Seed tolerance to post-fire temperature fluctuation of Cerrado legume shrubs with micromorphological implications

Highlights

• Heat tolerance drives regeneration from seed in Cerrado legume shrubs.

• No effect on physical dormancy break mediated by post-fire temperature fluctuation.

• Seed structures take part on water uptake in nondormant seeds within the seed lots.

Abstract

Multiple environmental cues may break seed physical dormancy (PY) in fire-prone ecosystems, but disentangling the roles of such cues remains challenging. Current research has mostly focused on the role of high temperatures (heat shocks) in PY alleviation, while post-fire environmental conditions are less understood. Here, we examined the role of post-fire temperature fluctuation on seed germination and viability of Cerrado legume shrubs. Specifically, we investigated (1) whether PY is broken by temperature fluctuation; (2) whether seed viability is affected by temperature fluctuations; and (3) micromorphological changes in the seed coat following exposure to temperature fluctuation. Seeds of nine legume species were collected in frequently burned Cerrado savannas. We simulated post-fire temperature fluctuation in germination chambers, with daily temperature fluctuation ranging from 18 to 55 °C for 90 days; controls consisted of untreated seeds (room temperature). We conducted germination trials with unblocked and blocked hilar region of seeds. In addition, seed samples were fixed and subjected to micromorphological analysis. Instead of dormancy break, three species showed decreased seed germination for both unblocked and blocked hilar regions, while other two species showed decreased germination in blocked seeds only. Nevertheless, random cracks formed across the seed coat were found in control and treated seeds, which displayed a proportion of nondormant seeds. Within nondormant seeds, both hilar and extra-hilar regions (such as the pleurogram) may take part on water uptake. Despite the lack of dormancy break, most Cerrado legume seeds persisted viable after post-fire temperature fluctuation, thus avoiding establishment under hazardous conditions. Seed tolerance to heat shocks and post-fire temperature fluctuation likely drives regeneration from seed in burned Cerrado areas.

Introduction

Legume seeds usually show an impermeable seed coat, that confers impermeability to water, referred as physical dormancy – PY (Baskin and Baskin, 2014). The hard seed coat is related to seed desiccation tolerance and confers mechanical protection, and helps to explain propagule persistence in the soil (Dalling et al., 2011; Tweddle et al., 2003). Therefore, legumes have successfully colonized most ecosystems worldwide, especially those with seasonal climates (Rubio de Casas et al., 2017). PY in legumes is caused by sclerified cells (sclereids) arranged in a palisade layer of cells in the testa, which is the outer integument, thus preventing water absorption (Baskin, 2003; Smýkal et al., 2014). In an evolutionary context, PY is a more derived type of dormancy and has evolved in response to different selective pressures, such as drought, fires, and seasonality (Lamont et al., 2019; Santana et al., 2020; Willis et al., 2014).

In fire-prone ecosystems, multiple environmental cues have been recognized to alleviate PY (Santana et al., 2013). Such cues include fires and/or daily fluctuations of soil temperature (Luna, 2020; Moreira and Pausas, 2012; Ooi et al., 2014; Santana et al., 2010). Higher temperature fluctuation is reached in the soil surface of vegetation gaps, formed by fire removal of aboveground vegetation (Bradstock and Auld, 1995; Daibes et al., 2017; Fidelis et al., 2012; Fidelis and Blanco, 2014). The amplitude between day and night temperatures often results in PY break under dry conditions (Baeza and Roy, 2008; Santana et al., 2010), thus facilitating seed recruitment earlier with subsequent rainfall events.

Because PY seeds might show dormancy break related to a flush of dry heat dose, different temperature thresholds seem to disrupt the impermeable seed coat in legume species (Liyanage and Ooi, 2017; Morrison et al., 1998). Therefore, PY break mechanisms in fire-prone ecosystems have been classified as ‘facultative’ vs. ‘fire-related’ (Moreira and Pausas, 2012; Ooi et al., 2014). Facultative species already show PY break when exposed to temperatures reached in the soil surface, up to 60 °C, contrasting to fire-related temperatures which rise above 80 °C (Ooi et al., 2014; Santana et al., 2013).

During PY break, seed coat structure shows specific regions – named ‘water gaps’ – which can be disrupted or pulled apart to allow imbibition, and thus germination to occur (Baskin, 2003; Gama-Arachchige et al., 2013; Serrato-Valenti et al., 1995). The seed coat of legumes usually includes the lens, typically argued as a specialized structure for water uptake (Baskin, 2003; Baskin and Baskin, 2014). The hilum and/or micropylar region have also been reported as the principal disruption points during PY break in Fabaceae (de Paula et al., 2012; de Souza et al., 2012). Other seed ornaments might play the functional role of water gaps, including the pleurogram in tropical Senna (Rodrigues-Junior et al., 2019).

The ability of seeds to tolerate high temperatures (heat tolerance) to warrant a propagule supply after disturbance is also considered an important trait in fire-prone ecosystems (Fidelis et al., 2016; Jaureguiberry and Díaz, 2015; Ribeiro et al., 2015). Because fires remove the aboveground vegetation, most seeds can be dispersed to vegetation gaps, subjected to higher temperature fluctuations in the soil surface compared to vegetated conditions. In the Cerrado (Neotropical savanna), both fire and temperature fluctuations might show little to no effect on dormancy alleviation (Daibes et al., 2017; Dairel and Fidelis, 2020; Zupo et al., 2016).

Some Cerrado legumes show variable proportions of nondormant (thus permeable) seeds able to germinate irrespective of pre-germination treatments (Daibes et al., 2019). Those cases can be classified as ‘intermediate’ dormancy, showing proportions ranging from 30 to 70% of nondormant seeds within the population (Baskin and Baskin, 2014; Dayrell et al., 2017). Small seeded legumes have been reported to show up to 40% of nondormant seeds within populations in Australia (Morrison et al., 1992). Nevertheless, it has not been identified whether the proportion of nondormant seeds show open water gaps or random cracks across the seed coat from the micromorphological point of view.

Anatomical changes during alleviation of PY have been assessed in some tropical forest tree legumes (e.g. de Paula et al., 2012; Geisler et al., 2017), but remain understudied in the Cerrado, where fire-related cues are expected to be common. The Cerrado harbors different vegetation types (Oliveira-Filho and Ratter, 2002; Ribeiro and Walter, 2008), including tropical savannas subjected to frequent fires (Miranda et al., 2009). This complex of vegetation types shows a megadiverse flora, considered as a global biodiversity hotspot (Myers et al., 2000), and recognized as diversification center for legume groups such as Mimosa (Simon and Proença, 2000).

Hence, we aimed to contribute to the debate on the factors that alleviate PY under natural conditions, we investigate the role of post-fire daily temperature fluctuation, simulating gap-reaching temperatures on germination and seed viability of Cerrado legume shrubs. In fire-prone ecosystems, the role of fire-related cues in promoting regeneration from seeds is relatively well-studied, but the role of post-fire temperature fluctuations is less understood. Specifically, we examined (1) whether PY seeds show dormancy break related to post-fire daily temperature fluctuation; (2) seed tolerance to temperature fluctuation, considering seed viability after post-fire simulations; (3) micromorphological changes caused by temperature fluctuation in the seed coat to get insight into putative structures allowing water uptake. We expect high tolerance to simulated temperature fluctuation, but little to null effect on PY break. Micromorphological analysis could elucidate the understudied role of seed structures as water gaps, helping us to understand the factors driving post-fire regeneration from seed in tropical savannas.