Humans have severely altered ecosystems worldwide (Ellis, 2015). Currently, it is estimated that more than three quarters of the Earth’s land surface has been directly altered by anthropogenic activities. These modifications have intensified in the last century, with almost a third of global land surface changes occurring during the last 60 years (Winkler et al., 2021). The remaining original ecosystems are increasingly threatened by ongoing land use change processes (Newbold et al., 2015), and for those habitats that persist, by increasing stress from climate change (Mantyka-Pringle et al., 2015; Segan et al., 2016). Climate change can be a severe threat to the long-term viability of species, particularly due to modifications of temperature and precipitation regimes (Urban, 2015; Román-Palacios and Wiens, 2020; Reynaert et al., 2021). Furthermore, habitat degradation and fragmentation induced by land use change processes can have synergistic effects with climate change, which may increase the current rates of biodiversity loss (García-Valdés et al., 2015; Selwood et al., 2015; Northrup et al., 2019). In the case of long-lived organisms with limited dispersal and long generation times, such as many tree species, viability of species will largely depend on the capacity of the remaining populations to reproduce on-site, which will be contingent climatic conditions within the area (Selwood et al., 2015). There is evidence that climate change is already affecting crucial processes in the life cycle of trees, namely seedling, germination, recruitment, and establishment (e.g., Walck et al., 2011; Yukich et al., 2023).

However, besides climatic conditions, other environmental factors at site level, such as vegetation cover, community composition, litter depth, and soil conditions, may also play a key role controlling the reproduction process of tree species (e.g., Dupuy and Chazdon, 2008; Montgomery et al., 2010; Promis and Allen, 2017). Therefore, small changes at site level during the establishment stage, which is one of the most vulnerable life stages of plants, can greatly influence seedling survivorship (e.g., Lloret et al., 2004). This, in combination with the inherent characteristic of seeds, can make reproduction success even more difficult. For example, plants with recalcitrant seeds (desiccation sensitive seeds) have a short viability span window (few weeks to a few months) for germination because of seed sensitivity to changes in site temperature and humidity (De Vitis et al., 2020; Brock et al., 2023). In addition, some species may present mast seeding strategies in response to climatic cues, meaning that changes in local climatic conditions may reduce the frequency of favorable signals, severely limiting the productions of seeds, leading to negative consequences for tree recruitment and forest dynamics (Bogdziewicz, 2022). Under these circumstances, site environmental conditions will have a pivotal effect on seeds germination once favorable climatic signals trigger the production of seeds.

The Mediterranean zone of Chile has been severely degraded due to the historical change in land use to agriculture, forestry, and urban development (Schulz et al., 2010), leading to large expanses of habitat loss and fragmentation (Fuentealba et al., 2021). Furthermore, models predict that this area could experience a temperature increase of 2°C and a decrease of precipitation (Araya-Osses et al., 2020; Vicuña et al., 2021). These factors combined (i.e., degradation and climate change) are becoming a threat to local biodiversity due to the synergetic effects on species extinction, which in turn can negatively impact the long-term viability of current ecosystems (Muñoz-Sáez et al., 2021). This is of particular concern because this part of the country is considered a global biodiversity hotspot due to high endemism of plant species (Martinez-Harms et al., 2021). Specifically, this hotspot has several endemic tree species with recalcitrant seeds such as Pouteria splendens (A.DC.) Kuntze, Beilschmiedia berteroana (Gay) Kosterm, Beilschmiedia miersii (Gay) Kosterm, Citronella mucronata (Ruiz et Pavón) D.Don, and Cryptocarya alba (Molina) Looser (Figueroa and Jaksic, 2004, Henriquez et al., 2012, Magni et al., 2023), which are already showing signs of recruitment limitations (e.g., Sotes et al., 2018; Magni et al., 2022, 2023).

The genus Beilschmiedia, within the Lauraceae family, is of particular interest because it is mostly distributed throughout the tropics (Nishida, 1999). However, in Chile this genus is distributed in a subtropical zone that corresponds to a Mediterranean climate, where climatic conditions are harsher, with lower water availability and a warmer and dryer summer season (Fernández and Morales, 2016). Beilschmiedia miersii, which is the species from the genus that has the most northern distribution in Chile, is a shade tolerant tree with a maximum height of approximately 25 meters (Henríquez and Simonetti, 2001). Previous dendroecology studies by Venegas-González et al. (2023) show that the age at different populations of B. mierssi can range from 34 to 348 years, with an average age among all sites of 114 years. Fruits are drupes (4 cm long, 2–3 cm wide), recalcitrant, with an unknown disperser (Henríquez and Simonetti, 2001; Marticorena and Rodríguez, 2001). B. miersii is usually associated with other species with recalcitrant seeds such as C. alba. It has been reported that B. miersii shows no seed production and a lack of seedlings in several sites during the last few years, a phenomenon that is not observed with the species Cryptocarya alba (although low germination in site is observed) (Magni et al., 2022; Fernández et al., 2023). Beilschmiedia species are known to show mast seeding processes; however, this has been hardly studied in Chile due to the lack of long-term data on seed fall (Yukich et al., 2023). Thus, B. miersii could be experiencing seeding and recruitment limitations due to the current megadrought, 13 consecutive years with rainfall deficits between 25 to 45%, affecting the central area of Chile (Venegas-González et al., 2023).

Although it is expected that the El Niño Southern Oscillation may still produce wetter years over central Chile (Cai et al., 2020), potentially triggering seeding production by B. miersii, there is scarce information on the current capacity of soil conditions for seed germination and recruitment, or other additional site-scale factors that may impede the first stages of plants establishment. Previous studies have shown that B. miersii germination and survival rates respond positively to leaf litter depth, litter water content, and canopy cover (Kremer et al., 2019), and also that herbivory and seed predations by small mammals and free-range cattle are prevalent limiting factors on the species recruitment (Morales, 2015a). Nevertheless, no studies have analyzed potential soil restriction induced by climate change on this species. Droughts can modify soil physical and biological process, leading to mineral stress, soil toxicity and accumulation of allelopathic compounds (Lynch and St Clair, 2004; Brevik, 2013; Zhong et al., 2023), which may restrict seed germinations and growth even if precipitations regimes are adequate 180 mm to 419 mm for B. miersii according to Becerra et al. (2004).

However, assessing potential soil restrictions for B. miersii establishment is very challenging due to the lack of recent seeds on-site and the short-term viability of seeds collected from previous years. Indeed, this is a common issue for long-lived laurifolious tree species that share their habitat with B. miersii, as they also produce recalcitrant seeds with limited seasonal availability, which has been exacerbated by increasing droughts driven by climate change (Magni et al., 2022). An alternative approach is the use of a surrogate species to evaluate if the soils of current B. miersii populations present unfavorable conditions for seed germination and growth. Surrogate species are selected substitutes used to represent the ecological responses of one or more species to environmental variables, particularly when direct assessment of the target species is impractical or constrained (Caro and O’Doherty, 1999). For the case of forest soil assessment, surrogate species should be able to quickly respond to several chemical and physical characteristics, including pH, salinity, organic matter, toxic elements, nutrients and texture (Schoenholtz et al., 2000).

Lens culinaris is an annual legume species belonging to the Fabaceae family, originally domesticated in the Eastern Mediterranean, but currently widely grown as a crop in Mediterranean, sub-tropical and temperate climates (Khazaei et al., 2016). Lens culinaris has been increasingly used as a bioindicator species for environmental analysis over the last decade. For example, Ostroumov (2016) used L. culinaris seedlings to test the toxic effects of surfactants on plant growth. Other authors have used L. culinaris as a bioindicator to assess metabolic toxicity induced by pesticides (Salazar and Maldonado, 2019; Salazar and Quintero, 2020; Salazar and Quintero, 2021). Salazar and Maldonado (2020) analyzed the effects of a widely used disinfectant on L. culinaris to evaluate its potential as a bioindicator for toxicology assessments. De Silva et al. (2022) used L. culinaris to evaluate the effects of microplastics on germination and seedling growth. Salazar and Vega (2023), evaluated the response of L. culinaris to the widely used medication acetaminophen, finding inhibitions of root development and cell abnormalities. Belasri et al. (2024) used L. culinaris to assess the effect of landfill leachate on vegetation, finding evidence of metabolic effects that reduced germination rates of the bioindicator species. These studies show that L. culinaris is highly sensitive and responds quickly to environmental factors, making it a suitable surrogate to evaluate potential limiting factors of soils for seed germinations. Nevertheless, as many legumes, L. culinaris has nitrogen-fixing capabilities making its seeds more tolerant to growth in nitrogen-restricted environments, which should be accounted for when interpreting the results.

In this work, we aim to test if the soil and other specific conditions from the sites currently holding populations of B. miersii have any limitations for germination and establishment of seeds, using L. culinaris as a bioindicator surrogate species. In particular, we are interested in assessing the direct effects of soil nutrients, and indirect effects of vegetation structure and historic changes of climatic patterns on the capacity of soils to provide suitable conditions for seed germination and growth. To determine if there is any restriction for germination in sites dominated by B. miersii we collected field data and soil samples from 15 B. miersii population sites, using these samples to run germination tests under controlled laboratory conditions. We hypothesize significant differences in germination and seedling growth rates between soil origins, which will be mainly explained by differences in soil available nutrients associated with changes in climatic patterns depending on site regions.

Table 1 shows a descriptive summary of the data collected in the 15 analyzed sites, including canopy cover, soil characteristics, plant species richness and climatic variability. All sites corresponded to forest patches, with an average canopy cover of 75.6% and 14.3 species per site. However, both the canopy cover and species richness show important variances between sites. The site showing the lowest canopy cover was Rincón, with 49.3%, while the one showing the highest cover was Árbol, with 91.1%. Species richness ranges from six species in Catapilco and Recreo, to 26 in Pangue. Our data revealed that the oldest trees were concentrated at Tantehue (center) and Aguas Claras (coast), with average ages of 186.8 and 152.9 years, respectively. The youngest trees on average were located in the center at the Colliguay and Pangue sites with an average age of 48.8 and 49.5 years, respectively. Interestingly, there was little variation in ages at site level in the center regions in comparison with the sites located toward the coast (Figure 2). In terms of sites region, we only find significant differences for species richness and precipitations (Wilcoxon rank-sum test p = 0.012 and p = 0.018, respectively, Table 1).

While we did find significant differences in germination and growth rates among soil origins, we did not find clear signs of germination limitations for any of the soils from the analyzed sites. Even though we were using a surrogate species, instead of seeds of B. miersii, our results suggest that all soil origins remain capable for recruitment, despite the observed variations in soils and sites conditions. These results were somewhat unexpected, given the lack of recruitment observed in the field at several sites, which was a pattern we were anticipating to also find in our laboratory experiment. An explanation for this may be related to the use a surrogate species (i.e., Lens culinaris) that requires different germination conditions that is also known to cope with limitation of nitrogen by fixing this nutrient from atmospheric sources. Nevertheless, we did not find nitrogen restrictions in any of the analyzed soils. Thus, other environmental limitations must be operating on sites, such as inadequate vegetation cover, harsh climatic conditions and increased herbivory, which were obviously not an issue at the laboratory setting.

We observed significant differences in the four analyzed variables (germination rate, germination speed, hypocotyl length, and radicle length) across different soil origins. Additionally, we identified specific site parameters associated with these variations, indicating that certain environmental factors at the site level play a crucial role in seedling recruitment. Yet, germination rates were relatively high, and mortality very low in all analyzed soils, indicating that the recruitment problems may be due to processes happening at seedling stages. Indeed, we were able to find some seedlings of B. miersii in 4 out of the 15 sites, ranging from 28 to 833 individuals per hectare, during our field work, but we did not see any evidence of seedling survival to sapling stage. Previous works have found that herbivory by exotic small-mammals [European rabbit (Oryctolagus cuniculus)] and dry summers may play a key role precluding the transition from seedlings to saplings in other mediterranean-type ecosystems (MacDougall et al., 2010; Granda et al., 2014), and the same effects has been also found for the Chilean Mediterranean ecosystem (Fleury et al., 2015). Thus, the same pressures could be precluding the recruitment of B. miersii on-site. Interestingly, we found that vegetation cover on-site was related to germination rates on the laboratory test, which may imply that vegetation cover could not only be relevant for decreasing water stress during the dry season but may also generate specific soil conditions that increases seed germination (Aponte et al., 2010).

The GSI tended to be higher for soil origins located in the coastal area, which could be attributed to variations on organic matter, canopy cover, species richness, nitrogen, and climatic conditions. However, among the aforementioned factors, only organic matter and nitrogen have been directly linked to germination speed changes according to the literature (Turk et al., 2004; Abazarian et al., 2011; De Souza Marinke et al., 2019; Cakir and Ceyhan, 2021). These two factors were relatively higher at the coastal sites, which may explain the differences on germination speed index (GSI). Nevertheless, our resulting explanatory model for GSI shows that organic matter is negatively related with GSI when all soil origins are considered, which may look counterintuitive. A deeper exploration of our data reveals that within the central sites there is a negative correlation (r = −0.57) between organic matter and nitrogen, which is not observed in coastal sites (r = −0.08). This difference may imply a differential effect of climatic change on these sites depending on their region. We argue that the decrease of precipitation and minimum temperature could cause the accumulation of organic matter and lower nutrients availability in soils due to reduced decomposition rates (e.g., Kirschbaum, 1995; Butenschoen et al., 2011). This process may be less pronounced in coastal areas, as there is a constant flow of humidity from the ocean that controls temperature and increases soil water content. Temperature is documented to have an influence on organic decomposition (e.g., Dwevedi et al., 2017), potentially increasing the decomposition rates in the hotter inland sites, but the dryer conditions of these areas may be limiting the decomposition processes. Despite the relationship observed between germination speed index and region, as well as the potential influence of certain independent variables, we acknowledge that our sample size is insufficient to draw definitive conclusions about these trends.

Both hypocotyl and radicle mean length showed significant differences among soil origins but not between regions. Patterns of hypocotyl and radicle growth were tightly related (r = 0.89), suggesting that these variables are responding to the same factors. Indeed, results from the mixed models show that both variables respond to the same factors (e.g., pH, EC, organic matter, potassium) but hypocotyl length also relates to site vegetation cover. Factors known to have an effect on hypocotyl growth include light intensity, temperature, water content, nutrient availability and organic matter (e.g., Xiong and Nilsson, 1999; Wang and Shang, 2020). However, in our study all these factors were controlled, except for soil nutrients and organic matter. Thus, the variability in nutrients such as nitrogen, phosphorous and potassium among sites may be a key factor explaining the differences in hypocotyl and radicle growth (Zhang and Wienhold, 2002; Abd El-hady et al., 2022). Nevertheless, contrary to what may be expected, potassium shows a negative effect on hypocotyl length. There is some evidence of a potential negative effect of an excess of potassium on plant growth which is mediated by modifications of the nutrient uptake processes (Xu et al., 2020, Papadakis et al., 2023). Our experimental setting had soils with relatively high concentrations of potassium; therefore, the excess of potassium could be a plausible explanation. However, this potential mechanism cannot be directly applied to B. miersii or other native species, as these have evolved to growth in naturally occurring high-potassium concentration soils.

Contrary to the prevailing notion that organic matter enhances plant growth, our study revealed a negative association between organic matter content and growth parameters (hypocotyl and radicle elongation). This seemingly paradoxical finding corroborates previous reports documenting the detrimental effects of organic matter on plant growth under specific conditions (e.g., Xiong and Nilsson, 1999). Nutrient depletion and the presence of allelopathic compounds, particularly polyphenols derived from evergreen litter, are potential mechanisms underpinning this negative relationship (Lopez-Iglesias et al., 2014). In addition, the potentially decreased availability of nutrients in soils due to slower decomposition rates can also influence growth (e.g., Kirschbaum, 1995, Butenschoen et al., 2011). While these factors could explain the negative relationship between organic matter and growth, this difference is more likely due to a positive relationship between organic matter and water retention (Lal, 2020). High moisture levels can hinder plant growth, for example, by reducing oxygen availability (Lacroix et al., 2021). In this study, the use of petri dishes without drainage, combined with high organic matter content in some soil samples, could have led to excessive moisture, thus causing the observed differences. In any case to have a conclusive answer field experiments will be needed.

Our results revealed a positive relationship between canopy cover and hypocotyl length. This finding is likely due to the close link between canopy cover and soil properties, particularly increased availability of nutrients like nitrogen and phosphorus (Gallardo, 2003). These nutrients have been shown to concentrate under tree canopies, with studies reporting contributions to the soil ranging from 49 to 90% (Prescott, 2002). Consequently, higher canopy cover could translate to potentially higher nutrient concentrations (N and P), which, due to their known positive impact on growth, can potentially influence hypocotyl length. This hypothesis appears to be supported by our data, as we observed a positive correlation between hypocotyl length and nitrogen levels (r = 0.57).

In terms of pH, results suggest a negative relationship between pH and hypocotyl and radicle length, meaning that there decrease in length when there is an increase in acidity of soil. Studies in other species such as red clover (Trifolium pratense L.) show that pH can have an impact on radicle length (Agić et al., 2009). However, it will depend on the tolerance range of the target species. Lens culinaris are tolerant to a wide range of soil pH, from moderately acid to slightly alkaline (Panuccio et al., 2022). The lentil cultivar we used in our study seems to present better growth responses in slightly acidic soils, but it did not show germination and growth problems in any of the assessed soils. We believe that the observed trends were perhaps explained by an interaction with other variables that were not studied.

The analyzed soil samples in this study do not show major restrictions for low recruitment of the surrogate species. Even the sites that have the worst performance still show reasonable germination and mortality rates, without major barriers to growth in the early stages of establishment. However, the main question of our study is whether the sites have the conditions to promote the recruitment of B. miersii. Most of these sites do not show evidence of successful recent regenerations as the youngest B. miersii individual we managed to find was dated to be 15 years old (Figure 1). From the 15 sites analyzed in this study, only four present seedlings, but no site had saplings. Interestingly, the age distribution of sites in the center was more homogeneous than in the coast and the four sites where seedlings were present were in coastal regions. These results are consistent with those of other species in the genus, such as Beilschmiedia tawa, where a lack of recruitment was also observed, and the youngest individual found was 45 years old (Morales, 2015a).

These results imply that other factors not analyzed in this study may explain the lack of B. miersii recruitment observed in the field, such herbivory, seed removal or specific microsite conditions. Seeds of B. miersii have short viability, and high humidity needs, especially after seeding in periods of drought. In addition, seeds require a leaf litter cover to secure germination and establishment (Becerra et al., 2004, Kremer et al., 2019). Studies have shown that B. miersii recruitment is highly dependent on regular water availability, with watering periods every two to 3 weeks to favor germination and promote the survival of seedlings. The amount of water in controlled studies have varied from 180 mm (dry year) to 419 mm (wetter year) (Becerra et al., 2004), and 502 mm (Kremer et al., 2019). The climatic data show that all the sites in our study have experienced a decrease in precipitation in the last 20 years (Table 1; Supplementary Figure S1). This reduction is significantly more intense during the first and second quarter of the year with an average decrease of 7 mm and 256 mm, respectively. A small decrease during the first quarter can be seen as marginal, however, is a critical time for the germination of B. miersii, as changes in humidity can have huge effects on the viability of recalcitrant seeds (Wen and Cai, 2014). Regarding litter, both studies highlight the importance of litter in successful germination, with a litter depth greater than 5 cm producing the best results (Kremer et al., 2019). However, it is important to note that in the present study, only 27% of the sites had a litter depth greater than 5 cm, which can also help to explain the lack of recruitment (Table 1).

We did not find any clear limitations, at least at the soil level, that could hinder germination and plant growth at the studied sites. B. miersii has recalcitrant seeds that have special requirements that are currently not met at all sites. The main current issues appear to be historical changes in climate variables (e.g., precipitation) and herbivory. Changes in other climatic variables, such as minimum and maximum temperature, can also cause disruptions in several processes and mechanisms that can exacerbate recruitment problems, and therefore are key for further analysis. Despite the limitation of using L. culinaris as a surrogate species, whose soil and water requirements may differ from B. miersii, it remains a valuable indicator of potential soil constraints for plant establishment. If this less tolerant species struggles to germinate and establish, it’s unlikely that native species like B. miersii would fare any better.

While this study employed a bioindicator species to assess soil limitations for B. miersii seed establishment, future studies in natural settings should prioritize native species, particularly B. miersii, to both minimize invasive risks and directly explore its recruitment strategies under shifting climatic conditions. In addition, coarse-scale climatic data, while valuable, lacks site-specific detail. To elucidate annual trends, sensor-driven research at site level on abiotic variables like temperature and soil moisture is essential to determine if the minimum conditions for recruitment are being met. Due to the complex interplay of different variables, future studies incorporating ecological modeling could help to evaluate the changes in these variables and their impacts on the viability of B. miersii dominated forests.

In conclusion, our study did not identify significant soil-related barriers to the germination and early growth of B. miersii, and the species’ recruitment appears to be influenced by other factors such as historical climate changes and herbivory. In Chile, efforts have been made to protect streams and gullies, which are the habitat of this species. However, changes in environmental factors, such as water availability, threaten their long-term viability. If this scenario continues, the species may face significant challenges. The use of a surrogate species highlighted potential soil constraints, but future research should focus on B. miersii itself to better understand its specific requirements and challenges. Detailed, site-specific climatic data and ecological modeling will be crucial in addressing the complex variables affecting B. miersii recruitment and ensuring the sustainability of its populations under changing environmental conditions.