INTRODUCTION

Variation in biotic and abiotic components, referred to as environmental heterogeneity, is a key evolutionary force in shaping phenotypic (e.g., the set of expressed characteristics), and phenological (e.g., temporal cycles of growth and reproduction) traits among plant populations (Baythavong and Stanton 2010). Heterogeneous environments are often epicenters for speciation processes, as the varying conditions act synergistically on population selection and adaptive processes (Fine 2015; Stiling 2002), thus catalyzing incipient speciation (Coyne and Orr 2004). Exploring phenological and phenotypic trends in populations subject to diverging environmental pressures may offer insights into life history processes and contribute with predictive information regarding adaptational changes at multiple spatial scales (Jägerbrand et al. 2005; Mallet et al. 2009; Pereira et al. 2013).

Bryophytes, small, non-vascular plants with ephemeral life history strategies, are second only to angiosperms in global terrestrial plant richness, keystone contributors of ecosystem services in most of the earth’s major biomes, and models for autecological studies (Vanderpoorten and Goffinet 2009; Rousk & Villareal 2024). Their generally broad distributions and wide ecological amplitudes make them ideal for addressing effects of environmental pressures on selection processes (Pócs, 1982; Gradstein 1992; Gradstein et al. 2001; Bowker et al. 2000; Zartman et al. 2015). For example, pressures exerted by contrasting conditions demonstrably promote phenotypic and phenological variations among individuals (Stark 2002), including fertilization rates and sex ratios (Maciel-Silva and Válio 2011; Maciel-Silva et al. 2012; 2013; Glime2021).

The Amazon Rainforest is an important biodiversity hotspot, and a primary source of lineages for other neotropical regions (Antonelli et al. 2018). It is a heterogeneous biome, with an array of vegetation types, such as “Terra-firme” forests comprised of ombrophilous, non-flooded forests which vary in structure and floristic composition over large geographic scales (Pires and Prance 1985; ter Steege et al. 2006). Within this matrix of terra-firme forests are patches of drier, oligotrophic white-sands habitats denominated “Campinas” or “Campinaranas” occurring near blackwater streams, and seasonally flooded blackwater rivers (Capurucho et al. 2020).

Amazonian plant diversity is roughly estimated to be ca. 60,000 species (Lewinsohn and Prado 2005; ter Steege et al. 2016; Cardoso et al. 2017). For bryophytes, the Amazon harbors the second largest species richness of any Brazilian Biome, hosting more than one-third (574 species) of the Nation’s bryoflora (Costa et al. 2020). Studies focusing on the influence of habitat heterogeneity on Amazonian bryophytes have typically been conducted at the level of quantitative floristics and community analyses (Mandl et al. 2009; Oliveira and ter Steege 2015; Sierra et al. 2018; Costa et al. 2020; Cerqueira et al. 2021), while comparative studies at the level of populations incorporating phenotypic and phenological data are, to our knowledge, nonexistent.

In this study, we aimed to conduct a meso-scale investigation of the morphometric and phenological patterns among populations of four ecological generalist bryophyte species in two common Amazonian habitats: white-sands (Campinarana), and plateau Terra-firme forests. Considering the markedly different environmental conditions attributable to these habitat types (Adeney et al. 2016), we set out to test the hypothesis that temporal patterns of gametangial expression and phenotypic attributes associated with vegetative characters differ among populations of these contrasting environments. Specifically, the greater solar radiation and lower humidity levels of white-sands forests would be reflected, for example, in temporally constricted intervals of bryophyte gametangial expression, in combination with reduced surface area of photosynthetically active structures such as gametophytic leaf surfaces. These predictions arise from previous studies showing seasonal effects on the expression and maturation of gametangia associated with variation in local humidity for tropical bryophytes across an elevation gradient (Maciel-Silva and Válio 2011). Furthermore, evidence from experimental studies on infra-specific phenotypic plasticity for vascular plants demonstrate inverse relations among leaf size and solar radiation (Pelabón et al. 2013) suggesting that bryophyte populations of lower statured, white-sands forests characterized by open canopies may be composed of smaller leaf phenotypes (e.g., reduced apices, reduced bases, and reduced total leaf length).

MATERIALS AND METHODS

Study Area

This study was conducted in the municipality of Presidente Figueiredo, Amazonas, Brazil (Figure 1), along the Sussuarana trail: a 4.5 km pathway adjacent to Balbina Village and the Balbina hydroelectric dam. The trail encompasses distinct phytophysiognomies such as terra-firme plateaus - clayey upland forests with relatively closed/denser canopies when compared to other formations in the area, such as white-sands vegetations including Campinaranas (Perigolo et al. 2017; Komura et al. 2021). These white-sands areas, like others alongside the Uatumã River, occupy ancient river terraces positioned between the river’s modern floodplains and the terra-firme plateaus, which surrounds the small, differentiated, isolated patches of sandy vegetation (Demarchi et al. 2021).

The region experiences a tropical wet climate (Köppen-Geiger: Af), with an average annual temperature of 27 °C and average annual rainfall of 1,920 mm (Beck et al. 2018, Resende et al. 2019). The rainy season (November to March) is characterized by an average monthly rainfall of 230 mm, with greater rainfall intensity in March, during which time the temperature ranges from a minimum of 26°C to 30°C. The dry season (May to September) has an average monthly rainfall of 92 mm and temperatures ranging from 27°C to 33°C, with the driest month being September. April and October are months of transition between the wet and dry seasons (INMET, 2018).

Figure 1
Map of the study area at the Sussuarana Trail, central Amazon. A: Approximate location of the study area within the Amazon biome (in Green), and city of Presidente Figueiredo (in Blue). B: Satellite imagery (Sentinel-2) of the Sussuarana Trail surroundings, with the trail location being marked with a white star. The tick marks represent SIRGAS 2000 / UTM Zone 21S projected coordinates.

Focal species and field sampling

Four bryophyte species, Octoblepharum albidum Hedwig, Octoblepharum pulvinatum Mitt. (Octoblepharaceae), Leucobryum martianum (Hornsch.) Hampe. (Dicranaceae) and Pilosium chlorophyllum (Hornschuch) Müller Hal. ex Brotherus (Stereophyllaceae), were chosen for study - Illustrations and photographs of the four species can be found in Figure 2 - as they are common Neotropical ecological generalists which share similar substrates (rotten logs and decaying organic matter). They are characterized as monoicous (autoicous), apart from the dioicous L. martianum (Costa and Peralta 2015; Oliveira-da-Silva and Ilkiu-Borges 2016; dos Santos et al. 2020).

Figure 2
Illustrations and images of the focal species (habit and leaves). A: Octoblepharum albidum Hedwig; B: Octoblepharum pulvinatum Mitt; C: Pilosium chlorophyllum (Hornsch.) Müll. Hal.; D: Leucobryum martianum (Hornsch.) Hampe.; E: Example of morphometrical measurements (AW = apical width, BW = basal width, BL = basal length, AL = apical length, TL = the total length taken across the sum of AW and BW). Illustrations: Anna Kelly N. C. da Silva.

A 5 cm² sample was collected for each of the four species from five populations once a month in both white-sands and upland forests (n = 40 monthly samples). The collections were made at the end of every month for one year (October to September 2016-17). Populations were initially marked with numbered stakes and were at least 15 m distance from each other. The specimens were deposited at the INPA Herbarium (INPA 0201185 to INPA 0201222).

Morphometry

From each sample, three gametophytes were randomly picked from which three leaves were separated for measurement with the use of a magnifying glass, tweezers and needles. Subsequently, leaves were rehydrated with tap water, and placed on slides for five measurements: AW = apical width (measured from one edge to the other in the apical region), BW = basal width (measured from one edge to the other in the basal region), BL = basal length (length from the shoulders, the widest part of the leaf, to the end of the blade), AL = apical length (measured from the tip of the leaf to the shoulders), TL = the total length obtained by adding AW and BW (Fischer et al. 2007; Luna and Velasco 2008) (Figure 2). Measurements were made with the use of a Leica DM2500 optical microscope (Leica Microsystems, Ernst-Leitz-Straße, Germany).

Phenology

From each sample, ten gametophytes were randomly chosen for phenological studies (Greene 1960; Stark 2002; Laaka-Lindberg 2005). If present, perichaetial and perigonial branches as well as the antheridia and archegonia harbored therein were counted and their phenophase classified based on the criteria detailed in Table 1.

Data analysis

Phenotypic variation among individuals from upland and white-sands forests was assessed by means of multivariate analysis, using principal component analysis (PCA) on standardized traits (centered and scaled to unit variance to account for differing measurement scales) - via a correlation matrix. In addition, an Analysis of Similarity (ANOSIM) was performed to test groupings (Supplementary Material, Appendix S1), where the R statistic varies from -1 to 1, and proximity to 1 reveals greater differences among groups (Clarke, 1993). To determine whether the morphometric variables of each species differ among upland and white-sands forests, a t-test was also performed. Additionally, the Mantel test was performed to assess whether geographic distance is correlated with phenotypic variation by comparing two dissimilarity matrices: a spatial distance matrix (Euclidean) and another representing floristic distances (Bray-Curtis) (Borcard et al. 2018).

For the phenological study, circular statistics were employed to assess differences in the seasonality of reproductive events between the two areas. This analysis aimed to determine whether there are distinct patterns of reproductive receptivity and temporal separation in reproductive periods between the two locations (Morellato et al. 2010). Firstly, the monthly occurrences of phenophases were recorded in absolute numbers for each environment. Afterwards, the months of the year were converted into degrees ranging from 0° to 360° (January as 0° and December as 360°), and the mean vector value (angles), the length of the mean vector to determine the directional consistency of the data, the median (angles), circular standard deviation (CSD) and Rayleigh test (Z and P) to verify the uniformity of the data distribution were calculated (Zar 1999; Morellato et al. 2010).

For the morphometric studies, the native GUI for R ver. 3.2.5 (R Core Team 2024) was used with the “vegan” package (Oksanen et al. 2024). For the phenological studies, the software “Oriana” ver. 4.02 (Kovach Computing Services 2024) was used.

RESULTS

A total of 4,320 leaves were analyzed, totaling 21,600 measurements of morphometric characters in the gametophytic phase. Multivariate analysis using principal component analysis (PCA) of the five gametophytic morphometric characters among the two environments indicated morphological discontinuities in three of the four focal species: Leucobryum martianum, Octoblepharum pulvinatum and Pilosium chlorophyllum (Figure 3; Supplementary Material, Figure S1). As for O. albidum, no significant morphological differences were detected among individuals inhabiting both environments.

Figure 3
Principal component analysis (PCA) of phenotypic variation in gametophytic characters for four bryophyte species: Leucobryum martianum (A), Octoblepharum albidum (B), Octoblepharum pulvinatum (C), and Pilosium chlorophyllum (D), across terra-firme (TF, triangles) and campinarana (WS, circles) habitats. Percentages indicate the variance explained by Principal Component 1 (PC1) and Principal Component 2 (PC2). Variable contributions (contrib) are proportional to their influence on the axes, with higher values denoting greater importance. Abbreviated phenotypic traits correspond to specific morphological characters analyzed (L.A = apical width, L.B = basal width, C.B = basal length, C.A = apical length, C.T = the total length taken across the sum of AW and BW.

When comparing gametophytic characters among environments using the t-test, these three species showed significant differences in more than one or in all morphometric characters analyzed. However, O. albidumshowed a difference in only one morphometric character: basal length (Table 2). The length-related characters, such as apical, basal and total length, showed the greatest frequencies of significant difference among the two habitats; however, the hypothesized trend of generally reduced leaf surface area for white-sands forests was not reflected in the results (Table 2). In fact, high inter-specific variability was documented among the phenotypic trends for leaf characteristics among the four study species.

For three of the species, L. martianum, O. pulvinatum and P. chlorophyllum, positive results from the Mantel test suggest that the differences found were correlated with geographic distance between populations (Figure 4).

Figure 4
Mantel Test of morphometric variation of the four studied bryophyte species related to the geographical distance (km).

Phenology

A total of 4,800 gametophytes were analyzed and 19,862 archegonia, 21,244 antheridia and 308 sporophytes were observed during the study period. Sexual expression was reported for all species in both habitats as either antheridia (male gametangia), archegonia (female gametangia) and/or sporophytes. White-sands specimens showed the highest rates of sexual expression when compared with their upland peers. For both environments, P. chlorophyllum had the highest percentages of sexual expression, and L. martianum showed the lowest (Figure 5A).

Figure 5
Relative percentage of sexual expression (A) and relative percentage of antheridial (♂) / archegonial (♀) expression / sporophyte output (B) for the four species.

Interestingly, gametangial frequencies observed in populations in both environments showed strong inter-specific variation with multiple cases of strong male bias em sex expression. For example, the sex ratio (SR = antheridia/archegonia) was nearly 2.5x higher (mean SR: 2.24) for males of L. martinanum in white-sands habitats, and more than 6x higher for males of O. albidum in upland forests. Nonetheless, interspecific variation in SR among white-sands populations was observed as L. martianum and O. albidum showed high percentages of antheridia expression, while O. pulvinatum and P. chlorophyllumexpressed proportionally greater archegonial expression (Figure 5B, Table 3). Furthermore, substantial intraspecific variation in SR among the two habitats was also observed across all species suggesting the influence of local environment on sex expression.

Only phenophases 3 (gametangia hyaline or green), 4 (gametangia brownish with broken apices), 5 (gametangia with zygote or sporophyte buds) and aborted (gametangia brownish and withered) were observed in the study (Figure 6; Supplementary Material, Table S1, Table S2). The four bryophyte species studied presented distinct seasonal patterns for the phenophases of both antheridia and archegonia among habitats, associated with rainfall seasonality (Figure 6; Supplementary Material, Table S1, Table S2). Sexual expression in white-sands forests was temporally restricted more so than those of upland forests. Furthermore, aborted archegonia were observed significantly more frequently in the rainy season in white-sands forests.

Figure 6
Circular histogram of Leucobryum martianum, Pilosium chlorophyllum, Octoblepharum albidum and O. pulvinatum showing the frequency of phenophases observed in white-sands and upland habitats.

Curiously, P. chlorophyllum was observed in dehiscence (phenophase 4) with greater frequency in the dry season, especially in white-sands forest. Regarding its archegonia, phenophases 3 and 4 were observed within both environments during the rainy season, while phenophase 5 during the dry season. The frequency of abortive phenophase observations among habitats was best represented in white-sands during the dry season, and upland forests during wet months.

For both antheridia and archegonia, the post-antherozoid maturation and post-oosphere receptivity of the dehiscent and abortive phases occurred throughout the entire collection cycle. Curiously however, during the rainy season of November/2016 to May/2017 a low occurrence of the mature phase was reported for both males and females. Circular analyses (Figure 6; Supplementary Material, Table S1, Table S2) revealed a seasonality only for the appearance of female gametangia during the post-maturation phase, for both environments.

DISCUSSION

To our knowledge, no studies of Amazonian cryptogams have yet to address the role of habitat differentiation on phenotypic and phenological variation. Considering that the Amazon is an epicenter for the evolution of neotropical plants (Antonelli et al. 2018), understanding patterns of inter-specific variation in reproductive phenology and phenotypic divergences is fundamental for revealing the contributions of Amazonian habitat differentiation to plant speciation processes. Our study reveals substantial infra-specific phenotypic variation among photosynthetically active characters in moss gametophytes, as well as phenological variation in the temporal expression of male and female gametangia among the two habitats. Furthermore, our observations of sex ratio expression reveal an atypically pervasive male bias for three of the four study species among all populations independent of habitat type.

White-sands and Terra-firme plateaus are characterized by different edaphic and microclimatic conditions. That is, white-sands are typically drier and subject to greater solar radiation due to the reduced density in canopy tree architecture and sandier soil types (Adeney et al. 2016; Fine and Baraloto 2016). Bryophytes, dependent on the presence of water for successful fertilization, are generally temporally restricted in sexual expression in drier micro-climates. Indeed, all four species in both environments displayed opportunistic strategies as gametangial production was frequent, thus allowing for the convergence of reproduction and gamete maturation with local water availability. In general, our phenological analysis reveals temporal restrictions in sexual expression particularly in males among white-sands populations. However, interspecific variation was observed as best illustrated in the case of L. martianum with unexpectedly high sexual expression during the dry season for populations inhabiting both habitats.

The biology of tropical bryophytes remains understudied; however evidence suggests that they employ a diverse array of reproductive strategies with characteristically high phenotypic variation within and among populations (Maciel-Silva et al. 2011; Pereira et al. 2013; Laenen et al. 2016; Silva et al. 2017; dos Santos et al. 2020). Such factors putatively contribute to their potential fitness advantages as best illustrated by their greater ecological amplitude when compared to Amazonian phanerogams (pers. obs. CEZ). Curiously, in our study however the ecological generalist and cosmopolitan species Octoblepharum albidum (Santos & Stech 2016) revealed less phenotypic variation than the other three species suggesting that for this species habitat heterogeneity is not the principal factor driving population divergence. Nonetheless, similar studies at regional spatial scales are necessary to test this hypothesis.

Leaf traits, such as surface area, typically respond to environmental exposure such as light and water resources, which when adequate result in the resource reallocation to other life history aspects such as growth and reproduction (Andrade et al. 2013; Taiz et al. 2017; Glime 2021). Although significant morphometric divergences were observed among the measured traits for all four focal species, the direction of these changes among habitat types was not consistent among species. Contrary to our hypothesis, that leaf surface area would diminish in white-sands habitats relative to terra-firme plateaus, no common trend across species was documented. In fact, phenotypes were mixed and complex with no established pattern among habitats and species. In summary, high observed phenotypic variation was not related to identifiable aspects of habitat differences.

Finally, and perhaps most surprisingly was the discovery of highly lopsided sex ratios towards male expression. Much interest has been dedicated to the nearly ubiquitous observation of female expression dominating bryophyte populations globally: a phenomenon coined the shy-male hypothesis (Stark et al. 2010). Interpretations of this anomalous reproductive behaviour (i.e., male expression generally more common among flowering plants (Barrett et al. 2010) point to the possibility that resource allocation, contrary to phanerogams, is relatively higher for male expression in bryophytes thus resulting in sex ratios (SR) skewed to female expression. Interestingly, our results show SR contrary to the historical pattern of female dominated expression documented for temperate species. A possible explanation may lie in the strong seasonality in rainfall which typically comes in tropical deluges reducing antherozoid densities in short episodes of heavy rainfall. Aquatic bryophytes adapted to a similar ecological challenge by allocating resources to male expression to augment antherozoid densities to compensate for water volume (Shevock et al. 2017). Whether reproductive strategies of terrestrial Amazonian bryophytes have converged with rheophytic species due to similar ecological pressures remains to be tested.

CONCLUSIONS

Our findings provide clues into the evolutionary potential of Amazonian bryophytes as we show that these four common neotropical bryophytes exhibit substantial phenotypic and phenological variation at the meso-scale in two common habitats. The unexpectedly high rate of male expression is noteworthy as such a reproductive strategy is globally uncommon among bryophytes. Future studies should aim to gather data across greater spatial scales to deepen our understanding of the mechanisms that drive survival, maintenance, reproduction and speciation of bryophytes in tropical environments.

ACKNOWLEDGMENTS

This research was funded by a master’s scholarship from the Brazilian Coordenação de Aperfeiçoamento de Pessoal de Ensino Superior (CAPES-DS/MEC), given to AKNCS. We thank the Botany Graduate Program at the Instituto Nacional de Pesquisas da Amazônia (PPG-BOT/INPA), Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) and Natura for the logistical, financial and institutional support. TAJ was partially supported by a master’s scholarship given by the Fundação de Amparo à Pesquisa do Estado do Amazonas (POSGRAD/FAPEAM) during the writing of this manuscript.

REFERENCES

Data availability

The data that support the findings of this study were published in this article in the “Supplementary Material”.

SUPPLEMENTARY MATERIAL

Silva et al. Exploring local phenological and phenotypic variation of mosses (Bryophyta) in two common Amazonian habitats

Figure S1
Analysis of Similarity (ANOSIM) Results

Edited by

Publication Dates

History