The study of vegetative and reproductive phenophases of plants is critical for understanding aspects related to plant behavior in different environments. In the tropics, there is a gap of understanding related to plant phenology since the theoretical framework on the topic has been built from perspectives of the temperate region. Furthermore, there are few studies in tropical regions influenced by anthropic conditions, which may be important for understanding these issues in the face of future climate scenarios. This study aimed to describe the vegetative and reproductive phenology of trees in an urban subtropical forest community and to test the influence of climatic variables on the tree community. In an urban forest fragment in Rio Grande do Sul, Brazil, eight individuals of locally dominant species were monitored: Allophylus edulis, Casearia sylvestris, Guarea macrophylla, Mimosa bimucronata, Myrsine coriacea, Myrsine umbellata, Schinus glandulosum, and Schinus terebinthifolia. The monitoring occurred every two weeks, for two years, with the recording of the presence of leaf flushing, leaf shedding, flowering, and fruiting phenophases for each tree. The seasonality of the species was tested using the Rayleigh test. We described the common pattern of community phenological activity by a Principal Component Analysis. Finally, we correlated the common patterns of each phenophase in the community with climatic variables of total precipitation, average temperature, and day length. All species showed a non-uniform phenological pattern for the evaluated phenophases despite the variable intensity. We evidenced common patterns for the community only for the vegetative phenophases. The reproductive phenophases of flowering and fruiting present themselves independently among species in the community. Finally, we identified influences only of temperature and day length on the vegetative phenophases.

Atlantic Forest, climate, flowering, fruiting, leaf flushing, leaf shedding, seasonality

Phenological studies have contributed to understanding the effect of climate on the seasonal life cycle of organisms (Garcia 2006; Forrest and Miller-Rushing 2010; Tang et al. 2016; Ponti and Sannolo 2022). In arboreal individuals, these studies mainly contribute to the description of the temporal variation of reproductive (i.e., flowering and fruiting) or vegetative (i.e., leaf flushing and leaf shedding) phenophases in a population or community level (Andreis et al. 2005; Polgar and Primack 2011; Garcia-Barreda et al. 2021). Furthermore, the influence of biotic interactions and environmental conditions (e.g., climate) that act on the physiological mechanisms of plants is a topic that has been widely explored (Almeida and Alves 2000; Ding and Nilsson 2016; Singh et al. 2016).

Temperate regions have been the focus of phenological studies for a long time because they present a well-defined thermal seasonality, where low winter insolation defines the dormant period of plants (Staggemeier et al. 2019). Recent advances in tropical dendrochronology have been breaking the paradigm that tropical trees generally lack seasonal growth rhythms (Quesada-Román et al. 2022; Zuidema et al. 2022). However, knowledge about the phenology of tropical and subtropical plants is still insufficient (Staggemeier et al. 2019; Davis et al. 2022). Thus, the theoretical framework on the topic has been built from a perspective that is contextually biased to temperate regions (Chambers et al. 2013; Davis et al. 2022). This paradigm makes it difficult to understand how tropical trees respond to climate change (Staggemeier et al. 2019; Davis et al. 2022).

Efforts to understand phenological patterns in tropical and subtropical regions face challenges, especially in the many distinct phenological patterns evidenced by the high diversity of species (Davis et al. 2022). Besides that, some knowledge gaps should be considered in future studies (Sakai and Kitajima 2018; Masson-Delmotte et al. 2021; Davis et al. 2022). First, most phenological studies in the southern hemisphere have temperature and day length as the only climatic variables tested (Chambers et al. 2013). Also, phenology studies in this region should cover a significant temporal scale, with more than one year of measurements (Sakai and Kitajima 2018; Davis et al. 2022). Furthermore, understanding the phenological cycles and patterns of organisms in areas influenced by anthropogenic conditions may reflect important interpretations from a conservation perspective, especially under future climate scenarios (Masson-Delmotte et al. 2021).

Urban forests are important components of urban landscapes as they provide environmental services such as reducing heat islands, preventing soil erosion, and improving air quality (Werner 2011). Furthermore, the existence of these ecosystems reflects a positive interaction for the maintenance and survival of animals and plants in urban landscapes (Jensen et al. 2021; Rigacci et al. 2021). In this sense, phenological studies of subtropical trees in urban forests can contribute to the understanding of plant behavior in environments influenced by anthropogenic conditions. Recent studies have been dedicated to understanding the effect of urbanization on trees and urban forests, identifying greater growth but lower longevity (Pretzsch et al. 2017) and in some cases delay in the onset of phenophase seasons (Li et al. 2022, in a temperate region). However, there is still a knowledge gap about the effects of urbanization on tree phenology in urban forests under subtropical conditions due to the absence of data series with sufficient spatial and temporal coverage (Chambers et al. 2013; Jochner and Menzel 2015).

This study aimed to (i) describe the annual vegetative and reproductive phenology of eight dominant tree species in an urban subtropical forest community, and (ii) test the influence of seasonal climate on the phenology of the community. The few studies of tree phenology in native forest environments in the South and Southeast regions of Brazil have shown responses related mainly to the seasonal cycle of day length (Morellato et al. 2000; Marques and Oliveira 2004; Marchioretto et al. 2007; Athayde et al. 2009; Rubim et al. 2010). Thus, we hypothesize that the climatic response of the tree community in this urban subtropical forest will be similar to previous studies for tropical regions, with a significant influence of day length on phenophases.

The tree species evaluated in this work are common in the early stages of ecological succession (Cappelatti and Schmitt 2009), and some have been widely used in agroforestry systems (A. edulis and M. bimucronata) and forest restoration programs (A. edulis and S. terebinthifolia) in the Atlantic Rainforest, especially in Rio Grande do Sul (Silva et al. 2019; Chemello et al. 2023; Domingues et al. 2023). Therefore, the results of this study provide knowledge about the ecology and both reproductive and vegetative phenological variation, as well as important information for sustaining future public policy programs for seed collection and promotion of plantations of these species and for the extraction of non-timber forest products, influencing the economy of small families (Domingues et al. 2023). Therefore, promoting knowledge about phenological data, especially through field observations in a natural environment, provides accurate information that can aid in future projections about species behavior (Tang et al. 2016; Zurell et al. 2020; Ponti and Sannolo 2022).

The leaf-flushing phenophase showed high seasonality values for most of the community, except for M. coriacea, M. umbellata, and S. terebinthifolia, which resulted in a low seasonality pattern with a more uniform expression of the phenophase throughout the years, although it is still classified as seasonal. This result corroborates the fact that these species are not classified as perennials and, therefore, their vegetative cycle is expected to be less regular compared to perennial plants (Bencke 2005; Bauer et al. 2012; Milani 2013). Tropical species may have different strategies regarding leaf-flushing, mainly related to defense against herbivore attacks (Corlett 2009; Yoneyama and Ichie 2019). Leaf-flushing activity throughout the year may be related to more resistant leaves and higher photosynthetic capacity and, on the other hand, sprouting occurring over a longer period may indicate some chemical influence against herbivore attack (Yoneyama and Ichie 2019). In this sense, the studied community may show distinct patterns against herbivore attacks related to variations in sprouting.

On the other hand, for leaf shedding, most species showed low seasonality. Similar results have been found in other locations for A. edulis (Milani 2013), C. sylvestris (Marchioretto et al. 2007), G. macrophylla (Bencke and Morellato 2002), M. coriacea (Bauer et al. 2012), and M. umbellata (Bencke 2005). Therefore, the community, regarding leaf shedding, can be considered slightly deciduous. This may be explained by damage associated with leaf and branch falls caused by winds and storms, which are common in the region (Janzen 1980). In this sense, future phenological studies related to the vegetative activity of tree species in the tropics should consider the frequency of these events throughout the observation period. Only S. glandulosum differed from this pattern of low seasonality, showing a concentration of its phenophase at the beginning of winter, corroborating what was found in other studies (Bencke 2005) and the fact that leaf deciduousness is considered a survival strategy for unfavorable periods, such as the cold winter months (Chabot and Hicks 1982).

Flowering was the phenophase with the best-defined seasonal pattern for all species. A. edulis and C. sylvestris had their flowering peak in early spring, a pattern recognized for these species (Andreis et al. 2005; Bencke 2005). Most other species flowered in autumn and winter, as already observed for M. bimucronata (Bencke 2005), M. coriacea (Bauer et al. 2012), and S. terebinthifolia (Milani 2013). It is also worth noting that no species showed flowering in January or July, months with extreme climatic conditions of temperature and day length, which may be associated with less favorable times for pollinators or even for the physiology of fertilization (Liberman 1982; Alencar 1994). Climate extremes tend to be more pronounced in urban areas (Qian et al. 2022), which may further potentiate their effects in our study location.

The fruiting peak of most species occurred in early winter or spring, with most species showing little marked seasonality. Fruit production in tree communities occurring during long periods of the year has been recorded in other studies, and the period between years can be highly variable (Spina et al. 2001). The fruiting period is strongly related to the dispersal strategy of the species (Freire et al. 2013). Among the species in this study, only M. bimucronata presents autochoric dispersal, with dispersal occurring in the autumn and winter months in agreement with the literature (Carvalho 2004). In other studies in Atlantic Forest environments, especially in pioneer secondary forests, the zoochory is also more common than autochory (Negrini et al. 2012; Freire et al. 2013). In this sense, it is worth noting the strong influence of fauna, especially avifauna, on the fruiting of the species in this study since disturbances in these groups can directly impact the seasonality of this phenophase (Negrini et al. 2012). Among the species that presented this well-defined phenophase, A. edulis and C. sylvestris are highlighted, both corroborating observations in other localities of the Atlantic Forest (Bencke 2005; Athie and Dias 2011). Finally, it is also important to note that plant age is related to the seasonality of fruiting events (Schorn 2003). This factor was not considered in this study, but, for future studies, it may be important to include this information.

We evidenced a common pattern of activity frequency among the species of the tree community only for the vegetative phenophases. The occurrence of independent reproductive phenophases among species in the community can be explained by the fact that flowering and fruiting are strongly dependent on factors such as age, pollination, dispersal, and competition, which can have their effects at different times of the year among species (Frankie et al. 1974; Alencar 1994; Forrest and Miller-Rushing 2010; Ding and Nilsson 2016; Viana et al. 2020; Garcia-Barreda et al. 2021). The common species pattern between vegetative and reproductive phenophases has also been found in other studies conducted in the subtropical region (Marchioretto et al. 2007; Athayde et al. 2009). However, it is worth noting that the occurrence of common patterns among species, that is, for the tree community, is little explored in phenological studies. Thus, we emphasize that new studies should seek to test these patterns beyond the description by species, aiming to better understand the leaf and reproductive phenology of tree communities in the subtropical region.

The vegetative phenophases were the only ones that showed a common pattern among the species in the community; therefore, their responses to climate were evaluated. We only found influences of temperature and day length, corroborating the few studies carried out in the subtropical region (Marques and Oliveira 2004; Bencke 2005). The leaf flushing activity in the community was negatively related to day length and temperature conditions. The production of vegetative structures usually occurs by avoiding less favorable climatic periods (Schaik et al. 1993; Ding and Nilsson 2016), which corroborates the results of this study since no species showed leaf flushing during the months with higher temperatures and longer days.

For leaf shedding, we evidenced two distinct patterns in the community related to a group of plants that present the phenophase during periods of lower day length and others in periods of higher day length. Leaf shedding in periods of longer days may be related to the change in the nutritional characteristics of the soil caused by the production of litter in early spring (Morellato and Leitão-Filho 1992; Santos and Fisch 2013). This change causes influence on the physiological aspects of the plant and, consequently, can affect the phenological patterns of the community (Costa et al. 2007; Cardoso 2009). We also evidenced a pattern for all species in the community regarding leaf shedding, which was characterized by a negative influence of temperature and day length. Responses of leaf shedding to day length are expected in the tropical region (Wright and Van Schaik 1994), but they have also been related to water changes (Marques et al. 2004; Marques and Oliveira 2004). This contrasts with our study, where we did not identify a significant relationship with precipitation. In this sense for this forest community, we evidenced inverse climatic correlations between leaf flushing and leaf shedding, as also evidenced by other studies with other species in southern Brazil (Viana et al. 2020).

In general, the phenological activity of the species in this urban subtropical forest environment in relation to natural forest environments was similar. We confirmed the hypothesis of our study since the phenophases of the tree community were mainly related to day length. However, some distinct patterns were observed, especially in the times of occurrence of the phenophases. Changes (delays or advances) in the typical period of phenophases have been observed in trees in urban conditions (Liang et al. 2016; Li et al. 2022), which synergistically with other factors may be affecting some species of the studied community. Future studies involving genetic factors in the physiology of these species are essential for a better understanding of plant ecology (Ding and Nilsson 2016; Singh et al. 2016). This knowledge, in addition to other surveys at larger temporal and local scales, is fundamental even for understanding the response of the tree community under future climate change scenarios (Polgar and Primack 2011).

During the two years of monitoring, we increased our knowledge about the ecological variation of vegetative and reproductive activities of eight common native species in secondary forests. We conducted high resolution phenological monitoring for eight tropical species; three of them (Mimosa bimucronata, Schinus glandulosum, and Schinus terebinthifolia) have few or no studies for southern Brazil. In addition, in this work we address phenology at a community scale in contrast to the species scale, which is usually analyzed. We found that the tree community presents common patterns of phenological activity for the vegetative phenophases, in contrast with the reproductive ones, which seem to present independent patterns among species in the community. We showed that climate seasonality is also a significant factor for vegetative phenology in this urban forest community in southern Brazil, especially considering variations in day length, as evidenced in other studies in the tropics. Finally, we emphasize that this study can serve as a basis for future comparisons regarding differences in the phenology of trees growing in urban and natural environments. In this sense, we contribute to further studies about the influence of urbanization on the phenology of tropical species.

The authors thank Universidade do Vale do Rio do Sinos for the peer mentoring program of the Community Ecology course for supporting the study project of forest fragments in the São Leopoldo Campus, and all the undergraduate and graduate studies involved in the project.