Geographic range of the species

Between December 2007 and June 2011, field excursions were made in 22 municipalities of the Sierra Madre Oriental (SMO) in Mexico (covering the states of Hidalgo, Veracruz and Querétaro) in order to collect plant material and record information on the species' geographic range. Collections are supported by collection permits: SGPA/DGVS/05612/09 and SGPA/DGVS/10163/10 by SEMARNAT. A total of 59 herbarium vouchers (Appendix 1) was deposited at HGOM (Universidad Autónoma del Estado de Hidalgo) and XAL (Instituto de Ecología, A.C.) and determined by A.P. Vovides, authority on Mexican cycads.

At each locality we contacted the local authority, explained the project, and obtained at least oral consent to proceed (permission was not issued in writing). At five localities, local authorities accompanied us in the fieldwork.

Our study required special permission because we worked with a protected species. The specific location of our study area is not published because of the risk of poaching, but can only be provided to the Journal privately.

Collections in Querétaro and Veracruz were infertile or had very immature reproductive structures and were determined as Ceratozamia affinis fuscoviridis. Although further taxonomic study of these populations could increase the species' range, we decided to limit the present population studies to Hidalgo as the populations appear to be restricted to this state.

Study Plots

With the aid of native guides, we visited all the 29 accessible populations found throughout the species' range. Inaccessible populations on extremely steep canyon faces, which could be accessed only with rappel equipment, were not sampled. Six populations of Ceratozamia fuscoviridis were selected within its range in Hidalgo. We selected the populations at extremes in latitude, longitude and elevation in order to capture the broadest variation, enabling us to evaluate the species' conservation status. In each area we selected the largest, best-conserved populations and avoided isolated individual plants. Sampling took place from 2008 to 2010 (Xochicoatlán in June 2008, Molango in July 2008, Juárez Hidalgo in April 2009, Tlanchinol in December 2009, Chapulhuacán in April 2010, and San Bartolo Tutotepec in October 2010, Fig. 1).

Fig. 1.

Municipalities in which the demographic study of Ceratozamia fuscoviridis took place. Places: a= Chapulhuacán, b= Tlanchinol, c=Xochicoatlán, d=Juárez Hidalgo, e=Molango, f= San Bartolo Tutotepec. Climate diagrams correspond to the Meteorological Stations (MS) closest to the populations sampled. Climatic data taken from [47].

A 400 m² (20 × 20 m) plot was measured out for each population, and then subdivided into four 10 × 10 m sub-quadrats with ropes and with wooden stakes at each corner. GPS coordinates, slope, altitude, and percentage tree cover were recorded in all sub-quadrats in every plot. Vegetation type, visible disturbance in the plots and surrounding area (garbage, fences, wood extraction, exotic plants, plantations, and/or roads) were also noted. Local guides provided information on current and historical land uses in the area (timber extraction, fires, agriculture, grazing, etc.). Records from the closest meteorological station were used to obtain climatic data.

Population size and density

Population size and density were calculated per plot; population size was the sum of individuals per plot, while density was the number of individuals per m², recording each individual found in the six plots. Additionally, the following data were recorded per individual: number of leaves, number of leaflets found on the longest leaf, length of longest leaf, and sex whenever there was a cone or cone remnants. The correlations among number of leaves, number of leaflets found on the longest leaf, and leaf length were used as an indicator of plant size to classify the individuals (see below).

Population structure

Size distribution in cycads has used diverse criteria such as number of leaflets [13, 26], trunk height classes [16], leaf number per individual [27], plant height [7], or leaf length and stem diameter [15]. In order to register differences within and between populations we used the number of leaflets on the longest leaf as a criterion (Appendix 2) to classify individuals in eight size categories: Seedling, Infant 1, Infant 2, Juvenile 1, Juvenile 2, Adult 1, Adult 2 and Adult 3 (Table 1). In general, breaks between categories were set by both number of leaflets and biological criteria to define smaller individuals and adults. We considered it more informative to divide smaller individuals into seedling and Infant 1 rather than pooling data in only one category. Seedlings were defined as plants with the seed still attached, while infants had no seed attached and had ≤ 29 leaflets on the longest leaf. Juvenile had among 30 to 49 leaflets. A plant was classified as adult if it had ≥ 70 leaflets on the longest leaf, based on our field observations that individuals with 69 or fewer leaflets are never reproductive.

Table 1.

Size categories used to describe the population structure of Ceratozamia fuscoviridis. X: present -: absent.

Life Table

We constructed a static life table [28, 29] using the total number of individuals per class for the six populations in order to assess mortality and survival rates for a population. Resulting data are treated as if a cohort was followed through time (i.e., the number of organisms decreases in every class). Even though this is static information, it can be indicative of population change through time, particularly in long-lived organisms.

Data analysis

Data from the field were analyzed with basic statistics to determine frequency and proportion of individuals for each size category. Population size, population density, population size structure, and spatial distribution were estimated for each plot. ANOVA and correlations between population size (total and adults) and elevation, tree cover, slope, annual precipitation and number of disturbance causes were explored.

The population structure among plots was compared using X² statistics. An a posteriori test (adjusted residual analysis described by Haberman [30]) to identify these differences was also carried out; the Haberman test enabled us to identify the categories with excess (marked with +) or missing (marked with -) individuals. The number of individuals in each category was used for X² and Haberman tests. Because the number of individuals in the adult category was often less than five (a limitation for Haberman test), we decided to merge the three categories of adults. Additionally, skew and kurtosis analyses were applied to compare the shape of the size distribution curve.

Finally, a static life table according to Begon et al. [29] and Krebs [28] was calculated using the sum of individuals in the six populations and estimate of the total population structure. I1 and I2 individuals were merged into one category for the sake of the analysis.

Geographical range

Twenty-nine populations of Ceratozamia fuscoviridis were found in twelve municipalities in the state of Hidalgo, occurring in Cloud Forests, Oak Forest, and Montane Rainforests, on pronounced slopes (34° to 44°), between 900 and 1,880 masl. All showed some level of human disturbance (Table 2).

Table 2.

General characterization of the study plots. MR: Montane Rainforest. CF: Cloud Forest. Data on rainfall, temperature and climate from [4748–49] and type of soil from [50].

Population size and density

Population size varied among populations: a 400 m² area had between 143 to 378 individuals, and the most dense population was three times that of the least dense. Populations at San Bartolo and Tlanchinol were the densest, with 0.945 and 0.938 individuals/m² respectively; Chapulhuacán was the least dense, with the remaining populations having intermediate densities (Table 3). There was a negative relationship between population size per plot and slope (F_{1, 4} = 39.8, P<0.001); the equation is Y = - 28.8 ^* slope + 1142.045.

Table 3.

Density and abundance of C. fuscoviridis individuals by size category in the sampled plots (400 m²). Categories: SL= Seedlings, I1= infant 1, I2= infant 2, J1= juvenile 1, J2= juvenile 2, A1= adult 1, A2= adult 2 and A3= adult 3.

Population structure

Population structure was statistically different among populations (χ² = 480.7, P < 0.05, d.f. = 25, Fig. 2); seedlings were proportionally more abundant in Tlanchinol, Juárez, and Xochicoatlán, and scarcer in Chapulhuacán, San Bartolo, and Molango (Fig. 2); and adults were proportionally more abundant in Tlanchinol and Molango and less abundant in Chapulhuacán, San Bartolo, and Juárez (Fig. 2).

Fig. 2.

Population structure (log 10 of the number of individuals) of Ceratozamia fuscoviridis, in the six 400 m² plots. Categories: S= seedling, I1= infant one, I2= infant two, J1= juvenile one, J2= juvenile two, A1= adult one, A2= adult two and A3= adult three.

The six populations studied presented a left-skewed frequency distribution. The degree of skew and kurtosis differed among populations. Ranked by decreasing degree of skew and kurtosis are: Xochicoatlán, Tlanchinol, Juárez, Molango, San Bartolo, Chapulhuacán (Table 4). The size structure showed two groups: 1) similar to an inverted J - Juárez, Xochicoatlán, Tlanchinol and Molango (Fig. 2). The populations at these sites had abundant seedlings with numbers decreasing in subsequent categories; the distribution was leptokurtic and displayed a skew around 2 units (Table 4); and 2) San Bartolo and Chapulhuacán displayed high numbers in infants 2 and juvenile 1 but fewer adults, suggesting a Deevey-I curve [31] (Fig. 2); they exhibited a platykurtic distribution and no skew (Table 4).

Table 4.

Kurtosis and skewness values for the six populations of Ceratozamia fuscoviridis studied.

In general, seedlings dominated. Infants and juveniles were less frequent and adults even more scarce (Fig. 2). No samples of A2 and/or A3 categories were found in San Bartolo and Chapulhuacán (Fig. 2). Seedlings represented 26% to 65% of the populations, although in some cases they were not the most abundant category; juveniles dominated in San Bartolo (50%), and infants in Chapulhuacán (47%). There were a large percentage of adults in three populations (23%, 18% and 12.5% in Molango, Tlanchinol, and Xochicoatlán respectively), while in the remaining populations there were fewer adults (1% in Chapulhuacán, 2% in San Bartolo and 7% in Juárez). The average number of fertile individuals sampled was 1.66 ± 0.49 in all populations.

Life Table

The life expectancy for the six populations of Ceratozamia fuscoviridis, from the sum of individuals observed per category, is indicated by l_x, the surviving proportion, and has a value of 0.036 in A3. This shows that only 3% of individuals passed on to the oldest category. The death rate, q_x, fluctuated between 0.138 in J1 to 0.555 in A1, and was higher in seedlings and adults (Table 5).

Table 5.

Life table for Ceratozamia fuscoviridis calculated from the sum of the six populations studied at Hidalgo.

Geographic range

Ceratozamia fuscoviridis was found to have a larger geographic range than just the Molango area, as previously thought [19]. The new populations found within Hidalgo possibly extend into the neighboring states of Veracruz and Querétaro, and further explorations may confirm this. However, this species still has a restricted overall geographical distribution, as it is known only from small populations in the cloud forests, oak, and montane rainforests of Hidalgo, all these being threatened, patchy ecosystems that are rapidly disappearing due to human activities.

Population size and density

The six populations varied in population size, apparently related to slope, while the population structure was not related to environmental or disturbance factors. The status of the present populations of Ceratozamia fuscoviridis, a long-lived species, is probably the result of historic multiple events such as deforestation and fragmentation. The life cycle of Ceratozamia typically extends about 30 years [32], although individual specimens can live for centuries.

The high density of C. fuscoviridis populations (0.358 ind/m² to 0.945 ind/m²) is unusual in comparison to other Mexican cycads. For example, Ceratozamia mexicana has densities of 0.28 individuals/m² [33] to 0.78 ind./m² in well-preserved sites, and up to 0.26 ind./m² in altered ones [34]; C. matudae has 0.10 ind./m² to 0.33 ind./m² [27]; Dioon edule or chamal has from 0.21 ind./m² to 0.44 ind./m² [35]; D. purpusii has 0.0037 ind./m² [36]; D. caputoi De Luca, Sabato & Vázq.Torres has 0.002 ind./m² and D. merolae De Luca, Sabato & Vázq.Torres 0.01 ind./m² [37]. Cycads are generally capable of producing a large amount of seedlings [38]. Such recruiting indicates viable populations even in very small areas, and suggests that reproduction is probably not a limiting factor under moderate habitat disturbance. Nevertheless, life cycles and population dynamics must be taken into account in making management decisions, even under good recruitment. Being very long-lived, cycads should be considered analogous to long-lived trees [17].

Although populations of Ceratozamia fuscoviridis have high densities, they occur only within a vegetation mosaic as small fragmented patches of approximately 0.1 ha. They were never seen in populations of a large area (>10 ha). We believe that the current geographic range could reflect a time when this species belonged to larger continuous populations; gene flow studies in the future may confirm or refute this.

Differences in population densities could be related to the conservation status of the habitat, as observed for C. mirandae [16] and Zamia melanorrhachis D.W. Stev. [26], where there were higher densities in well-preserved forests than in disturbed ones. However, the evaluation of the relationship between population characteristics and disturbance (particularly disturbance index, not only number of disturbances) needs to be measured.

Population structure

The six populations studied resulted in two groups. The first, in Juarez, Molango, Tlanchinol and Xochicatlán, was leptokurtic. These four populations had an inverted J static population structure corresponding to a Deevey-III [31] or Bongers et al. [39] type I distribution, common for shade-tolerant plants that maintain a more or less constant recruiting rate but higher mortality in the earlier life stages. This structure is the one most often seen in cycads such as Zamia debilis L.f. [40], Ceratozamia matudae Lundell [27], C. mirandae Vovides, Pérez-Farr. & Iglesias [16], Dioon edule [7], D. purpusii Rose [36] and D. spinulosum Dyer ex Eichler [41]; it was observed in four of the six populations of C. fuscoviridis studied. In contrast, the second group, the Chapulhuacán and San Bartolo populations showed a Deevey-I or Bongers-III distribution pattern. In part this is due to the absence of A2 and A3 individuals, possibly caused by disturbance or illegal extraction of the larger individuals, although direct evidence for this is lacking.

Population structure is extremely dynamic and very sensitive to environmental changes, so distribution Bongers-I can easily change to a Bongers-II if recruitment rates drop or are interrupted, or changed to a type III if there are other obstacles to regeneration, such as severe drought or fire [7], [42].

Life table

The analysis of six populations of Ceratozamia fuscoviridis as a whole showed that the proportion of surviving plants, (l_x), greatly decreases in the infants category, whilst the decrease is almost constant in the remaining categories. Therefore, l_x (log) has a tendency towards Deevey type II. This is a predictable result because it represents the pooling of the six populations, some of them with a Deevey I feature and others exhibiting a Deevey III. In general, the remarkable change occurred in the transition from seedlings to infants, while the other changes were of minor magnitude. While these data represent a static life table, they provide some information about population dynamics. In species with long life cycles, the population structure reflects multiple events occurring over long time periods. Therefore, this life table analysis gives only limited population dynamics, needing studies over several years to calculate elasticity matrix and lambda values for a better resolution of events.

Conservation status

We recommend an update of the species IUCN Red List category from “critically endangered” (CR) to “endangered” (EN). The extent of occurrence of Ceratozamia fuscoviridis is approximately 2,075 Km² and the area of occupancy is about 500 Km² (Fig. 3). According to IUCN criteria B [47] the species can be considered as “endangered” (EN).

Fig. 3.

Quantification of “extent of occurrence” (polygon) and “area of occupancy” (shaded squares), according to IUCN terminology [51] for the species Ceratozamia fuscoviridis. Each square represents 25 Km².