Zamia loddigesii is a perennial plant characterized by subterranean stems, chartaceous cataphylls, one to ten leaves per crown, each bearing prickles on the petiole and rachis, and ellipsoid to conical ovulate strobili with an acute apex (Fig. 1). This species, endemic to Mexico, occurs from southern Tamaulipas along the seaboard in Veracruz to the northern region of Oaxaca (Fig. 2). The vegetation types are oak forest, evergreen tropical forest, dry forest, as well as secondary vegetation zones (Nicolalde-Morejón et al. 2009, 2024). The region has two distinct seasons: a dry season from October to May, and a wet season from June to September. The mean annual temperature ranges from 22 to 26°C, and the mean annual precipitation varies between 1000 and 1500 mm (Vidal-Zepeda 1990a, 1990b). We monitored 17 wild populations in Veracruz and Oaxaca States. These populations were grouped into three areas for analyses: 1) a northern area that includes four populations located at the north of Veracruz State with some close to Tamaulipas State, 2) a central area that covers 11 populations in central Veracruz, and 3) a southern area with two populations found in Oaxaca State (Fig. 2; Suppl. material 1). Most populations are located in disturbed habitats such as pastures and cornfields, with the exception of the populations in the southern area and two in the central area of distribution that are relatively undisturbed (Suppl. material 1). Vegetation types were categorized according to the most recent land use and vegetation cover dataset provided by the Mexican Instituto Nacional de Geografía y Estadística (INEGI 2018).

The census was carried out from October 2022 to January 2024. All individuals in the 17 wild populations covering the distribution range of the species were tagged. To investigate the development of reproductive and vegetative structures, all structures produced by individuals were monitored monthly. Some populations were visited more than once a month when there were a large number of strobili or when the phenophases involved in pollination were observed (receptivity and open pollen). The categorization of the reproductive phenophases for both ovulate and pollen strobili followed the scheme of Martínez-Domínguez et al. (2018), whereas the leaf phenophases were developed and classified based on their timing and development as recorded here (Suppl. material 2). The shape of reproductive structure, trichome colour, and separation of the sporophylls for all strobili were recorded to describe ontogenetic stages in this species. For characterizing and describing vegetative events, the shape, colour, texture, and arrangement of the leaves were recorded. These data were used to develop a categorization of phenophases for leaves applicable to this and future studies.

Additionally, the length and diameter of the fertile portion and the peduncle (infertile portion) of ovulate and pollen strobili were measured to analyze the annual growth dynamics. To better understand the life cycle of these structures and the population dynamics of Zamia loddigesii, these data were visualized using boxplots. The analysis was performed in R v.3.0.1 (R Core Team 2013).

Circular statistics were applied to determine the reproductive and vegetative phenological patterns within and among populations, evaluate the synchrony among phenophases, as well as the concentration of the data using the mean vector (r) and the mean angle. Rayleight’s test (z) was calculated to evaluate deviations. We tested data for normality and homoscedasticity. All analyses were performed in ORIANA v.4.02 (Kovach 1994). Months of the year were converted to angles at 30° intervals from January through December (Morellato et al. 2010).

The Fournier index (FI) was applied to evaluate the intensity each reproductive phenophase expressed at population level according to a semi-quantitative scale with categories from 0 to 4. The categorical scale was defined considering the percent of reproductive structures in each phenophase, with 0 indicating individuals without strobili, 1 for 1–25%, 2 for 26–50%, 3 for 51–75%, and 4 for 76–100%. This index was obtained by adding the intensity values of each month. The sum was divided by the total number of individuals that exhibited one of the phenophases multiplied by four, then multiplied by 100 to calculate the intensity percentage for each population (Bencke and Morellato 2002; Segalla et al. 2022).

Temperature and precipitation data were obtained directly from the National Meteorological System of Mexico (SNM). We used the monthly temperature and precipitation of each population from the nearest meteorological station. In order to explore the potential correlation between these climatic variables and the phenophases at species level, the standardized normal data of the populations were averaged. Spearman correlation was used to evaluate the correlation among climatic variables, three reproductive phenophases (receptivity, open pollen, disintegration) and two vegetative phenophases (emergence and expansion) (Spearman 1904). The analysis was performed in R v.3.0.1 (R Core Team 2013).

A total of 1,874 plants were recorded for all populations. The percentage of adults per population ranged from 20 to 70%, with more than half of these being non-reproductive individuals. Within the total of adult individuals, 4 to 43% of the individuals in each population were polliniferous plants and 2.32 to 30.86% were ovuliferous plants. We observed the timing and abundance of vegetative structures in all 1,874 plants. Some plants have multiple apices capable of producing strobili simultaneously. Only 333 plants were in the reproductive phase, and these were recorded for reproductive phenological patterns and phases. We found 723 reproductive structures, of which 134 were ovulate strobili and 589 pollen strobili.

Across all populations, 266 strobili were measured and analyzed for growth dynamics. There were 87 ovulate strobili and 179 pollen strobili. The growth trends in diameter and length of pollen and ovulate strobili were markedly different (Fig. 3). Growth in pollen strobili in both length and diameter increased exponentially until the fourth month, with only minimal increase in the fifth month compared to previous months. In contrast, ovulate strobili doubled in size in the first three months, after which growth increased gradually month by month (Fig. 3). In the fifth month of development, pollen strobili showed a marked decrease in diameter (Fig. 3D).

The emergence phenophase for pollen strobili lasted four to six weeks and was characterized by abundant whitish trichomes and peduncles almost entirely covered by cataphylls (Fig. 4A). The closed pollen phenophase lasted five to six weeks and was marked by a gradual elongation of the central portion of the strobilus axis and an increase in the size of the microsporophylls which remain compacted without spaces between them. Towards the end of this phenophase, the hexagonal faces of the microsporophylls expanded and the trichomes turned reddish brown. In the open pollen phenophase, the microsporophylls were yellowish green with reddish brown trichomes, separated slightly from one another, and microsporangia opened allowing the release of pollen in a process that lasted two to three weeks (Fig. 4C). The senescence phenophase was characterized by dehydration of the pollen strobilus with a curvature of the fertile portion from the apex to the base and the trichomes turned dark brown to greyish brown (Fig. 4D).

The life span of ovulate strobili began with elongation of the fertile portion, a light brown colour and cataphylls covering the peduncle. This emergence phenophase lasted four to five weeks (Fig. 4A). The receptivity phenophase lasted two to three weeks and was characterized by separation of megasporophylls during elongation of the axis starting from its base and the presence of brown to reddish trichomes (Fig. 4C). In the late ovulate phenophase, the strobili were brown and turned greyish brown while maturing. The adaxial face of the megasporophylls was widened towards the end of week 21 to 23 and the centre of each became slightly sunken (Fig. 4D). Finally, the disintegration phenophase involved a progressive detachment of the megasporophylls from the central axis and the strobili turned dark brown with the red sarcotesta exposed. This phase lasted two weeks (Fig. 4E).

The leaf development cycle was divided into four phenophases: emergence, elongation, expansion, and maturation (Fig. 5). The emergence phase lasted four weeks. It was characterized by leaves with compact, imbricate leaflets emerging from the shoot apex. The leaves had brown to reddish brown trichomes that became light brown and withered with a few remaining on the rachis, the base of the petiole, and the margins of the leaflets. The elongation phenophase only lasted one week and was characterized by an erect petiole and rachis with a progressive decrease in trichomes that became whitish. Towards the end of this phenophase, the leaflets began to separate at the apex of the leaf. The expansion phenophase lasted four to six weeks. The leaflets separated from each other and expanded longitudinally. The leaf apex became curved, and the leaf was mostly glabrous with a papyraceous texture. The petiole elongated longitudinally and its trichomes turned whitish grey. The mature phenophase was characterized by the full elongation of the rachis and the separation of the linear to lanceolate leaflets from one another. The leaves reached their final size at approximately week 10.

For Zamia loddigesii, the reproductive phenophases of open pollen and receptivity were highly synchronous and seasonal (Fig. 6; Table 1). The open pollen phenophase started in March and lasted until July with a maximum intensity in May (Fig. 6A). The receptivity phenophase lasted five months from April to August with a maximum intensity in June (Fig. 6A; Table 2). These reproductive phenophases between pollen and ovulate strobili were synchronous during four months from April until July. The peak for open pollen and receptivity was May, with the exception of the northern area in which it was June (Suppl. material 3).

The emergence and expansion vegetative phenophases had the shortest duration (i.e. a few days). These phases were selected to evaluate patterns of leaf production. Each plant produced one new leaf flush per year, and the timing varied among the individuals within populations (Fig. 6B). The production of new leaves was recorded from February through December, with the exception of January and August (Fig. 6B). The peak for these phenophases was April and the emergence phenophase lasted five months, from March to June and December while the expansion phenophase was recorded throughout the year.

The number of reproductive individuals varied among populations. The Cerro Gordo and Pinoltepec populations (Fig. 7A) in the central area were those with the largest number of reproductive individuals producing pollen and ovulate strobili, whereas the La Cueva (central area) and San Felipe Usila (south area) populations had the lowest number of both pollen and ovulate strobili (Suppl. material 1). The Cerro Gordo and Pinoltepec populations occur in the “acahual” of dry forest and Cedrela odorata L. plantations with low forest cover. In contrast, La Cueva (central area) and San Felipe Usila (south area) populations occur under closed canopies in secondary evergreen tropical forest. Likewise, another canopy-covered population in the central area, Mozomboa, produced fewer strobili and leaves (Fig. 8B). Overall, a greater number of pollen strobili than ovulate strobili were recorded in each population because pollen plants often produce multiple pollen strobili per plant, whereas ovulate plants usually produced one and occasionally two strobili per plant.

At the population level, the reproductive phenophases involved in pollen release from polliniferous plants and pollen reception in ovuliferous plants (open pollen and receptivity, respectively) were highly synchronous among the three areas of the distribution range. However, receptivity was slightly asynchronous within the areas, with some individuals starting receptivity in April and others in July. The central area was the first to start the receptivity phenophase, and the northern area was the last (Suppl. material 3). Not all populations within the same area commenced receptivity at the same time. Also, two populations in the central area started receptivity simultaneously, with one population in the northern area (Suppl. material 3).

In the northern area, the open pollen phenophase was recorded from May to June, with May recorded as the peak; three out of four populations showed receptivity from May to August, with July recorded as the peak. In each of these three populations (Cinco de Mayo, Ojo de Agua, and Rodolfo Curtí), the open pollen phenophase lasted two months (Suppl. material 3). The largest number of strobili was recorded in the Rodolfo Curtí population during May, whereas in Santa Rosalía no individuals with pollen strobili were found. In all populations, the receptivity phenophase lasted two months. Ojo de Agua and Rodolfo Curtí were the first populations where this phenophase started and the latter recorded the highest number of pollen strobili. The Santa Rosalía population was the last to express receptivity (i.e. July to August).

In the central area, the open pollen phenophase occurred from March to July, and receptivity in April and May. The peak month for both phenophases was May (Suppl. material 3). Open pollen lasted four months (March to June) in most of the populations; however, extending into July in three populations. Of the 11 populations of the central area, eight were releasing pollen during the peak period of the open pollen phenophase. Receptivity was synchronous in all populations during May, with the exception of one population, Dos Caminos, that peaked in June and July. (Suppl. material 3).

In the southern area, the open pollen phenophase occurred from April to June with a peak in May, whereas receptivity occurred from June to July with a peak in June. In this area, a low number of ovulate and pollen strobili were recorded (Suppl. material 3). This seems to be related to population sizes, which were the smallest compared to the other areas.

Leaf production was highly asynchronous across phenophases for all three areas. In the northern area, emergence and expansion occurred continuously for five months, starting in February with a peak in June. Santa Rosalía was the population where the emergence phenophase lasted the longest (i.e. four months), and Rodolfo Curtí was the population that produced the greatest number of leaves. The expansion phenophase mainly occurred in the months of February, April, June, July, and December, with a peak in April (Fig. 6B). Overall, the production of leaves exhibited an irregular pattern. In the central area, emergence and expansion phenophases were recorded in two separate periods: 1) March to June in seven populations (Mozomboa, Vaquería, La Vega, Llano Grande, El Cañal, La Cueva, Pinoltepec, La Poza, Cerro Gordo), and 2) October to December in five populations (La Poza, Vaquería, Llano Grande, Dos Caminos, Cerro Gordo) (Suppl. material 4). Peak emergence occurred in April, whereas expansion peaked May. The number of leaves produced was variable among populations.

Temperature significantly influenced three reproductive phenological events: receptivity, open pollen, and disintegration (Table 3). Open pollen and receptivity coincided with months when the highest temperatures were recorded. The interaction between temperature and precipitation had a strong negative correlation for the disintegration phenophase (Table 3). Overall, these patterns were consistent among areas. The most precipitation occurred between July and September, when the phenophases of receptivity and open pollen had ended (Fig. 7). The open pollen phase was more consistently related with high temperature. At lower temperatures, the number of pollen strobili decreased (Fig. 7A).

In relation to leaves, the greatest leaf production occurred during the rainy season (Fig. 7B); however, emergence and expansion phenophases were not significantly correlated with either of these climatic variables (Table 3). Overall, new leaf production occurred intermittently throughout the yearly cycle in most populations.

The difference in the duration of the life cycle between ovulate and pollen reproductive structures is common in dioecious species, including Cycadales (Espírito-Santo et al. 2003; Escobedo-Sarti and Mondragón 2016; Martínez-Domínguez et al. 2018; Segalla et al. 2022). In Zamia loddigesii, pollen strobili emerged at different times among individuals of the same population; this asynchronous production ensures that pollen remains available for a longer period of time at the population level. In some populations in the northern area, a delay in the receptivity phenophase was found compared with populations in the central and southern areas (Table 2). This phenophase lasted for several months for Z. loddigesii, longer than what had been reported in the genus Ceratozamia Brongn. (Martínez-Domínguez et al. 2022); this suggests a lower risk for the survival of Z. loddigesii. Also, in some populations, no pollen strobili were recorded, but the ovulate strobili in these populations were fertilized, leading to ovule maturation. This illustrates the intra- and interpopulation dynamics that lead to successful pollination (Lazcano-Lara and Ackerman 2018; Toon et al. 2020; Martínez-Domínguez et al. 2022; Segalla et al. 2022). Pollinator insects could have migrated from the nearest populations due to their apparent mobility (Martínez-Domínguez et al. 2020).

The reproductive phenological pattern of open pollen and receptivity was seasonal and correlated with temperature variation (Fig. 7A). Changes in the timing of phenophases induced by climatic variables could affect the close relationship that plants have with their pollinating insects (Thackeray et al. 2016). In Zamia loddigesii, open pollen and receptivity lasted for several months, covering most of the dry season, ensuring that some pollen strobili provided pollen to ovulate plants. These phenophases occur during the warmest season of the year in each of the areas, although not occurring in the exact same month in every area. As for strobilus disintegration, this phenophase occurs in the driest months with lower temperatures (Table 3). Precipitation may not be relevant for this species, as seeds of Z. loddigesii can germinate shortly after their detachment from the ovulate strobili (Norstog and Nicholls 1997). This is in contrast with the pattern found in Ceratozamia, where disintegration occurs during the rainy season (Martínez-Domínguez et al. 2022). The seeds from this genus take more than a year to germinate because the embryo is not fully developed (Norstog and Nicholls 1997). Indirectly, this suggests that climatic variables could influence the timing of reproductive phenophases in the Cycadales (Davies et al. 2013).

On the other hand, in areas with low environmental disturbance and shaded understory, where the incidence of sunlight is low, a smaller proportion of ovulate and pollen strobili are produced (Clark and Clark 1987). Most populations of Zamia loddigesii have been altered by anthropogenic activities such as crop establishment, in which there is even use of herbicides. These anthropized populations recorded the highest number of ovuliferous and polliniferous strobili (Fig. 8A). More than 60% of the cycad species are classified as at risk, with habitat loss and reproductive failure as the major threats (Mankga and Yessoufou 2017); however, individuals in these anthropized populations produced a greater number of apices and more strobili. Overall, these results show that despite habitat disturbance, Z. loddigesii can still achieve pollination and ovulate strobili development.

The vegetative phenology of cycads has received far less attention and is often overlooked. Some taxonomic studies have documented leaf production in some species (e.g. Stevenson 1981; Martínez-Domínguez et al. 2017). With some exceptions, studies on the quantity and periodicity of leaf production are uncommon (Clark et al. 1992). Here, we describe development stages of the leaves for Zamia loddigesii (Fig. 5). The leaf growth accelerates in the early stages, and then slows down. Cycads invest in defence against herbivory (Schneider et al. 2002); however, until leaves complete ontogeny, these structures are susceptible to herbivory, which in some cases results in the complete loss of the leaves (Prado et al. 2014). New studies focusing on vegetative patterns and herbivory are necessary to understand the factors influencing leaf production.

It has been reported that environmental factors such as light availability affect leaf production (Clark et al. 1992). In Zamia loddigesii, leaves at emergence were recorded almost every month of the year, suggesting the absence of a seasonal pattern (Fig. 6B). This could be because most of the populations are in anthropized areas (Fig. 8; Clark and Clark 1987). In fact, some leaves emerged after accidental or intentional cuts in fields for orange or lemon cultivation. In contrast, the lowest leaf production was recorded in April and June in populations under canopy cover. This data coincides with the peak intensity of leaf emergence for all populations regardless of their degree of disturbance. Regarding climatic variables, no correlation was found (Table 3). This suggests that leaf production responds to other types of variables. The species appears to have an intrinsic biological pattern but retains the ability to adjust to different environmental situations. Partially, it could be related to insects of the genus Eumaeus Hübner, 1819 (Fig. 1), with which the species interacts closely and that are known as defoliators (Prado et al. 2014). The ability to produce new leaves shortly after losing some structures could represent an advantage allowing Z. loddigesii to survive in anthropized environments where disturbances are common, a strategy not reported for most other cycad species (Stevenson 1981).