 2003}).

The date and duration of each phenotypic phase was recorded and the percentage of presence of each phenological character was evaluated through the BBCH general scale (^{Hess et al., 1997}). This scale considers two digits, where the development cycle of the plant is divided into 10 recognizable and distinguishable phases called main growth stages numbered from zero to nine (first number); secondary growth stages (second number) were used to identify more accurately the growth progression of C. guatemalensis during the main growth stage.

The weather data of temperature and monthly pluvial precipitation were provided by the network of climatological stations of the National Water Commission and the National Institute of Research, Agriculture and Livestock of Chiapas. The altitude was obtained by a manual receiver of the global geo-positioning system (Garmin Leyend Etrex). Additionally, the predominant type of vegetal association was recorded according to the classification of ^{Rzedowski (1994)} to determine the presence of vegetable cover. Finally, a simple sampling of the soil was carried out to determine its quality through the collection of five samples per plot in each protected area, of one kilogram of soil each, with a recommended depth of 0 cm -20 cm and 20 cm - 40 cm. These depths are more practical as more uniform samples are obtained, quick to perform and cause less damage to the ground surface (^{Fournier, 1974}). The samples were taken to the Plant Germplasm Bank Laboratory of the Institute of Biological Sciences at the University of Sciences and Arts of Chiapas (UNICACH) to determine their pH, porosity, texture and organic matter through the procedure of ^{Jaramillo (2002)} . Table 1 shows the environmental variables registered for each NPAs.

Fruits and seed collection

Once the fructification of the species was present, 100 fruits were collected from each tree monitored in each NPAs. The fruits were stored in labeled paper bags to be transferred to the Plant Germplasm Bank Laboratory of the Biological Sciences Institute at the University of Sciences and Arts of Chiapas (UNICACH). A total of 500 seeds were extracted from each NPA to determing the germination and growth percentage of each seedling in the nursery.

Seed germination. Under a completely randomized experimental design, the germination test was performed, with three repetitions of 30 seeds of homogeneous measurements that were taken from the total lot collected for each NPAs (30 seeds × 3 repetitions × 4 NPAs). After two days of harvesting, the 360 seeds were sown at a depth of two cm with the micropyle down in 12 Koper block® type unicel trays for forest species (60 cm × 35 cm × 12 cm), with substrate collected from each NPAs. The experiment was conducted in a nursery with shaded mesh (70% shade), maintained temperature of 26 °C and humidity of 78%; irrigation and data collection were taken once every 3 days. The seeds were considered germinated when they presented epicotyl emergence on the substrate (> 5 mm) (Hartmann & Kester, 2001). Each tray was considered as an experimental unit, with time as an independent variable and as response variables the percentage average of final germination (GP), the average time of germination (T), the germination speed (GS), the germination index (GI) and the speed coefficient (GC) (^{González-Zertuche & Orozco-Segovia, 1996}). The observation and annotation corresponding to the growth and development of the seedlings, type of germination, primary leaves, length and thickness of the stem, number of leaves and length of the root for the identification of the species in its early stage (seedling) were made (^{López-Ríos, 2005}). The diameter of the stem and height of the seedling were determined with an Ohaus trademark 0 mm - 150 mm caliper digital calibrator with precision of 0.1 mm. The leaves were only counted according to the number of shoots. The data were taken at (15, 20, 30, 45 and 60) days, and irrigation was given once every 3 days (^{López-Ríos, 2005}).

Statistical analysis

To determine the possible relationship between the environmental variables present in each protected area and flowering and fructification of C. guatemalensis, a simple correlation analysis was carried out (^{Ríos García, 2015}). For the production of fruits, it was verified that the data had a normal distribution (mesocúrtica) with standardized kurtosis values ​​and standardized bias within the range -2 to +2 according to the normality test (Shapiro test). Having no normal distribution, the non-parametric Kruskal-Wallis test was carried out to compare the production of fruits from the four PNAs (^{Duplancic et al., 2015}). In terms of germination, descriptive statistics were first carried out to obtain the mean and standard deviation of the germination in each NPAs. Subsequently, it was verified that the data obtained had a normal distribution with values of standardized kurtosis and standardized bias within the -2 to +2 range, according to the normality test (Shapiro test); as it did not have a normal distribution, the non-parametric Kruskal-Wallis test was carried out to compare the germination among the seeds from the four NPAs, and subsequently, the Mann-Whitney U test was performed, using the Bonferroni fit for the identification of possible groups (^{Duplancic et al., 2015}). To determine which environmental variables influenced the final germination result, a generalized linear model (MLG) was carried out (^{R Core Team, 2018}); where the variables of temperature and precipitation were excluded from this analysis since these variables were the same throughout the experimental design of germination because it was carried out in the same nursery for all the seeds. Finally, the observation and annotation corresponding to the growth and development of the seedlings, germination type, primary leaves, length and thickness of the stem, number of leaves and length of the root for the identification of the species in its early stage (seedling) were carried out (^{López-Ríos 2005}). All statistical analyzes were performed using the R 3.24 software and the statistical data package Past® and Statgraphics® (^{Hammer et al. 2001}, ^{R Core Team 2018}).

Flowering and Fructification

The phenological events of flowering and fructification of C. guatemalensis occurred only once a year in each NPAs; flowering started in the month of July (25% of individuals n = 5 ± 4.82) and extended until April (25% of individuals), with December and January being the months with the highest percentage of individuals with flowers (100% of individuals n = 20 ± 8.4). Fructification began in March (25% of individuals), extending until September (25% of individuals); the months of May and June presented the highest amount of fruits (75% n = 15 ± 3.33 and 100% of the individuals n = 20 ± 3.42, respectively (Table 2, Fig. 2). According to the Kruskal-Wallis analysis, there were significant differences in fruit production between the four NPAs (K = 12.38, P = <0.006). ZSCELP (28248 ± 2578.7) was the area with the largest fruits, followed by PNCS (35460 ± 1670.6), REBISO (37313 ± 2060.2) and finally ZPFVVA (43103 ± 5730.7). In all the NPAs, a moderately strong relationship was observed between flowering and temperature, flowering and rainfall, fructification and temperature, and fructification and rainfall (Table 2).

A: Viable seed, BC: Germination type epigeous, DE: False leaves (cotyledon), FGH: appearance of 2 to 3 leaves, I: formation of flower spikes, JK: Flowering and beginning of fructification, LMN: Complete fructification and ripening, O: End of growth, perennifolia plant.

Figure 2 Main and secondary growth stages of C. guatemalensis. 

Germination

Table 3 shows the means and standard deviations of C. guatemalensis germination in each NPAs According to the Kruskal-Wallis analysis, there were statistically significant differences in the germination percentage (GP) (K = 9.31, P = <0.025), time (T) (K = 8.56, P = <0.036), germination index (GI) (K = 9.49; P = <0.023), germination speed (GS) (K = 10.38, P = <0.016) and germination coefficient (GC) (K = 8.56; P = <0.036) of C. guatemalensis seeds from the four NPAs. Both ZSCELP and REBISO had the highest percentage of final germination (98% ± 3.84% and 88% ± 11.70%, respectively), followed by PNCS (72% ± 8.38%) with moderate germination and ZPFVVA (6% 3.84; Table 3) the lowest.

Through the generalized linear model (GLM) is was observed that of all the environmental variables evaluated for the final germination percentage, soil porosity was the only significant variable (t = 2.237, P = 0.0503), with a positive relationship with the germination; the greater the porosity of the soil, the greater the percentage of final germination that was obtained (Table 4).

Growth and development of seedlings

The germination of the seeds was of the epigeal type, the cotyledons emerged from the soil due to a considerable growth of the hypocotyl (portion between the radicle and the insertion point of the cotyledons). From day three to five the seed coat, cotyledons and epicotyl of the seedling emerged from the ground, thus forming the primary root. After five days, the plumule (cotyledonary leaves) originated, from which the adventitious roots developed. From 20 days to 30 days the first leaves appeared (true leaves) and at 35 days the process of lignification of the stem, increase in thickness and length began (Fig. 2, Table 5).