Squirrels reduce post-fire regeneration potential in serotinous pines

Serotiny is a key trait for population persistence in fire-prone ecosystems, allowing species to accumulate an aerial seed bank until fire triggers seed release. However, serotinous cones provide a food source for specialized pre-dispersal seed predators, potentially counteracting the benefits of serotiny. These predators can also exert selection pressures on pine cone traits that could further reduce the size of the canopy seed bank. Therefore, we hypothesized that in populations of serotinous species, seed predation reduces the size of the canopy seed bank not only through the ecological effect of cone removal but also through negative selection of serotiny and increased seed defences (which leads to a reduction in seed production). To test this, we sampled six sites dominated by mature Pinus halepensis forests (three with European red squirrels Sciurus vulgaris and three without) and conducted a literature review for further validation. Our findings indicate that, in populations with squirrels, pre-dispersal seed predation reduces the canopy seed bank by approximately 65 %. This decline in the canopy seed bank could compromise population regeneration after fire, which is particularly concerning in the context of increasing fire activity in the Mediterranean basin. Squirrels reduce post-fire regeneration potential in serotinous pines

Serotiny is a key trait for population persistence in fire-prone ecosystems, allowing species to accumulate an aerial seed bank until fire triggers seed release. However, serotinous cones provide a food source for specialized pre-dispersal seed predators, potentially counteracting the benefits of serotiny. These predators can also exert selection pressures on pine cone traits that could further reduce the size of the canopy seed bank. Therefore, we hypothesized that in populations of serotinous species, seed predation reduces the size of the canopy seed bank not only through the ecological effect of cone removal but also through negative selection of serotiny and increased seed defences (which leads to a reduction in seed production). To test this, we sampled six sites dominated by mature Pinus halepensis forests (three with European red squirrels Sciurus vulgaris and three without) and conducted a literature review for further validation. Our findings indicate that, in populations with squirrels, pre-dispersal seed predation reduces the canopy seed bank by approximately 65 %. This decline in the canopy seed bank could compromise population regeneration after fire, which is particularly concerning in the context of increasing fire activity in the Mediterranean basin.

pre-dispersal seed predation, plant-animal interactions, fire ecology, fire traits, evolutionary ecology, Pinus halepensis, Sciurus vulgaris.

Fire has been part of the natural dynamics of many ecosystems for millions of years (Pausas and Keeley, 2009). Therefore, plant species inhabiting fire-prone environments have developed traits that allow them to survive or recruit after fires (He et al., 2011; Keeley and Pausas, 2022; Pausas and Lamont, 2022). One of the most recognized traits associated with fire is serotiny, which is the accumulation of a canopy seed bank in woody structures (cones or fruits) until the heat of a fire opens them, releasing the seeds (Lamont et al., 1991a; 2020). In environments where return intervals of crown fires fall between the age of reproductive maturity and the plant life span, serotiny maximizes the amount of seeds available when recruitment conditions are optimal, i.e., post-fire, as resource availability is high and competition is low (Causley et al., 2016; Lamont et al., 2020; best-bet strategy sensu Pausas et al., 2022). Therefore, variations in serotiny can influence the structure and dynamics of ecosystems, making it essential to understand the factors determining its prevalence. Even though it is well known that fire regimes shape intraspecific serotiny variability in many species and regions (Borchert, 1985; Cowling and Lamont, 1985; Gauthier et al., 1996; Givnish, 1981; Groom and Lamont, 2011; Hernández-Serrano et al., 2013; Lamont et al., 2020; Radeloff et al., 2004), there is still considerable variation in serotiny within populations and among populations with similar fire regimes (Benkman and Siepielski, 2004; Hernández-Serrano et al., 2013; Muir and Lotan, 1985; Tapias et al., 2004). This indicates that factors other than fire likely contribute to shaping this trait. One disadvantage of storing a canopy seed bank is that seeds become a year-round available food source to seed predators (Lamont et al., 1991b; Talluto and Benkman, 2014). This increased exposure of cones may increase pre-dispersal seed predation and thus reduce tree fitness; that is, predation can select against serotiny (Benkman and Siepielski, 2004; Talluto and Benkman, 2013). Therefore, the spatial variability in serotiny could be mainly due to the strength of these two opposing selective forces, fire and pre-dispersal seed predation (Talluto and Benkman, 2013; 2014). In short, although we have plenty of evidence that fire regimes shape serotiny, the potential of biotic interactions, such as seed predation, influencing the degree of serotiny and the size of the seed bank has received limited attention. Pre-dispersal seed predators also influence the evolution of pine cone characteristics, making them more resistant to being eaten (Benkman, 1995; Parker and Benkman, 2020; Siepielski and Benkman, 2007; Smith, 1970; 1975). This greater investment in defensive structures implies a reduction in the relative seed production per cone (Benkman, 1999; Mezquida and Benkman, 2005; Parker and Benkman, 2020; Smith, 1970), contributing to reducing the seed bank and, thus, potentially post-fire pine regeneration (Siepielski and Benkman, 2008; Talluto and Benkman, 2014). Although this cost should be particularly significant for serotinous species (as they primarily depend on their stored canopy seed bank for persistence), little is known about the combined impact of reduced serotiny and fewer seeds per cone as a result of seed predation. In this study, we aim to understand the implications of the pre-dispersal seed predation on post-fire regeneration potential in a serotinous pine. We hypothesize that pressure exerted by seed predators reduces the canopy seed bank, and consequently the post-fire regeneration potential of populations, and that this reduction is mediated by (i) the direct ecological effect of cone removal (and therefore, seeds); (ii) negative selection against serotiny, which reduces the frequency of highly serotinous individuals in populations; and (iii) an increase in seed defences, which leads to a decrease in the number of seeds per cone. We tested this hypothesis in the iconic and widespread Mediterranean pine Pinus halepensis Mill. (Aleppo pine), a species whose reproduction is tied to the fire regime (Guiote and Pausas, 2023; Hernández-Serrano et al., 2013), and its major pre-dispersal seed predator, Sciurus vulgaris (European red squirrel) . We first explore, based on an extensive field sampling, the mechanisms that can lead to a reduction in the canopy seed bank by analysing the effects of defensive traits of pine cones, serotiny degree, and crop size on predation pressure at level of individual trees (within a population; Fig. 1a). Then we verify whether the patterns found are consistent in explaining the variation in seeds defences, serotiny degree, and size of the canopy seed bank among populations by comparing populations under contrasting seed predation pressures (regional analysis; Fig. 1b). Finally, we validated the relationship between serotiny and squirrels with an independent dataset. Specifically, we searched the literature for estimations of serotiny and combined them with records of squirrel’s presence in P. halepensis forests across the Mediterranean basin (geographical analysis; Fig. 1c). Figure 1: Conceptual diagram illustrating the working hypothesis about the impact of seed predation by squirrels on the size of the canopy seed bank in a serotinous pine species. Trees within a population with larger crop sizes, higher serotiny levels, and weaker cone seed defences are more likely to be preferentially harvested by squirrels, leading to a reduction in tree fitness (local scale; panel a). This selection pressure is reflected in lower serotiny levels (validated at regional and geographical scales; panels b and c) and increased seed defences (validated at the regional scale; panel b) in populations where squirrels are present (panel b) or where their density is higher (panel c).

Study species P. halepensis is a common and widespread serotinous tree species in the Mediterranean basin and is dominant in the eastern Iberian Peninsula (Quézel, 2000). Its level of serotiny is controlled by fire regimes (Hernandez-Serrano et al., 2013; 2014; Ne’eman et al., 2004); however, to what extent pre-dispersal seed predation modulates the serotiny level for a given fire regime remains to be evaluated. European red squirrels are widely distributed in mature P. halepensis forests of the Iberian Peninsula (Purroy, 2017). They are efficient pre-dispersal seed predators of conifer cones and negatively affect the abundance of other P. halepensis pre-dispersal seed predators such as crossbills (Mezquida and Benkman, 2005). Consequently, this squirrel species is considered the main selective agent influencing cone evolution in P. halepensis populations of the Iberian Peninsula. Field sampling We selected six sites dominated by mature P. halepensis forests in two contrasting seed predation environments. In three of these sites, squirrels were absent: two were located in Ibiza, a continental island about 90 km from the Iberian Peninsula and part of the Balearic Archipelago, Spain; and the third was in a relatively isolated population on the Iberian Peninsula. In contrast, squirrels were present—although at different densities—at the other three sites, all located on the Iberian Peninsula (Table S1). Site selection was based on distribution maps (www.vertebradosibericos.org; www.bdb.gva.es) and field observations of remains of consumed cones (cone cores) on the ground. We ensured that cone cores resulted from the feeding activity of squirrels and not of other vertebrate seed predators (such as rats or crossbills; following Rima et al., 2007). The current presence of squirrels in Iberian P. halepensis forests likely results from both a long-standing ecological association (dating back at least to the Middle Ages) and more recent recolonization following the forests’ 20th-century regeneration. On the other hand, the historical absence of squirrels in the Balearic Islands is well-documented (e.g., lack of fossil records), and thus it provides a unique opportunity for comparing the effects of the presence and absence of squirrels in relatively close distance and similar conditions. Identifying regions in the Iberian Peninsula where squirrels were absent was more challenging, as squirrels are very common and the available information is scarce (Rochat et al., 2014). However, squirrels depend on habitat quality and connectivity, so we located a relatively isolated P. halepensis population within the peninsula where squirrels were absent. To minimize the effect of fire regime (Hernandez-Serrano et al., 2013; 2014) and tree age (Martín-Sanz et al., 2016; Moya et al., 2008) on serotiny degree, all populations were located in sites with a similar fire regime (relatively frequent crown fires) and in mature stands with similar tree sizes (Table S2). Yet, the age of the trees and variations in fire regimes were considered in the analyses (see below). The selection procedure was based on the available fire maps, field evidence, and in the knowledge about the local relation between fire activity, altitude, and summer drought (Hernandez-Serrano et al., 2013; Verdú and Pausas, 2007). In the six sites, w e haphazardly selected pine trees (separated by at least 10 m, excluding trees with overlapping crowns ) and measured their size (diameter at breast height; DBH) and the distances to the two closest conspecific trees (we use the mean of these values; hereafter average distance). Next, using binoculars, we counted the number of cones from the last six cone cohorts and recorded whether they were open or closed (closed cones older than six years are uncommon in this pine species; Hernández-Serrano et al., 2013; Notes S1). For each tree, we estimated serotiny degree, predation pressure (in sites with squirrels), and the crop size for each cohort (Table 1, Notes S1). Finally, we collected at least six cones per tree from the most recent mature cohort, while avoiding malformed or damaged cones. For this, we climbed the pine crown and cut at least three terminal branches to represent the variability in cone traits within individuals. Table 1: Definition of the variables used in the six studied sites. | Pre-predation serotiny degree | (Closed grey cones in the canopy + grey cone cores at the base of the pine) / (closed and open grey cones in the canopy + grey cone cores at the base of the pine) | | Post-predation serotiny degree | Closed grey cones in the canopy / closed and open grey cones in the canopy | | Predation pressure* | Cone cores at the base of the pine / (closed cones in the canopy + cone cores at the base of the pine) | | Relative predation pressure | Predation pressure of the pine / mean value of predation pressure of the population | | Crop size* | Closed cones + cone cores at the base of the pine | | Canopy seed bank | Crop size x mean number of seeds per cone of the tree | * Cones were classified by cohorts (both in the canopy and at the base of the pine) as defined in the text. In each site, the 29-30 trees were sampled during spring-early summer (Table S1). One of the sites with a high density of squirrels (Serra; the focal site) was selected to explore the influence of seed defences, serotiny degree, and crop size on predation pressure at the individual level (Fig. 1a). For a squirrel, the rewards offered by a given tree often vary seasonally. For instance, a tree might produce numerous cones, making it highly profitable in the spring, but it could be poorly serotinous and therefore less rewarding during the autumn and winter. Thus, in this site, we sampled also during winter (77 trees) when the ripened cones have passed the critical period (summer) and some have opened. Consequently, in this focal site, we collected a minimum of six cones per tree in each sampling season. Cone trait measurements Cones of all six sites were taken to the lab for measuring morphological traits following the methodology in (Mezquida and Benkman, 2005; see more details in Notes S2). Additionally, we monitored the time required for each cone to open (in an oven; hereinafter ‘cone opening duration’ ), measured its width 45 minutes after opening (see detailed protocol in Notes S2 ), and calculated the difference between the width of the open and closed cone divided by the width of the closed cone as an estimate scale rigidity (a potential defensive trait; Notes S2). As in previous studies (for example, Parker and Benkman, 2020; Mezquida and Benkman, 2005), we also estimated the ‘seed-to-cone mass ratio’ as the total seed mass (the number of full seeds, i.e., excluding empty seeds, multiplied by the mean mass of individual seeds) divided by the total cone mass. This variable is considered a defensive trait in such a way that lower levels of this ratio indicate higher defence value as squirrels need to chew more woody tissue relative to the energy intake (lower feeding efficiency; Benkman et al., 2001; 2003; 2010; Mezquida and Benkman, 2005; Smith, 1970). Finally, we calculated mean values per tree for all measured traits to be used in the statistical analyses. Bibliographic review (geographical analysis) We searched the literature for articles where serotiny had been estimated in P. halepensis . We found four articles with data for sixty-one populations distributed in six geographical regions throughout the Mediterranean basin (Table S3). To test the influence of predation, we used georeferenced data of S. vulgaris presence obtained from GBIF database (GBIF.org, 2024). Then, using QGIS version 3.16.13 (QGIS Development Team, 2021), we set an influence area of 0.1-degree (approximately 10 km) radius around the studied populations and calculated the density of squirrel records per km 2 (Table S3). In all the studies considered, the serotiny degree was estimated following the same criteria (Notes S3). Statistical analyses All climatic variables were obtained from the CHELSA V.2.x database (a gridded map with 0.0083° resolution and average values for the 1981–2010 period; Karger et al., 2021). Statistical analyses were performed using R software version 4.3.1 (R Core Team, 2023). Generalized linear mixed models (GLMMs) and zero-inflated gamma models (ZIGs) were fitted using the ‘glmmTMB’ package (Brooks et al., 2017) while linear mixed models were fitted using the ‘lme4’ package (Bates et al., 2015). Models were constructed through the sequential addition of variables, using Akaike’s Information Criterion (AIC) to guide model selection. The ‘DHARMa’ package (Hartig, 2022) was used to assess model diagnostics via residual analysis, and the ‘ggeffects’ package (Lüdecke, 2018) was used to compute predicted values. Within population (local analysis) In the first sampling (winter), we found that 94.8 % of the depredated cones were serotinous (grey cones), while in the second sampling (early spring), the majority of them (81.1%) belonged to the first mature cone cohort (orangish-brown cones). Therefore, we studied seed predation by fitting a model for each sampling period. To identify the most important variables in explaining relative predation pressure at the tree level, we fitted a ZIG model with the log link function. In the zero-inflated part of the model (i.e., probability of lack of seed predation), we included crop size from the target cohort in each sampling (grey cones in the first case, orangish-brown cones in the second), as it is crucial in the selection of individual plants by seed predators (Jordano, 2000). In the conditional part (i.e., predation pressure in depredated trees), we included cone traits, pre-predation serotiny degree, and their interactions. Additionally, we performed pairwise linear regressions between the relative predation pressure and cone traits or serotiny degree (standardized values to a mean of 0 and a variance of 1). Among populations (regional analysis) We tested whether serotiny at the individual level differed among the six sites depending on squirrel presence. To do so, we used a GLMM with a Betabinomial error distribution and the logit link function. Serotiny can be characterized based on two components; the proportion of serotinous cones and the time these cones remain closed (temporal component of serotiny; Lamont, 2020). We calculated post-predation serotiny degree (Table 1) using the two oldest studied cone cohorts for these analyses because it is the estimation that best reflects the temporal component of serotiny. In any case, we also made the analysis with all serotinous cone cohorts together. We included squirrel presence as predictor and site as a random factor. We also included proxies for the most relevant factors explaining serotiny in this species: fire regime (as fire proneness), tree size (as DBH) and density effects (as average distances to the two closest trees; Goubitz et al., 2004). In Mediterranean conditions low summer precipitation is often related to high fire-proneness (Keleey et al., 2012) and, specifically in our study area, it is correlated with the area burned (Pausas, 2004; Pausas and Paula, 2012). Thus, we used the mean monthly precipitation of the warmest quarter as a proxy for fire proneness. We also considered an alternative proxy, the cumulative burned area over a 37-year period within a 10 km buffer around the sites (following Guiote and Pausas, 2023). To investigate the selection pressures of depredation on the canopy seed bank by the increase in seed defences, we began by testing whether the seed-to-cone mass ratio differed between populations with and without squirrels. This ratio has previously been identified as a defensive trait against squirrels and is known to correlate with the number of seeds per cone (Mezquida and Benkman, 2005; Parker and Benkman, 2020). Next, we examined whether the number of seeds per cone was different depending on squirrel’s presence. To do this, we fitted two LMMs in which the response variables were the seed-to-cone mass ratio and the total number of seeds per cone. In both we included squirrel presence as predictor and site as random factor. The size of the cone and the thickness of the scales are traits related to the amount of seeds (Mezquida and Benkman, 2005), so we also included cone length (highly correlated with width; r = 0.73, p = <0.001) and scale thickness when modelling the number of seeds per cone. Also, since seed production can be conditioned by factors such as water availability and intraspecific competition (Ayari et al., 2011; Moya et al., 2007), we tested climatic variables related to precipitation and humidity (Table S4) in addition to tree size (DBH) and the average distance to the closest trees. We excluded from these analyses cones with missing values for any of the traits (less than 6 % for seed-to-cone mass ratio, and less than 1 % for the number of seeds per cone). Since we used tree mean values, we excluded trees with fewer than three cones with data for seed-to-cone mass ratio (11 cases) and trees with fewer than three cones with data for the amount of seeds per cone (5 cases). Additionally, we checked whether the other cone traits measured differed among sites with and without squirrels by fitting LMMs. We use each cone trait as a response variable, squirrel’s presence as predictor, and site as random factor. To corroborate the impact of predation pressure on the canopy seed bank and estimate its magnitude, we tested whether the size of the canopy seed bank differs among populations depending on squirrel presence. For this, we fitted a GLMM with a Gamma error distribution and the log-link function. We included annual cone production (crop size of the more recent mature cone cohort; in Serra, we consider cone production from the spring sampling, as sampling at the other sites was conducted during spring-summer) and squirrel’s presence as predictor variables, and site as a random factor . In this case, we considered not only serotinous cones since mature cones from more recent cohorts are also depredated by squirrels and the seeds from these cones contribute to a certain extent to post-fire regeneration (Greene et al., 2024; Lamont and Enright, 2000). Additionally, to assess how predation pressure impacts each of the two components of the canopy seed bank (number of cones and number of seeds per cone), we fitted two models with the same structure and error distribution family as our previous analysis. In this case, to estimate the canopy seed bank, in the first model, we hold the number of seeds per cone constant at its mean value, while in the second model, we keep the number of cones constant at its mean value. We then determine the effect size of predation pressure on each trait by calculating Cohen’s d between the levels of squirrel presence (No, Yes) in both models. Finally, we examined whether the association between serotiny and seed-to-cone mass ratio (i.e., seed defences) differed between populations with and without squirrels. Finding a stronger association at sites with squirrels could indicate that both traits are being jointly selected due to predation, potentially counteracting the negative effects of seed predation on serotiny (Parker and Benkman, 2020). In pine species where serotiny is a qualitative trait (i.e., individuals produce predominantly serotinous or non-serotinous cones), such association has been noted for serotinous individuals within a population (Parker and Benkman, 2020). Therefore, since serotiny is a quantitative trait in our species (Nathan et al., 1999), a stronger association might only be evident in the most serotinous individuals of the population. For this reason, we categorized trees as strong versus intermediate-weak serotinous (above and below the 75th percentile, respectively). We considered the interaction between this categorization and the seed-to-cone mass ratio as a predictor in the model. We fit a LMM with serotiny (arcsine-transformed) as the response variable and site as a random factor. This analysis was also performed using a ZIG model rather than the arcsine transformation, but the poor fitting (deviation from residual normality) made us use this transformation (no deviation of the residuals). Before constructing the model, we performed an exploratory analysis to assess potential relationships between serotiny and cone traits across all populations. This initial analysis identified covariates potentially associated with serotiny (independent of seed predation pressure by squirrels) for inclusion in the main model, enhancing its accuracy and accounting for possible confounding factors. Across regions (geographical analysis) Using the bibliographic review data, we tested the influence of predation pressure on serotiny across the Mediterranean basin. To do so, we fitted a LMM including serotiny degree (proportion of closed cones respect to total cones excluding the two youngest cone cohorts) as a response variable, squirrels’ density as a predictor variable, and site nested within study, as random factor. As in the regional analyses , we also took into account factors that can affect serotiny degree: fire proneness (as an indicator of fire regime) and the height of the trees. We use the arcsine square-root transformation of serotiny to fit the model. Within population Crop size significantly affected the relative predation pressure at both sampling seasons. The smaller the crop size, the lower the probability of being depredated. This is indicated by the negative relation between probability of not being predated and crop size in the zero-inflation part of the models (χ² = 8.63, p = 0.003, Table S5.1a; and χ² = 8.38, p = 0.004, Table S5.2a). The conditional part of the model for the winter sampling indicated that the most serotinous trees with a higher seed-to-cone mass ratio experienced the highest relative predation in winter (χ² = 7.05, p = 0.029, Table S5.1b; Fig. 2a). During early spring, scale rigidity was the most relevant variable in explaining the differences in relative predation pressure among trees, being trees having cones with softer scales preferred by squirrels (χ² = 9.35, p = 0.002, Table S5.2b; Fig. 2b). Figure 2: Fitted lines of the models testing the effects on trees’ relative predation pressure against: a) the interaction between serotiny degree and seed-to-cone mass ratio in winter, and b) scale rigidity in early spring. For the former (a), we provide fitted lines for the 25 th (P 25, blue) and 75 th (P 75, orange) percentile of serotiny. Lower values in seed-to-cone mass ratio and higher in scale rigidity are associated with enhanced seed defences as they reduce feeding efficiency for the squirrels. See Table S5 for the statistics. Pairwise relationships between relative predation pressure and cone traits, as well as serotiny degree, showed that, during the winter sampling, only serotiny had a positive effect. In early spring, both scale rigidity and seed-to-cone mass ratio exhibited a correlation with relative predation (negative and positive, respectively; Table S6). Among populations (regional analysis) Tree size (DBH) was the main explanatory variable of serotiny degree of the trees (negative relationship; χ² = 12.09, p = <0.001, Table S7a). After including this variable, the presence of squirrels still showed a significant negative relationship with serotiny, reducing serotiny degree by approximately 55% (on average) in the populations where they are present (χ² = 4.17, p = 0.04, Table S7a; Fig. 3a). When we considered all grey cone cohorts (instead of the oldest cohorts), the trend towards lower serotiny in populations with squirrels is maintained, although the effect becomes marginal (χ² = 2.58, p = 0.109, Table S8). Fire regime was not significant, regardless of whether we considered climate-based fire-proneness or cumulative burned area (Table S9). There were significant differences in seed defences (seed-to-cone mass ratio) between populations with and without squirrels. Locations where predators were present had seed defence values that were twice as high (i.e., a lower seed-to-cone mass ratio) on average compared to those without predators (χ² = 15.08, p = <0.001, Table S7b; Fig. 3b). As expected, this resulted in cones with fewer seeds in populations with squirrels (on average, 43.3% fewer seeds), after accounting for the variables related to the amount of seeds per cone (cone length and scale thickness; χ² = 7.64, p = 0.006, Table S7c; Fig. 3c). None of the environmental covariates tested (Table S4) were related to the number of seeds per cone, and no other cone traits differed between populations with and without squirrels (Table S10). The size of the canopy seed bank was significantly lower in sites with squirrels (64.5 % lower; χ² = 6.48, p = 0.012, Table S7d, Fig. 3d). Specifically, seed predation appeared to have a greater impact on the number of cones than on the number of seeds per cone (65.6 % and 34.4 % of the total reduction, respectively; Table S11, Fig. S3). Figure 3: Comparison between absence and presence of squirrels (“No” in blue; “Yes” in orange, respectively) in serotiny degree (closed-to-total cones of the two oldest cone cohorts; a), seed-to-cone mass ratio (b), seeds per cone (c), and size of the canopy seed bank (log-scaled y-axis; d). Coloured symbols represent raw data per tree, while black-bordered coloured symbols with error bars indicate model-predicted values with confidence intervals (see Table S7 for the statistics). Of all the cone traits measured, cone opening duration and scale rigidity were the only traits associated with serotiny across all populations (Table S12). When we considered the effect of cone opening duration, scale rigidity was no longer significant (Table S13). In populations where squirrels were present, serotiny was negatively related to seed-to-cone mass ratio (i.e., positively related to seed defences) in the strong serotinous trees (above the 75 th percentile; t = -2.022, p = 0.044, Table S14, Fig.4). This analysis was also performed using a zero-inflated binomial approach (instead of the arcsine-square root transformation), despite this approach exhibits a deviation from residual normality (Kolmogorov-Smirnov normality test, p = 0.01), the results were the same (Table S15). Figure 4: Fitted lines for the model testing the interaction between squirrel presence (“No” in blue, “Yes” in orange) and the serotiny categorization (Low- Intermediate Vs Strong) in explaining the association between serotiny degree and seed-to-cone mass ratio (inverse of seed defences; a) and scale rigidity (b). The left plots (Intermediate-Weak) include trees with serotiny degree below 75 th percentile, and the right plots (Strong) include those above this value. Serotiny values were back-transformed to generate the plot. The result of the model is shown in Table S11. Among regions (geographical analysis) Tree height and fire proneness were the most significant variables explaining the variability in serotiny degree among regions. Specifically, shorter trees and those from sites more fire-prone exhibited higher levels of serotiny. After accounting for these two factors, squirrel density remained significant, demonstrating a negative relationship with serotiny (χ² = 4.93, p = <0.026, Table S16, Fig 5), which is consistent with our regional analysis (Fig. 3a). As in the regional analysis, the results using untransformed serotiny values (Table S17) were similar, although this analysis exhibits a slight deviation from residual normality (Kolmogorov-Smirnov normality test, p = 0.01). Figure 5: Fitted line of the model testing squirrel density (squirrels’ records/km 2 of land) influence over serotiny variability (closed-to-total cones) across the Mediterranean basin. The model includes tree height and fire proneness as covariates (Table S15) . The plot was generated using back-transformed serotiny values.

Squirrels are voracious predators of pine seeds, whose direct and indirect effects contribute to an approximate 65% reduction in the canopy seed bank of P. halepensis. This decline is driven by the ecological effect of cone removal, the stronger selection pressure on the most serotinous individuals, and the increased investment in seed defences by trees in response to depredation (to the detriment of the number of seeds per cone). Squirrels reduce serotiny level For frugivorous and granivorous animals, although reward quality ultimately guides fine-scale foraging choices, resource abundance also plays a significant role in the overall decision-making process (Palacio et al., 2016). This is why crop size accounted for within-population variation in predation pressure (in both winter and spring; Table S5). This may also explain why, during winter, when the availability of newly ripened cones is reduced due to seed predation in spring and summer, the presence of serotinous cones retained in the canopy becomes a significant factor in explaining predation levels (Table S5.1b, Fig. 2a). This increased predation pressure on trees with higher levels of serotiny (local scale) is consistent with the lower level of serotiny observed in the populations with squirrels (regional scale; Table S7a, Fig. 3a), and with the decrease of serotiny degree in sites with higher density of squirrels across the Mediterranean basin (geographical scale; Table S15, Fig. 6). That is, the negative effect of squirrels on serotiny is observed at the three different spatial scales considered and using independent datasets. This negative relationship could simply be a direct consequence of the ecological effect of seed predation (the removal of serotinous cones). However, serotiny is a key and heritable trait that increases fitness in fire-prone environments (Castellanos et al., 2015; Keeley and Zedler, 1998). Therefore, the increased predation pressure on the most serotinous individuals (at least during winter) likely reduces their relative fitness, resulting in a decreased frequency of highly serotinous individuals in the next generation (i.e., low serotiny is favored). Considering this, it is reasonable to assume that among-population variation in serotiny levels (for a given environmental conditions) results from both the ecological impact of seed predation and its evolutionary consequences. Pines defend themselves against squirrels We found evidence of tree selection based on the seed-to-cone mass ratio during spring (Table S6), although this trait became insignificant in early spring once scale rigidity was included in the model (Table S5.2b, S5). This is probably because both traits are linked to feeding efficiency. Scale rigidity is associated with lignin content, which influences the compaction and hardness of cones (Moya et al., 2008; Smith, 1970), so higher rigidity likely decreases feeding efficiency. Similarly, the seed-to-cone mass ratio is related to feeding efficiency, as a lower ratio indicates a higher proportion of woody tissue that squirrels must chew through to access a given amount of food (Benkman, 1999; Benkman et al., 2001). Previous research indicates that squirrels prefer softer cones when available (e.g., Smith, 1970), and our spring sampling aligns with this pattern (Table S5.2b). However, we found no significant relationship between scale rigidity and predation pressure during the winter sampling. This absence of a relationship does not necessarily negate its role as a defensive trait, as it might result from a reduced variability in this trait once the cones become serotinous (after summer), making them uniformly hard and compact (Greene et al. 2024; Moya et al., 2008). The intrinsic positive association between serotiny and scale rigidity (Table S12), combined with the role of this trait as a defence against seed predation during spring (Table S5.2b), suggests that being potentially more serotinous (i.e., having recently matured cones with more rigid scales) may serve a protective function against seed predation during certain periods. However, when squirrels prey on serotinous cones, the impact on plant fitness is greater than when they target newly ripened cones since they contribute little to fitness in fire-prone environments; Goubitz et al., 2002). The result of the negative selection against trees with a higher seed-to-cone mass ratio is translated into lower values of this trait in populations with squirrels (Table S7b, Fig. 3b). Additionally, the number of seeds per cone was lower in populations with squirrels (approximately 43% lower, on average; Table S7c, Fig. 3c). Previous studies have reported a similar pattern of reduced investment in seeds in populations under seed predation pressure by squirrels. In some cases, it has been observed that squirrels select trees producing a higher number of seeds per cone (Parker and Benkman, 2020), while in other cases they avoid trees with a higher investment in woody cone tissue (larger cones or with relatively thicker scales; Mezquida and Benkman, 2005). In our study, the form of selection exerted by squirrels within a population appears to cause in the long term a decrease in the number of seeds per cone at a population level, as they select against trees with high seed-to-cone mass ratio within populations. However, in contrast to previous findings in P. halepensis (Mezquida and Benkman, 2005), we did not find evidence of tree selection based on cone size or scale thickness (Table S5 and Table S6). Although populations with squirrels had cones that were, on average, larger (52.58 mm vs. 48.68 mm) and had thicker scales (3.40 mm vs. 2.89 mm), these differences were not statistically significant (p = 0.37 and 0.22, respectively). These contrasting results may be due to differences in pine cone sizes between studies. Squirrels tend to select larger pine cones when the average cone size is relatively small, with this relationship reversing when cones are larger, within- and among populations (Mezquida and Benkman, 2005; Siepielski and Benkman, 2007). The pine cones in our squirrel-populated sites (mean = 55.43 mm) were smaller than in previous studies (Mezquida and Benkman, 2005; mean = 75.79 mm). We did not investigate the causes of these size differences, but this might explain the absence of size-based selection in our study. Whether the temporal disruption of the possible evolutionary trajectory towards an increase in cone size could be explained by, for example, the drought experienced by Mediterranean ecosystems in recent decades and its potential impact on cone and seed development (Ayari et al., 2011; IPCC, 2021; Moya et al., 2007; Vogel et al., 2021), remains to be explored. An alternative explanation could be the contribution of a secondary pre-dispersal seed predator such as crossbills ( Loxia curvirostra ) to the evolution of cone traits in our population (see Mezquida y Benkman, 2005 for a geographic selection mosaic on cone traits influenced by the presence and abundance of crossbill and squirrels). That is, it is possible that on our sites that lacked or had a low density of squirrels (Table 1), a secondary seed predator may have obscured the pattern of selection by squirrels on cone traits. In any case, as a response to predation pressure, populations with squirrels showed a lower number of seeds per cone, which may contribute to compromising post-fire regeneration of populations. Serotiny and seed defences association In a Pinus contorta population, serotiny and seed defences showed a positive association, suggesting that the most serotinous individuals invest more in defences, counteracting the selection exerted by squirrels ( Tamiasciurus hudsonicus ) against serotiny (Parker and Benkman, 2020). In P. halepensis, we found that the seed-to-cone mass ratio decreased with the increasing serotiny (i.e., seed defence positively associated with the degree of serotiny), but this only occurs among the most serotinous individuals, and in populations with squirrels (Fig. 4; Table S14). Since higher levels of seed predation cause stronger selection (Benkman 2013), the association between serotiny and seed-to-cone mass ratio is consistent with the observed enhanced predation pressure on the most serotinous trees, leading to defences being more intensely selected in these individuals. Considering these results, together with the lower values of serotiny observed in sites with squirrels, we conclude that, although serotiny and seed defences can be selected together as a response to seed predation by squirrels, this does not completely counteract the negative effect of predation on serotiny. Reduction of the canopy seed bank Predation pressure effectively reduced the canopy seed bank by approximately 65 % (regardless of the annual production of cones; Table S7d, Fig 3d), Our findings suggest that predation impacts both components of the canopy seed bank; the number of cones and the number of seeds per cone but the first was more affected (65.6 % and 34.4 % of the total reduction respectively; Table S10; Fig. S3). Although P. halepensis may retain some capacity for regeneration via non-serotinous cones (Daskalakou and Thanos, 1996; Ne’eman et al., 2004), the canopy seed bank remains it is his primary source of recruitment in Mediterranean fire-prone ecosystems (Goubitz, 2001; Goubitz et al., 2002; Ne’eman et al., 2004). So, while the 65% reduction in P. halepensis may not be an extreme (as has been observed in Pinus contorta populations in the Rocky Mountains, where red squirrels can reduce the canopy seed bank by up to 95%, severely limiting post-fire regeneration; Benkman, pers. comm.), it still represents a substantial loss of reproductive potential, particularly under suboptimal post-fire conditions (e.g., drought, poor soils, or competition; Pausas et al., 2002). Therefore, such a reduction could compromise both forest recovery and the long-term persistence of populations in frequently burned landscapes. Certainly, other factors such as seed size can influence seedling establishment and regeneration of populations, although the magnitude of the benefits is highly contingent on environmental conditions and species (Bladé and Vallejo, 2008; Daskalakou and Thanos, 2010). In any case, we did not find robust evidence of larger seeds in populations with squirrels (individual seed biomass = 1.59 mg vs. 1.10 mg in sites without squirrels; p = 0.23).

Serotinous species rely on canopy seed banks for post-fire recruitment, making it critical to understand the factors that regulate their dynamics. Increasing evidence suggests that climate change may lead to demographic collapse in these species (Agne et al., 2022; Enright et al., 2015; Enright and Agne, 2025; Souto-Veiga et al., 2022; Paneghel et al., 2024; Vincenzi and Piotti, 2014). Shorter fire intervals limit canopy seed bank development, while intensified drought conditions reduce the window for seedling emergence, delay reproductive maturity, and decrease seed production. These factors, combined with the pre-dispersal seed predation, which significantly depletes the seed bank, could severely threaten the future of populations in certain regions. This study, therefore, highlights the importance of closely monitoring serotinous populations in areas with abundant squirrels, as they could be especially susceptible to regeneration failure. Nevertheless, a reduced seed bank is not always detrimental; in certain Mediterranean areas, historic management practices have resulted in overly dense pine stands that constrain natural regeneration and biodiversity (Gómez-Aparicio et al. 2009). Thus, incorporating knowledge of seed predation effects into management and restoration planning, taking into account site-specific and ecological objectives, may be crucial for sustaining ecosystem resilience under future climate scenarios.

Agne, Michelle C., Joseph B. Fontaine, Neal J. Enright, and Brian J. Harvey. 2022. “Fire Interval and Post-Fire Climate Effects on Serotinous Forest Resilience.” Fire Ecology 18 (1): 22. https://doi.org/10.1186/s42408-022-00145-4. Ayari, A., D. Moya, M. N. Rejeb, et al. 2011. “Geographical Variation on Cone and Seed Production of Natural Pinus Halepensis Mill. Forests in Tunisia.” Journal of Arid Environments 75 (5): 403–10. https://doi.org/10.1016/j.jaridenv.2011.01.001. Bates, D., M. Mächler, B. Bolker, and S. Walker. 2015. “Fitting Linear Mixed-Effects Models Using Lme4.” Journal of Statistical Software 67 (1). https://doi.org/10.18637/jss.v067.i01. Benkman, Craig W., and Adam M. Siepielski. 2004. “A Keystone Selective Agent? Pine Squirrels and the Frequency of Serotiny in Lodgepole Pine.” Ecology 85 (8): 2082–87. https://doi.org/10.1890/04-0177. Benkman, C.W. 1995. “The Impact of Tree Squirrels ( Tamiasciurus ) on Limber Pine Seed Dispersal Adaptations.” Evolution 49 (4): 585–92. https://doi.org/10.1111/j.1558-5646.1995.tb02295.x. Benkman, C.W. 1999. “The Selection Mosaic and Diversifying Coevolution between Crossbills and Lodgepole Pine.” The American Naturalist 153 (S5): S75–91. https://doi.org/10.1086/303213. Benkman, C.W., W.C. Holimon, and J.W. Smith. 2001. “The Influence of a Competitor on the Geographic Mosaic of Coevolution Between Crossbills and Lodgepole Pine.” Evolution 55 (2): 282–94. https://doi.org/10.1111/j.0014-3820.2001.tb01293.x. Benkman, C.W., T. Parchman, A. Favis, and A.M. Siepielski. 2003. “Reciprocal Selection Causes a Coevolutionary Arms Race between Crossbills and Lodgepole Pine.” The American Naturalist 162 (2): 182–94. https://doi.org/10.1086/376580. Benkman, C.W., T. Parchman, and E. Mezquida. 2010. “Patterns of Coevolution in the Adaptive Radiation of Crossbills.” Annals of the New York Academy of Sciences 1206 (1): 1–16. https://doi.org/10.1111/j.1749-6632.2010.05702.x. Bladé, C., and V.R. Vallejo. 2008. “Seed Mass Effects on Performance of Pinus Halepensis Mill. Seedlings Sown after Fire.” Forest Ecology and Management, Large-scale experimentation and oak regeneration, vol. 255 (7): 2362–72. https://doi.org/10.1016/j.foreco.2007.12.039. Borchert, M. 1985. “Serotiny and Cone-Habit Variation in Populations of Pinus Coulteri (Pinaceae) in the Southern Coast Ranges of California.” Madroño 32 (1): 29–48. Brooks, M.E., K. Kristensen, K. Benthem van, et al. 2017. “glmmTMB Balances Speed and Flexibility Among Packages for Zero-Inflated Generalized Linear Mixed Modeling.” The R Journal 9 (2): 378. https://doi.org/10.32614/RJ-2017-066. Castellanos, M. C., S. C. González-Martínez, and J. G. Pausas. 2015. “Field Heritability of a Plant Adaptation to Fire in Heterogeneous Landscapes.” Molecular Ecology 24 (22): 5633–42. https://doi.org/10.1111/mec.13421. Causley, C.L., W.M. Fowler, B.B Lamont, and T. He. 2016. “Fitness Benefits of Serotiny in Fire- and Drought-Prone Environments.” Plant Ecology 217 (6): 773–79. https://doi.org/10.1007/s11258-015-0552-y. Cowling, R.M., and B.B. Lamont. 1985. “Variation in Serotiny of Three Banksia Species along a Climatic Gradient.” Australian Journal of Ecology 10 (3): 345–50. https://doi.org/10.1111/j.1442-9993.1985.tb00895.x. Daskalakou, E. N., and C. A. Thanos. 1996. “Aleppo Pine ( Pinus Halepensis ) Postfire Regeneration: The Role of Canopy and Soil Seed Banks.” International Journal of Wildland Fire 6 (2): 59–66. https://doi.org/10.1071/wf9960059. Daskalakou, E., and C.A. Thanos. 2010. “Postfire Seedling Dynamics and Performance in Pinus Halepensis Mill. Populations.” Acta Oecologica 36 (5): 446–53. https://doi.org/10.1016/j.actao.2010.05.001. Enright, N. J., and M. C. Agne. 2025. “Climate Drying Reduces Serotinous Seedbanks and Threatens Persistence in Two Fire-Killed Shrubs.” International Journal of Wildland Fire 34 (2). https://doi.org/10.1071/WF24046. Enright, Neal J, Joseph B Fontaine, David MJS Bowman, Ross A Bradstock, and Richard J Williams. 2015. “Interval Squeeze: Altered Fire Regimes and Demographic Responses Interact to Threaten Woody Species Persistence as Climate Changes.” Frontiers in Ecology and the Environment 13 (5): 265–72. https://doi.org/10.1890/140231. Gauthier, Sylvie, Yves Bergeron, and Jean-Pierre Simon. 1996. “Effects of Fire Regime on the Serotiny Level of Jack Pine.” Journal of Ecology 84 (4): 539–48. https://doi.org/10.2307/2261476. GBIF.Org User. 2024. “Occurrence Download.” The Global Biodiversity Information Facility. Text/tab-separated-values,application/zip, 56145620. https://doi.org/10.15468/DL.8N9WK2. Givnish, Thomas J. 1981. “Serotiny, Geography, and Fire in the Pine Barrens of New Jersey.” Evolution 35 (1): 101–23. https://doi.org/10.2307/2407945. Gómez-Aparicio, Lorena, Miguel A. Zavala, Francisco J. Bonet, and Regino Zamora. 2009. “Are Pine Plantations Valid Tools for Restoring Mediterranean Forests? An Assessment along Abiotic and Biotic Gradients.” Ecological Applications 19 (8): 2124–41. https://doi.org/10.1890/08-1656.1. Goubitz, S., M.J.A. Werger, and G. Ne’eman. 2002. “Germination Response to Fire-Related Factors of Seeds from Non-Serotinous and Serotinous Cones.” Plant Ecology 169 (2): 195–204. https://doi.org/10.1023/A:1026036332277. Goubitz, Shirrinka. 2001. “Reproduction Ecology of Pinus Halepensis : A Monoecious, Wind-Pollinated and Partially Serotinous Mediterranean Pine Tree.” Dissertation. November 5. https://dspace.library.uu.nl/handle/1874/205. Greene, David F., Jeffrey M. Kane, Edith Pounden, and Sean T. Michaletz. 2024. “Cone Allometry and Seed Protection from Fire Are Similar in Serotinous and Nonserotinous Conifers.” New Phytologist 242 (1): 93–106. https://doi.org/10.1111/nph.19578. Groom, Philip K., and Byron B. Lamont. 2011. “Regional and Local Effects on Reproductive Allocation in Epicormic and Lignotuberous Populations of Banksia Menziesii.” Plant Ecology 212 (12): 2003–11. https://doi.org/10.1007/s11258-011-9935-x. Guiote, Carmen, and Juli G. Pausas. 2023. “Fire Favors Sexual Precocity in a Mediterranean Pine.” Oikos 2023 (3): e09373. https://doi.org/10.1111/oik.09373. Hartig, Florian. 2022. DHARMa: Residual Diagnostics for Hierarchical (Multi-Level / Mixed) Regression Models . Version 0.4.6. Released. https://CRAN.R-project.org/package=DHARMa. He, Tianhua, Byron B. Lamont, and Katherine S. Downes. 2011. “Banksia Born to Burn.” New Phytologist 191 (1): 184–96. https://doi.org/10.1111/j.1469-8137.2011.03663.x. Hernández-Serrano, Ana, Miguel Verdú, Santiago C. González-Martínez, and Juli G. Pausas. 2013. “Fire Structures Pine Serotiny at Different Scales.” American Journal of Botany 100 (12): 2349–56. https://doi.org/10.3732/ajb.1300182. Hernández-Serrano, Ana, Miguel Verdú, Luís Santos-del-Blanco, José Climent, Santiago C. González-Martínez, and Juli G. Pausas. 2014. “Heritability and Quantitative Genetic Divergence of Serotiny, a Fire-Persistence Plant Trait.” Annals of Botany 114 (3): 571–77. https://doi.org/10.1093/aob/mcu142. Intergovernmental Panel On Climate Change (Ipcc). 2021. Climate Change 2021 – The Physical Science Basis: Working Group I Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change . 1st ed. Cambridge University Press. https://doi.org/10.1017/9781009157896. Jordano, P. 2000. “Fruits and Frugivory.” Seeds: The Ecology of Regeneration in Plant Communities, CABI Books, 125–65. https://doi.org/10.1079/9780851994321.0125. Karger, Dirk Nikolaus, Adam M. Wilson, Colin Mahony, Niklaus E. Zimmermann, and Walter Jetz. 2021. “Global Daily 1 Km Land Surface Precipitation Based on Cloud Cover-Informed Downscaling.” Scientific Data 8 (1): 307. https://doi.org/10.1038/s41597-021-01084-6. Keeley, J. E., and P.H. Zedler. 1998. “Evolution of Life Histories in Pinus.” In Ecology and Biogeography of Pinus, pp. 219–250. Cambridge (GB): Cambridge University Press. Keeley, Jon E., William J. Bond, Ross A. Bradstock, Juli G. Pausas, and Philip W. Rundel. 2012. Fire in Mediterranean Ecosystems: Ecology, Evolution and Management . Cambridge (GB): Cambridge university press. Keeley, Jon E., and Juli G. Pausas. 2022. “Evolutionary Ecology of Fire.” Annual Review of Ecology, Evolution, and Systematics 53 (Volume 53, 2022): 203–25. https://doi.org/10.1146/annurev-ecolsys-102320-095612. Lamont, Byron B., Stephen W. Connell, and Stephen M. Bergl. 1991. “Seed Bank and Population Dynamics of Banksia Cuneata: The Role of Time, Fire, and Moisture.” Botanical Gazette 152 (1): 114–22. https://doi.org/10.1086/337870. Lamont, Byron B., and N. J. Enright. 2000. “Adaptive Advantages of Aerial Seed Banks.” Plant Species Biology 15 (2): 157–66. https://doi.org/10.1046/j.1442-1984.2000.00036.x. Lamont, Byron B., D. C. Le Maitre, R. M. Cowling, and N. J. Enright. 1991. “Canopy Seed Storage in Woody Plants.” The Botanical Review 57 (4): 277–317. https://doi.org/10.1007/BF02858770. Lamont, Byron B., Juli G. Pausas, Tianhua He, Ed T. F. Witkowski, and Mick E. Hanley. 2020. “Fire as a Selective Agent for Both Serotiny and Nonserotiny Over Space and Time.” Critical Reviews in Plant Sciences 39 (2): 140–72. https://doi.org/10.1080/07352689.2020.1768465. Lüdecke, Daniel. 2018. “Ggeffects: Tidy Data Frames of Marginal Effects from Regression Models.” Journal of Open Source Software 3 (26): 772. https://doi.org/10.21105/joss.00772. Martín-Sanz, Ruth C., Luis Santos-del-Blanco, Eduardo Notivol, M. Regina Chambel, Roberto San-Martín, and José Climent. 2016. “Disentangling Plasticity of Serotiny, a Key Adaptive Trait in a Mediterranean Conifer.” American Journal of Botany 103 (9): 1582–91. https://doi.org/10.3732/ajb.1600199. Mezquida, E. T., and C. W. Benkman. 2005. “The Geographic Selection Mosaic for Squirrels, Crossbills and Aleppo Pine.” Journal of Evolutionary Biology 18 (2): 348–57. https://doi.org/10.1111/j.1420-9101.2004.00846.x. Moya, D., J. M. Espelta, F. R. López-Serrano, M. Eugenio, and J. De Las Heras. 2008. “Natural Post-Fire Dynamics and Serotiny in 10-Year-Old Pinus Halepensis Mill. Stands along a Geographic Gradient.” International Journal of Wildland Fire 17 (2): 287–92. https://doi.org/10.1071/WF06121. Moya, Daniel, Josep M. Espelta, Iraima Verkaik, Francisco López-Serrano, and Jorge Las Heras. 2007. “Tree Density and Site Quality Influence on Pinus Halepensis Mill. Reproductive Characteristics after Large Fires.” Annals of Forest Science 6 (64): 649–56. https://doi.org/10.1051/forest:2007043. Muir, Patricia S., and James E. Lotan. 1985. “Disturbance History and Serotiny of Pinus Contorta in Western Montana.” Ecology 66 (5): 1658–68. https://doi.org/10.2307/1938028. Nathan, Ran, Uriel N. Safriel, Imanuel Noy-Meir, and Gabriel Schiller. 1999. “Seed Release without Fire in Pinus Halepensis, a Mediterranean Serotinous Wind-Dispersed Tree.” Journal of Ecology 87 (4): 659–69. https://doi.org/10.1046/j.1365-2745.1999.00382.x. Ne’eman, Gidi, Shirrinka Goubitz, and Ran Nathan. 2004. “Reproductive Traits of Pinus Halepensis in the Light of Fire – a Critical Review.” Plant Ecology 171 (1): 69–79. https://doi.org/10.1023/B:VEGE.0000029380.04821.99. Palacio, Facundo X., Juan Manuel Girini, and Mariano Ordano. 2016. “Linking the Hierarchical Decision-Making Process of Fruit Choice and the Phenotypic Selection Strength on Fruit Traits by Birds.” Journal of Plant Ecology, June 29, rtw063. https://doi.org/10.1093/jpe/rtw063. Paneghel, Mara, Gil Torné, Xavier Morin, Josu G. Alday, and Lluís Coll. 2024. “Increasing Temperature Threatens Post‐fire Auto‐successional Dynamics of a Mediterranean Obligate Seeder.” Journal of Ecology, October 16, 1365-2745.14433. https://doi.org/10.1111/1365-2745.14433. Parker, Anna L., and Craig W. Benkman. 2020. “Enhanced Seed Defenses Potentially Relax Selection by Seed Predators against Serotiny in Lodgepole Pine.” Ecology and Evolution 10 (12): 6001–8. https://doi.org/10.1002/ece3.6339. Pausas, J.G., N. Ouadah, A. Ferran, T. Gimeno, and R. Vallejo. 2002. “Fire Severity and Seedling Establishment in Pinus Halepensis Woodlands, Eastern Iberian Peninsula.” Plant Ecology 169 (2): 205–13. https://doi.org/10.1023/A:1026019528443. Pausas, Juli G. 2004. “Changes in Fire and Climate in the Eastern Iberian Peninsula (Mediterranean Basin).” Climatic Change 3 (63): 337–50. https://doi.org/10.1023/B:CLIM.0000018508.94901.9c. Pausas, Juli G., and Jon E. Keeley. 2009. “A Burning Story: The Role of Fire in the History of Life.” BioScience 59 (7): 593–601. https://doi.org/10.1525/bio.2009.59.7.10. Pausas, Juli G., and Byron B. Lamont. 2022. “Fire-Released Seed Dormancy - a Global Synthesis.” Biological Reviews 97 (4): 1612–39. https://doi.org/10.1111/brv.12855. Pausas, Juli G., Byron B. Lamont, Jon E. Keeley, and William J. Bond. 2022. “Bet‐hedging and Best‐bet Strategies Shape Seed Dormancy.” The New Phytologist 236 (4): 1232. https://doi.org/10.1111/nph.18436. Pausas, Juli G., and Susana Paula. 2012. “Fuel Shapes the Fire–Climate Relationship: Evidence from Mediterranean Ecosystems.” Global Ecology and Biogeography 21 (11): 1074–82. https://doi.org/10.1111/j.1466-8238.2012.00769.x. Purroy, Francisco, J. 2017. “Ardilla Roja - Sciurus Vulgaris.” Enciclopedia Virtual de Los Vertebrados Españoles. https://www.vertebradosibericos.org/mamiferos/scivul.html. QGIS Development Team. 2021. QGIS Geographic Information System . Version 3.16.13. Released. https://qgis.org. Quézel, P. 2000. “Taxonomy and Biogeography of Mediterranean Pines ( Pinus Halepensis and P. Brutia ).” In Ecology, Biogeography and Management of Pinus Halepensis and P. Brutia Ecosystems in the Mediterranean Basin., Ne’eman, G. and Trabaud, L. (eds). Leiden. R Core Team. 2023. R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing. https://www.R-project.org/. Radeloff, Volker C., David J. Mladenoff, Raymond P. Guries, and Mark S. Boyce. 2004. “Spatial Patterns of Cone Serotiny in Pinus Banksiana in Relation to Fire Disturbance.” Forest Ecology and Management 189 (January): 133–41. https://doi.org/10.1016/j.foreco.2003.07.040. Rima, P., G. Aloise, M. Cagnin, and L. Wauters. 2007. “The Use of Species‐specific Cone Remains of Sympatric Arboreal Rodents to Monitor Their Distribution.” Italian Journal of Zoology 74 (3): 289–96. https://doi.org/10.1080/11250000701286647. Rocha, Rita Gomes, Lucas Armand Wauters, Maria da Luz Mathias, and Carlos Fonseca. 2014. “Will an Ancient Refuge Become a Modern One? A Critical Review on the Conservation and Research Priorities for the Red Squirrel ( Sciurus Vulgaris ) in the Iberian Peninsula.” Hystrix, the Italian Journal of Mammalogy 25 (1). https://doi.org/10.4404/hystrix-25.1-9496. Siepielski, A. M., and C. W. Benkman. 2007. “Selection by a Predispersal Seed Predator Constrains the Evolution of Avian Seed Dispersal in Pines.” Functional Ecology 21 (3): 611–18. https://doi.org/10.1111/j.1365-2435.2007.01261.x. Siepielski, Adam M., and Craig W. Benkman. 2008. “Seed Predation and Selection Exerted by a Seed Predator Influence Subalpine Tree Densities.” Ecology 89 (10): 2960–66. https://doi.org/10.1890/08-0072.1. Smith, Christopher C. 1970. “The Coevolution of Pine Squirrels ( Tamiasciurus ) and Conifers.” Ecological Monographs 40 (3): 349–71. https://doi.org/10.2307/1942287. Smith, Christopher C. 1975. “The Coevolution of Plants and Seed Predators.” In Coevolution of Animals and Plants, L. E. Gilbert and P. H. Raven, (eds), pp. 53–77. Austin: University of Texas Press. Souto-Veiga, Rodrigo, Juergen Groeneveld, Neal J. Enright, Joseph B. Fontaine, and Florian Jeltsch. 2022. “Declining Pollination Success Reinforces Negative Climate and Fire Change Impacts in a Serotinous, Fire-Killed Plant.” Plant Ecology 223 (7): 863–81. https://doi.org/10.1007/s11258-022-01244-7. Talluto, Matt V., and Craig W. Benkman. 2013. “Landscape-Scale Eco-Evolutionary Dynamics: Selection by Seed Predators and Fire Determine a Major Reproductive Strategy.” Ecology 94 (6): 1307–16. Talluto, Matthew V., and Craig W. Benkman. 2014. “Conflicting Selection from Fire and Seed Predation Drives Fine-Scaled Phenotypic Variation in a Widespread North American Conifer.” Proceedings of the National Academy of Sciences 111 (26): 9543–48. https://doi.org/10.1073/pnas.1400944111. Tapias, Raul, José Climent, Jose A. Pardos, and Luis Gil. 2004. “Life Histories of Mediterranean Pines.” Plant Ecology 171 (1): 53–68. https://doi.org/10.1023/B:VEGE.0000029383.72609.f0. Verdú, M, and J. G Pausas. 2007. “Fire Drives Phylogenetic Clustering in Mediterranean Basin Woody Plant Communities.” Journal of Ecology 95 (6): 1316–23. https://doi.org/10.1111/j.1365-2745.2007.01300.x. Vincenzi, Simone, and Andrea Piotti. 2014. “Evolution of Serotiny in Maritime Pine ( Pinus Pinaster ) in the Light of Increasing Frequency of Fires.” Plant Ecology 215 (7): 689–701. https://doi.org/10.1007/s11258-014-0342-y. Vogel, Johannes, Eva Paton, Valentin Aich, and Axel Bronstert. 2021. “Increasing Compound Warm Spells and Droughts in the Mediterranean Basin.” Weather and Climate Extremes 32 (June): 100312. https://doi.org/10.1016/j.wace.2021.100312. Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License.

Authors Metrics & Citations Metrics Article Usage 259views 141downloads Citations Download citation Carmen Guiote, Eduardo Mezquida, Juli Pausas. Squirrels reduce post-fire regeneration potential in serotinous pines. Authorea. 29 October 2025. DOI: https://doi.org/10.22541/au.176173040.09360567/v1 DOI: https://doi.org/10.22541/au.176173040.09360567/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.