1 Introduction
Despite being recognized as a foundation for the theory of evolution by natural selection, the importance of intraspecific trait variation (ITV) has been neglected over time in ecology (Bolnick et al., 2011). Recently, there has been a resurgence of ecological interest in ITV, stimulated by the proliferating studies that underline the tremendous effects of ITV on community assembly and ecosystem functioning (Violle et al., 2012; Cardou et al., 2022). However, the patterns and sources of ITV in themselves are still unclear (Cope et al., 2022). For example, ITV can be structured at various spatial scales in relation to different drivers. Between-population ITV tends to be shaped by large-scale environmental gradients, whereas within-population ITV is more likely the consequence of heritable differences and/or plastic responses to the local environment (Albert et al. 2010; Martin et al., 2017). Alternatively, ITV can be unstructured if the trait is mainly determined by stochastic processes or constrained by fitness tradeoffs (Kendall & Fox, 2003; Moran et al., 2016). Elucidating ITV distribution across multiple scales is yet crucial for understanding and predicting ecological responses to global changes (Cochrane et al., 2015; Moran et al., 2016).
Trait-based plant studies to date have largely focused on vegetative traits (e.g., leaf and root functional traits), with less attention placed on seed traits despite the significance of seed traits for plant regeneration and long-term persistence (Saatkamp et al., 2019). Knowledge of seed traits is particularly urgent for mangroves, which once covered over 200,000 km2 of sheltered tropical and subtropical coastlines but nowadays has been disappearing worldwide for decades at an extremely high rate (Duke et al., 2007; Friess et al., 2019; Giri et al., 2015). Due to the property of low species richness and redundancy, mangrove degradation is always followed by pronounced losses of ecological multifunctionality (Arifanti et al., 2022; Donato et al., 2011; Feller et al., 2010). Such situation emphasizes mangrove recruitment and thereby calls for characterizing seed/propagule traits that represent key dimensions of the ‘regeneration niche’ (sensuGrubb, 1977) in mangroves (Feller et al., 2010; Peterson & Bell, 2012).
Many mangroves (e.g., Rhizophoraceae, Avicenniaceae) have evolved a special reproductive strategy of viviparous seeds (germinating precociously while attached to the maternal plant), probably reflecting an adaptation to the salty and flooding intertidal environments (Feller et al., 2010). Due to the lack of seed dormancy, mangrove forests destroyed by severe disturbances such as hurricanes and tsunamis may not have local seed reserves, necessitating recolonization through long-distance seed dispersal from relatively undisturbed populations (Nettel & Dodd, 2007). This is particularly the case for Rhizophoraceae forests, as Rhizophoraceae species lack the capacity of resprouting from damaged trees as Avicenniaceae species (Baldwin et al., 2001). Therefore, dispersal/retention is a key dimension of propagule functions where focal traits should be targeted for mangroves (Van der Stocken et al., 2019). Whether dispersed or retained, establishment is a prerequisite for any propagules to contribute to the regeneration, representing another critical dimension of seed trait (Krauss et al., 2008).
Most ITV studies in mangroves, though few in absolute number, have investigated propagule traits (e.g., size and weight) in relation to establishment, rarely considering the dispersal/retention dimension (Saenger & West, 2018; Yang et al., 2020; Zhu et al., 2021). These studies showed that temperature and/or precipitation may have given rise to structured ITV in mangroves across geographical gradients (Saenger & West, 2018; Yang et al., 2020; Zhu et al., 2021), a pattern commonly reported in terrestrial forests (Kumordzi et al., 2019). Nevertheless, as mangroves are coastal vegetation, their trait variation may also be shaped by oceanic factors including salinity and tidal currents (Richards et al., 2021; Sousa et al., 2007). Additionally, maternal plants can also affect propagule traits, both through genetically fixed differences and environmentally induced transgenerational plasticity (Alam et al., 2018; Cochrane et al., 2015). However, the relative importance of these abiotic and biotic factors in shaping the distribution of ITV across multiple scales, and whether the patterns differ between the propagule function dimensions (establishment vs. dispersal/retention) are still poorly understood.
Here we use Kandelia obovata as the model species to investigate the issues. Kandelia obovata (Rhizophoraceae) is the most cold-tolerant true-mangrove species and has a wide latitudinal distribution along the southeast coast of China (Sheue et al., 2003; Wang et al., 2011), providing an ideal system for studying the ITV of mangroves. Using a stratified sampling design across a 9° latitudinal gradient, we analyzed the structure of intraspecific variability for five hypocotyl functional traits in relation to propagule dispersal/retention and establishment. The stratified sampling design and contrasting environmental conditions allowed us to address the following questions: (i) How is ITV structured spatially (between populations, maternal trees, hypocotyls)? (ii) What is the major driver (climatic, oceanic, or maternal factors) shaping the distribution of ITV? Based on the results from previous studies (Saenger & West, 2018; Zhu et al., 2021), the variability of traits on the establishment axis is expected to be higher between populations than within populations and predominantly shaped by abiotic factors. By contrast, the trait on the dispersal/retention axis, likely more reflecting a tradeoff between post-disturbance recolonization and local recruitment (Sousa et al., 2007; Van der Stocken et al., 2019), may be less structured or even unstructured with regard to particular environmental gradients.
2 Materials and methods