Abstract Aims Fires play a crucial role mediating species interactions in the Mediterranean Basin, with one prominent example being the nursing effect of post-fire resprouting shrubs on tree recruits, which then outcompete their benefactors throughout succession. Yet, the community structuring role of resprouting shrubs as potential facilitators of post-fire recruiting subshrub species, which are commonly outcompeted in late post-fire stages, has been overlooked. The aims of this work were to investigate (i) whether proximity to resprouting shrubs increased the demographic performance of a fire-adapted carnivorous subshrub and (ii) whether mature shrubs negatively affected the performance of established plants through interference with prey capture. Methods To evaluate the facilitative effects of resprouting shrubs, we sowed seeds of Drosophyllum lusitanicum, a carnivorous, seeder pyrophyte, into two microhabitats in recently burned heathland patches defined by proximity to resprouting shrubs. We monitored key demographic rates of emerged seedlings for 2 years. To test for competitive effects of shrubs on plant performance at a later habitat regeneration stage, we placed greenhouse-reared, potted plants into distinct microhabitats in neighboring burned and unburned heathland patches and monitored prey capture. Both experiments were performed in the Aljibe Mountains at the Northern Strait of Gibraltar and were replicated in 2 years. Important Findings Resprouting shrubs significantly improved survival, juvenile size and flowering probability compared with open microhabitats, and had no significantly negative effects on the growth of recruits. Prey capture was significantly lower in unburned heathland patches compared with burned ones, thus partly explaining the decrease in survival of Drosophyllum individuals in mature heathlands. However, microhabitat did not affect prey capture. Our findings suggest that not only periodic fires, removing biomass in mature stands, but also resprouting neighbors, increasing establishment success after fire, may be important for the viability of early successional pyrophytes. competition, Drosophyllum, lusitanicum, early successional species, habitat succession, Mediterranean heathlands, pyrophyte INTRODUCTION Species interactions such as facilitation and competition drive biodiversity and community composition (Bertness and Callaway 1994; Holmgren et al. 1997; Wright et al. 2014). The prevalence and importance of facilitation over competition in natural communities, meanwhile, have been linked to increases in abiotic stress such as drought or frost (Bertness and Callaway 1994; but see Holmgren and Scheffer 2010), where facilitation has been shown to increase biodiversity (He et al. 2013; Soliveres et al. 2015). Mediterranean shrublands are diverse ecosystems where seasonal drought typically constitutes a severe, periodic stress (Moreno et al. 2011). In the Mediterranean Basin, shrub species resilient to droughts (Zeppel et al. 2015) may increase the survival of tree and shrub seedlings recruiting under or nearby them (Gómez-Aparicio et al. 2004). Facilitation in this setting happens via the increased soil moisture and/or nutrient content, low evapotranspiration, protection from ultraviolet radiation or herbivory, and/or buffering against high temperatures and winds provided by the nursing plant (Callaway 2007; Baraza et al. 2006; He et al. 2013). Fires play a key role in mediating facilitative plant interactions in Mediterranean heathlands (Keeley et al. 2012). In early post-fire stages, where the effects of drought stress are most pronounced (Peñuelas et al. 2007), plant diversity is also highest (Ojeda et al. 1996; Keeley et al. 2012). This diversity is largely driven by a high abundance of post-fire recruiting plant species, usually short-lived herbs and subshrubs, which disappear as shrub cover increases with time since fire (Calvo et al. 2002; Ojeda et al. 1996; Yates and Ladd 2010). It is well known that after fire, resprouting, drought-resilient shrubs act as ecosystem engineers promoting biodiversity by facilitating survival and growth of juvenile stages of other long-lived shrubs and trees (reviewed in Vilà and Sardans 1999; Gómez-Aparicio et al. 2004). However, although nursing effects of resprouting shrubs on recruiting seedlings of short-lived, post-fire dwelling shrub and subshrub species have been assumed in Mediterranean systems (Verdú et al. 2009), they have rarely been investigated experimentally. The question of whether growing under or in close proximity to resprouting shrubs improves the performance of a short-lived, post-fire dweller is complicated by the fact that such facilitative interactions change with habitat succession. With shrub maturing into dense heathlands as rapidly as 4 years after fires in Mediterranean ecosystems (Calvo et al. 2002; Céspedes et al. 2014), facilitation by resprouting shrubs may eventually turn into competition for light and resources against smaller post-fire recruiting subshrubs (Vilà and Sardans 1999). They may thus be preferentially found in open patches, risking higher mortality due to adverse environmental conditions (e.g. solar radiation, wind, drought) in early post-fire habitats but avoiding competition for longer through habitat succession. The presence of post-fire recruiting species close to resprouting shrubs may then simply be the result of propagule concentration by shrubs (Callaway 1995). To investigate the net effects of facilitation vs. competition, a study must span a reasonable time interval to capture habitat succession (He et al. 2013). In addition, it has been repeatedly demonstrated that the choice of demographic performance estimator (e.g. survival, growth and reproduction) by which the effects of facilitation or competition are measured may strongly affect results (Maestre et al. 2005; He et al. 2013). For example, several studies investigating facilitation under high abiotic stress in arid ecosystems have found no effect of neighbors on plants survival and growth but a strong facilitative effect on fecundity of target plants (Donovan et al. 2000; Maestre et al. 2005). Therefore, studies must consider several performance estimators to gain a full picture of the effects of neighbors on the performance of a target species. Disentangling the roles of facilitation and competition along ecological succession will shed light on the dependence of early successional species on the presence of community structuring species (Dickie et al. 2005). For example, selective removal of resprouting shrubs for lignotuber harvesting (Ojeda et al. 1996) or heathland afforestation with pines (Andrés and Ojeda 2002) may subsequently affect the success of post-fire dwelling species in fire-prone Mediterranean heathlands if resprouting shrubs act as nursing plants in early post-fire regeneration stages. To test the hypothesis that resprouting shrubs provide key facilitative services to short-lived, post-fire subshrub species, we quantified the effects of shrub neighbors on the demographic performance of the Mediterranean heathland endemic Drosophyllum lusitanicum (Drosophyllaceae), a short-lived, carnivorous pyrophyte (Paniw et al. 2016). In order to base our results on several performance measures, we estimated seed germination as well as survival, growth, and reproduction in two microhabitats, close to shrubs and open, from a recently burned heathland patch. To account for habitat succession, we monitored the performance of emerged seedlings for two consecutive years. We also recorded trapped insects on the sticky leaves of greenhouse-reared, potted young individuals placed in recently burned and neighboring unburned sites in order to test for negative effects of fully developed shrubs in mature communities via interference with nutrient acquisition (i.e. insect capture). Such interference may occur directly through shading of plants (Schulze et al. 2001) and/or indirectly due to a decreased prey insect abundance in mature shrub communities (Potts et al. 2003). In addition, we noted leaf damage on the potted plants in order to test whether shrubs protect individuals from solar radiation and/or wind. MATERIALS AND METHODS Study species and sites Drosophyllum lusitanicum (L.) Link (Drosophyllaceae; Drosophyllum hereafter) is a short-lived, carnivorous perennial subshrub endemic to the Western Mediterranean Basin (Garrido et al. 2003) and tightly associated with fire-prone Mediterranean heathlands (Müller and Deil 2001). These heathland habitats are characterized by a Mediterranean climate regime and occur on highly acidic, infertile, siliceous soils (Ojeda et al. 2010). Their dominant vegetation consists of shrubs in the Ericaceae (Calluna vulgaris, Erica australis, E. umbellata and E. scoparia) and Fabaceae families (e.g. Stauracanthus boivini, Pterospartum tridentatum, Genista tridens). Drosophyllum populations are threatened by habitat degradation (Correia and Freitas 2002; Garrido et al. 2003). In natural heathlands, population dynamics of this species are linked to recurrent fires, since its seed germination is stimulated by both direct (heat; Paniw et al. 2016) and indirect (opening of vegetation; Paniw et al. 2015) fire-related cues. Consequently, Drosophyllum’s highest population densities are attained during early post-fire stages, typically 1–3 years after fires (Paniw et al. 2016). During such stages, emerging seedlings are typically exposed to seasonal (summer) drought stress (Adlassnig et al. 2006), and resprouting shrubs may thus act as nursing plants. However, germination is increasingly inhibited by mature shrubs and accumulation of ground litter in heathlands ≥4 years after fires (Correia and Freitas 2002; Paniw et al. 2016, 2017). In addition, shrubs may interfere with insect capture as has been shown for other carnivorous plant species (Schulze et al. 2001), but this has not been investigated for Drosophyllum. To quantify the extent of facilitation and/or competition by shrubs during post-fire habitat succession, we carried out two experiments replicated at two Mediterranean heathland sites within the Aljibe Mountains, at the northern side of the Strait of Gibraltar (Fig. 1). Parts of the two study sites burned by wildfires (see below). Natural Drosophyllum populations occur at both sites, but were located >200 m away from the experimental settings. Figure 1: View largeDownload slide location of the two sites within the Aljibe Mountains at the Northern Straits of Gibraltar (red box) used in this study and experimental design at each study site. At Sierra Carbonera, the seed-sowing experiment was designed as random plots (P) in a burned heathland patch adjacent to an unburned one. Within each plot, two microhabitats (open and close to shrub) were used. At Retin, the prey capture experiment was designed at two sites as paired plots in adjacent burned and unburned heathland patches. Figure 1: View largeDownload slide location of the two sites within the Aljibe Mountains at the Northern Straits of Gibraltar (red box) used in this study and experimental design at each study site. At Sierra Carbonera, the seed-sowing experiment was designed as random plots (P) in a burned heathland patch adjacent to an unburned one. Within each plot, two microhabitats (open and close to shrub) were used. At Retin, the prey capture experiment was designed at two sites as paired plots in adjacent burned and unburned heathland patches. Seed-sowing experiment To test the interspecific effects of resprouting shrubs on the vital rates germination, survival, growth and reproduction of Drosophyllum individuals, we conducted a seed-sowing experiment at a burned heathland site (Sierra Carbonera; fire in August 2011). We established seven paired plots perpendicular to the main elevation gradient (Fig. 1) and sowed two cohorts of seeds, in August 2012 and in 2013, to track the aforementioned vital rates. The experiment assessed relative germination in distinct microhabitats during early stages of habitat succession, but not germination in response to direct fire cues. We therefore did not pre-treat seeds, e.g. exposing them to heat, and instead observed germination in response to indirect cues (Paniw et al. 2015). In each of the seven plots, we distinguished two types of microhabitats: ‘open’ and ‘shrub’ (Fig. 1). We chose the most abundant shrub species in each plot as shrub microhabitat, which were either Erica scoparia or Stauracanthus boivini. Both species had similar, rounded/conical crowns and show similar growth rates after fire (M. Paniw, unpubl.), and we assumed that neighbor identity would not significantly affect our results (Correia and Freitas 2002). We chose exemplars of ≥20 cm crown radius as neighbors to ensure that neighbors were large enough to potentially affect the performance of Drosophyllum seedlings. In each microhabitat, we sowed 50 seeds, randomly collected from >80 individuals across five Drosophyllum populations, in squares (20 × 20 cm2), using one and two squares per microhabitat treatment for the 2012 and 2013 cohort, respectively. The edges of the squares were randomly positioned >25 cm and <20 cm away from the edge of resprouting shrubs for the open and shrub microhabitats, respectively. The sowing depth was 0.5–1 cm, within the natural seed depth in the soil (Paniw et al. 2016). In three of the seven plots, we created a control treatment by digging up soil in three 20 × 20 cm2 squares in ‘open’ and ‘shrub’ microhabitats, respectively, without seed sowing so as to control for potential germination of naturally occurring Drosophyllum seeds after mechanical disturbance. In this control treatment, the ‘open’ microhabitats were picked randomly, while ‘shrub’ microhabitats consisted of the same shrub that was used as a neighbor for a sowing treatment (see above). We only used three plots because active seed dispersal does not occur in Drosophyllum (Ortega-Olivencia et al. 1995), and we therefore did not expect to find naturally occurring seeds in our experimental plots. We recorded the proportion of seeds that germinated and their consequent number and length (cm) of leaves in each square in April 2013 and 2014, 8 months after sowing when germination rates in natural Drosophyllum populations are highest (Paniw et al. 2016). We tagged all emerged plants and followed their fate, i.e. survival, growth and reproduction (production of flowering stalks), in 6-month intervals until April of 2014 and 2015, or 20 months after sowing, for the 2012 and 2013 cohort, respectively. Although some recruits likely died before tagging in April, we assumed that this mortality did not differ between open and shrub microhabitats (see online supplementary Fig. S1). Lastly, using HOBO data loggers (Onset Computer Corporation 2013), we took hourly measurements of temperature and relative humidity for the length of the experiment to monitor potential changes in weather during the two study periods. Prey capture experiments In order to test for any negative effects of shrub cover on established Drosophyllum individuals, we quantified prey capture rates in 9-month-old potted plants placed in burned and unburned habitat patches. Drosophyllum individuals in this experiment were grown in clay pots under greenhouse conditions (20°C, 85% humidity, and daily 50 ml watering with decalcified water) at the University of Cadiz from seeds collected randomly in five natural populations from southern Spain (Fig. 1). We initially performed this experiment in early May 2013, within the growing season of natural Drosophyllum populations (Paniw et al. 2016), at a heathland site in Sierra Retin (Fig. 1). The last fire occurred in August 2010 at this site. We established seven paired plots perpendicular to the main elevation gradient, each plot consisting of burned and unburned (>30 years after last fire) subplots. After another fire occurred at a different site in Sierra Retin in August 2013, we implemented the same experimental design in early May 2014 at the newly burned and an adjacent unburned (>30 years after last fire) patch (Fig. 1). Although the burned patches in 2013 and 2014 were not in the same post-fire successional stage, we grouped them into the ‘burned’ treatment as both were in an early post-fire stage (<3 years since fire). Within each early post-fire/burned and unburned subplot for 2013 and 2014, we distinguished two microhabitats, ‘open’ and ‘shrub’. In early post-fire/burned subplots, we used the aforementioned criteria for Sierra Carbonera to place pots within the microhabitats. In unburned subplots, pots were placed >25 cm from the edge of mature shrubs and underneath the shrub canopy for the ‘open’ and ‘shrub’ microhabitats, respectively. Mature shrubs in unburned subplots had a crown radius of 89.7 (±37.1 SD) cm and were 51.6 (±17.0 SD) cm tall. We placed two potted individuals of Drosophyllum free of insect prey at each microhabitat within each subplot across the 14 plots (seven per year/site). As in the sowing experiment in Sierra Carbonera, we chose the most abundant shrub species in each subplot as shrub microhabitat, which were Erica scoparia, Stauracanthus boivini or Calluna vulgaris. Individuals were watered daily with 100 ml of decalcified water during seven days to prevent desiccation. We then took the potted individuals to the laboratory for quantification, size estimation and identification to at least the taxonomic order of every trapped insects per individual. To assess whether shrubs protect Drosophyllum individuals from physical damage, we also examined all mature leaves for visual damage such as broken-off parts (see online supplementary Fig. S2). Statistical analyses Statistical analyzes of both the seed-sowing and pre-capture experiment were performed separately for each year because we did not have enough temporal replicates to include year as a random effect in our models. For the seed-sowing experiment, we fitted generalized mixed effect models (GLMMs) to describe each vital rate separately (seedling recruitment, seedling and juvenile survival probability and flowering probability) as a function of microhabitat (open vs. shrub). We fitted GLMMs with a binomial error distribution for germination, survival, and flowering, and used a normal error distribution to model size, measured as leaf number × length of longest leaf (cm) on the logarithmic scale. For the prey capture experiment, we fitted GLMMs describing number of insects and number of visibly damaged leaves as functions of post-fire habitat state (burned vs. unburned), microhabitat (open vs. shrub) and their interaction. We used number of leaves per plant as an offset in the models, thereby treating the two responses as proportions but allowing the models to be fit as count data in a GLMM framework. We fitted the two models using a negative binomial error distribution as simple Poisson models showed overdispersion, i.e. the ratio of squared Pearson residuals and residual degrees of freedom was >1 (χ2, P < 0.01; Ver Hoef and Boveng 2007). All analyses were performed with the lme4 package in R (Bates et al. 2014). In all models, we used plot as a random effect on the model mean. We used likelihood ratio tests to determine significant differences between treatments (Vuong 1989). These tests compare increasingly complex, or nested, models to simpler ones (starting with intercept-only models). When the effect of microhabitat or post-fire state was significant, we applied a post-hoc Tukey’s honestly significant difference (HSD) test to the linear predictors using the R package multcomp (Hothorn et al. 2008) to detect significant pairwise differences between treatment levels. RESULTS Seed-sowing experiment The microhabitat (open vs. shrub) in burned plots markedly, positively affected several vital rates Drosophyllum individuals, although only for seeds sowed in 2012, 1 year after fire (Figs. 2 and 3). Seedling survival was significantly higher when seeds were sowed close to resprouting shrubs than in the open for the 2012 cohort (χ2 deviance = 4.6, df = 1; P < 0.05; Fig. 2b). On the other hand, none of the three recruitment vital rates (germination, seedling size and seedling survival) differed significantly between open and shrub microhabitats in the burned subplots for the 2013 cohort (Fig. 2). For the 2012 cohort, juvenile size was also significantly higher in shrub microhabitat (χ2 deviance = 5.1, df = 1; P < 0.05; Fig. 3b); and recruiting plants in the shrub microhabitat had a significantly higher probability of flowering after 20 months than recruiting plants in the open (χ2 deviance = 4.7, df = 1; P < 0.05; Fig. 3c). No recruiting plant from the 2013 cohort flowered 20 months after emergence, so that statistical analyses of flowering were not possible (Fig. 3c). No germination occurred in the control treatment, and we therefore excluded it from statistical analyses. In addition, temperature and relative humidity did not vary substantially between the 2 years (see online supplementary Fig. S3). Figure 2: View largeDownload slide effects of microhabitat (open and close to shrubs) in burned heathland patches on (a) germination rate, (b) seedling size (cm) and (c) seedling survival probability 8 months following sowing of Drosophyllum lusitanicum seeds sown in September 2012 and 2013. Error bars indicate ±1 SE. Different small (large) letters indicate significant differences (Tukey’s HSD, P < 0.05) of group means between microhabitat for the treatment in 2012 (2013, respectively). Figure 2: View largeDownload slide effects of microhabitat (open and close to shrubs) in burned heathland patches on (a) germination rate, (b) seedling size (cm) and (c) seedling survival probability 8 months following sowing of Drosophyllum lusitanicum seeds sown in September 2012 and 2013. Error bars indicate ±1 SE. Different small (large) letters indicate significant differences (Tukey’s HSD, P < 0.05) of group means between microhabitat for the treatment in 2012 (2013, respectively). Figure 3: View largeDownload slide effects of microhabitat (open and close to shrubs) in burned heathland patches on (a) survival probability, (b) size (cm) and (c) flowering probability of individuals 20 months following sowing of Drosophyllum lusitanicum seeds in September 2012 and September 2013. Error bars indicate ±1 SE. Different small (large) letters indicate significant differences (Tukey’s HSD, P < 0.05) of group means between microhabitat for the treatment in 2012 (2013, respectively). Figure 3: View largeDownload slide effects of microhabitat (open and close to shrubs) in burned heathland patches on (a) survival probability, (b) size (cm) and (c) flowering probability of individuals 20 months following sowing of Drosophyllum lusitanicum seeds in September 2012 and September 2013. Error bars indicate ±1 SE. Different small (large) letters indicate significant differences (Tukey’s HSD, P < 0.05) of group means between microhabitat for the treatment in 2012 (2013, respectively). Prey capture experiment Overall, insect capture was significantly higher in potted individuals placed in the burned subplots than in neighboring unburned ones in both years/sites in the study (2013: χ2 deviance = 37.0, df = 1, P < 0.01; 2014: χ2 deviance = 14.9, df = 1, P < 0.01), regardless of shrub/open microhabitat (Fig. 4a). The majority of prey consisted of flies (Diptera) of various sizes, and we did not detect a difference in prey diversity between post-fire state or microhabitat (see online supplementary Table S1). Potted individuals showed a significant higher proportion of damaged leaves in burned subplots (2013: χ2 deviance = 13.1, df = 1, P < 0.01; 2014: χ2 deviance = 50.3, df = 1, P < 0.01) and, within burned subplots, in open microhabitats (2013: χ2 deviance = 10.2, df = 1, P < 0.01; 2014: χ2 deviance = 4.2, df = 1, P < 0.05; Fig. 4). Figure 4: View largeDownload slide effects of fire state of the habitat (burned and unburned) and microhabitat (open and close to shrubs) on (a) general insect capture and (b) proportion damaged leaves of Drosophyllum lusitanicum (L.) plants. Error bars indicate ±1 SE. All tests were performed separately for both years/sites. Different small (large) letters indicate significant pairwise differences (Tukey’s HSD, P < 0.05) between burned/unburned and open/shrub microhabitats for experiments in 2013 and 2014, respectively. Figure 4: View largeDownload slide effects of fire state of the habitat (burned and unburned) and microhabitat (open and close to shrubs) on (a) general insect capture and (b) proportion damaged leaves of Drosophyllum lusitanicum (L.) plants. Error bars indicate ±1 SE. All tests were performed separately for both years/sites. Different small (large) letters indicate significant pairwise differences (Tukey’s HSD, P < 0.05) between burned/unburned and open/shrub microhabitats for experiments in 2013 and 2014, respectively. DISCUSSION Fires in Mediterranean heathlands remove biomass and, while this provides post-fire recruiting species with the opportunity to colonize otherwise rather competitive habitats, it also exposes them to harsh environmental conditions typical of Mediterranean summers. Resprouting shrubs provide mosaics of relatively dense vegetation cover in otherwise barren post-fire landscapes and are known to facilitate survival and growth of other dominant shrub or tree species (Gómez-Aparicio et al. 2004). At the same time, only few studies have tested facilitation of shrubs on shorter-lived plant species, and rarely in the context of habitat succession (Verdú et al. 2009). Our results show evidence of a strong yet transient facilitative effect of resprouting shrubs on the demographic performance of a short-lived, post-fire dwelling, subshrub species and highlight the importance of resprouting shrubs on the post-fire recovery of heathland biodiversity. In addition, as our interpretations of the role of facilitation on the performance of Drosophyllum clearly depended on the demographic performance estimator used, we highlight the importance of including several measures of demographic performance when assessing species interactions (He et al. 2013; Maestre et al. 2005). We show that seedling survival and flowering probability of established individuals of Drosophyllum, the epitome of short-lived, post-fire recruiting species in Mediterranean heathlands (Andrés and Ojeda 2002; Paniw et al. 2015), are significantly improved by a nursing effect of resprouting shrubs in early post-fire stages (Figs. 2 and 3). In Mediterranean ecosystems, facilitation occurs typically when the abiotic stress, which the benefactor species alleviates, constitutes a non-resource stress for interacting species, such as temperature (Maestre et al. 2003). If a resource stress, e.g. water or nutrients, is the main abiotic stress and the niches of interacting species overlap, a shift to facilitation is not likely (Maestre et al. 2003, 2009). In our case, resprouting heathland shrubs and Drosophyllum plants do not compete for resources since the carnivorous Drosophyllum obtains nutrients from prey capture and a substantial amount of water in form of dew and mist absorbed by the mucilage droplets on stalked leaf glands (Adlassnig et al. 2006; Adamec 2009). Since Drosophyllum relies on the production of sticky mucilage on leaf stalked glands for prey capture, microhabitat conditions mitigating desiccation (Adlassnig et al. 2006; Gómez-Aparicio et al. 2004), such as close proximity to shrubs, providing protection from direct exposure to sun and wind, are likely to significantly improve the performance of individuals (Adlassnig et al. 2006; Bertol et al. 2015). These key facilitative effects were however transient, as they only occurred in the earliest post-fire stages. With increasing time since fire and vegetation recovery, our study showed that facilitative effects of resprouting shrubs diminish dramatically. A close proximity to resprouting shrub neighbors benefited recruitment only in the 2012 seed cohort, where seeds were sowed in a patch that burned just 1 year before. By 2014, when the 2013 cohort emerged from the soil, the differences in vital rates between the open and shrub microhabitat vanished. As the experiment was only performed in 2 years, we cannot discard that facilitation may have been an effect of year. However, another explanation for the overall lower germination, survival and size in the 2013 cohort is that post-fire succession occurs relatively rapidly in burned heathlands (Calvo et al. 2002; Ojeda et al. 1996), and the increase in vegetation cover negatively affects the germination and seedling size and survival of a post-fire dweller (Verdú et al. 2009). When we quantified seedling emergence in spring 2014, woody vegetation cover in some of our burned subplots had increased two-fold in the spring of 2013, from 30% to 60%. As a result, germination and survival of Drosophyllum may have been impeded. This conclusion seems reasonable when we consider that the temperature and relative humidity remained stable across the 2 years of the experiment (see online supplementary Fig. S3). Other studies in fire-prone systems have also shown that post-fire facilitative species interactions are transient and change throughout succession (Bullock 2009; Vilà and Terradas 1995). It is notable however that, although our study detected facilitation only in the first year after fire and primarily acting on seedling survival and subsequent probability of flowering of established individuals, none of the measured vital rates in either the 2012 or 2013 seed cohorts were significantly negatively affected by shrub cover in the first 2–3 years of post-fire succession. These results are in accordance with several other studies in burned Mediterranean ecosystems where competition between resprouting shrubs and post-fire seeder species was reported to be low (Calvo et al. 2002; Vilà and Sardans 1999). In long-unburned, mature heathland patches, the negative effects of biomass accumulation on post-fire dwellers appear predominant. In the case of Drosophyllum, individuals trapped fewer insects under mature shrubs compared with resprouting ones. At the same time, prey capture was not affected by microhabitat in unburned patches, with plants catching few insects regardless of whether they were located close to shrubs or in open microhabitats (Fig. 4a). Our results therefore suggest that a high density of mature heathland shrubs may both directly and indirectly interfere with prey capture and therefore survival of Drosophyllum individuals. Indeed, the increase in direct and apparent competition with post-fire habitat succession has been noted in several studies (Ojeda et al. 1996; Vilà and Terradas 1995) including some on carnivorous plant species (Brewer 2003; Schulze et al. 2001). For carnivorous plants in particular, the direct interference of mature shrubs with light and prey capture has been identified as one important reason why some taxa cannot survive in late post-disturbance habitat states (Schulze et al. 2001). At the same time, along with decreasing plant species diversity, insect diversity and abundance has also been shown to decrease with time since fire in Mediterranean ecosystems (Mateos et al. 2011; Potts et al. 2003). Dense shrublands may therefore indirectly affect prey capture rates of Drosophyllum by decreasing the overall availability of prey. Drosophyllum efficiently attracts prey (Bertol et al. 2015), and one may hypothesize that low numbers of prey insects caught in unburned heathland patches (even in open microhabitats within these patches) indicate low availability of prey. However, detailed studies on insect (prey) abundance and diversity in distinct heathland habitat patches are required to corroborate this hypothesis. Transient facilitation by resprouting shrubs may have important consequences for population dynamics and viability of Drosophyllum, as has been shown for other species (Maestre et al. 2009). Demographic census data of Drosophyllum populations show that most individuals in natural heathlands perish after one or two reproductive events, and that populations persist mostly in the seed bank after the third year after fire (Paniw et al. 2016, 2017). Such a ‘weedy’ life-history strategy, typical of an early successional species, may avoid severe effects of competition with growing shrub neighbors (Bazzaz 1979). Because seed input into the seed bank, ensuring mass germination after fires, is a key life-history strategy for many post-fire dwelling species (Menges and Quintana-Ascencio 2004; Quintana-Ascencio et al. 2003), including Drosophyllum (Salces-Castellano et al. 2016), shrub facilitation of reproduction in early post-fire stages when the majority of individuals reproduce is likely to significantly affect viability of populations. With post-fire succession, reproductive Drosophyllum individuals may show plastic responses to the interference with prey capture exercised by shrubs, such as reduced leaf size and reproductive structures, as has been demonstrated for pitcher plants (Brewer 2003; M. Paniw, pers. obs.). Indirect competition via interference with nutrient acquisition through prey capture may therefore be a price worth paying because the facilitative effects of the surrounding community in initial post-disturbance stages can outweigh these future costs. Fire-prone heathlands are particularly diverse habitats partly because resprouting shrubs, in their community engineer role, help to maintain biodiversity (Soliveres et al. 2015), particularly in early post-fire stages (Sedláková and Chytrý 1999). In our case, resprouting shrubs facilitate a rare, carnivorous plant species. It is alarming therefore that these shrub communities are being increasingly altered by habitat degradation such as active afforestation campaigns, which permanently change heathland community structure and composition by replacing shrub vegetation with trees (Andrés and Ojeda 2002). In addition, fire suppression has also become a common goal in current heathland management policies, drastically altering the natural fire regime to which the endemic flora is adapted (Bartolomé et al. 2005; Keeley et al. 2012). Fire suppression that results in fire-return intervals exceeding seed bank viability will likely pose a severe threat to post-fire dwelling, short-lived species relying on fire to alleviate direct and indirect competitive effects of shrubs (e.g. Bartolomé et al. 2005). Therefore, the preservation of periodic fire regimes within Mediterranean shrub communities must become a priority for conservation management (Keeley et al. 2012). It shall be stressed that, since Drosophyllum is an endemic, red-listed species, our study was limited by the amount of seeds we could collect in any given year. This in turn limited the number of treatments we could perform in this study. For example, the assumption that the identity of the resprouting neighbor did not affect demographic parameters may not hold if different neighboring species are considered and should be tested by including plant neighbor as a fixed effect in future studies. Therefore, our results should be corroborated by future studies in other areas across the range of Drosophyllum. SUPPLEMENTARY MATERIAL Supplementary data are available at Journal of Plant Ecology online. ACKNOWLEDGEMENTS We thank the Spanish military Campo de Adiestramiento de la Armada in Sierra del Retín (Barbate, Cádiz) for allowing access to the Monte Retin sites and facilitating the study. The Andalusian Consejería de Medio Ambiente provided the necessary permits to work with Drosophyllum lusitanicum, an endemic, red-listed species. This contribution has benefited from comments by two anonymous reviewers on a previous version of the manuscript. The study was supported by project BREATHAL (CGL2011-28759/BOS; Spanish Ministerio de Ciencia e Innovación). M.P. was supported by a Subprograma de Formación de Personal Investigador (FPI) fellowship (BES-2012–053075). Conflict of interest statement. None declared. REFERENCES Adamec L( 2009) Ecophysiological investigation on Drosophyllum lusitanicum: why doesn’t the plant dry out? Carniv Plant Newslett 38: 71– 4. Adlassnig W Peroutka M Eder Get al. ( 2006) Ecophysiological observations on Drosophyllum lusitanicum. Ecol Res 21: 255– 62. Google Scholar CrossRef Search ADS Andrés C Ojeda F( 2002) Effects of afforestation with pines on woody plant diversity of Mediterranean heathlands in southern Spain. Biodivers Conserv 11: 1511– 20. Google Scholar CrossRef Search ADS Baraza E Zamora R Hódar JA( 2006) Conditional outcomes in plant–herbivore interactions: neighbours matter. Oikos 113: 148– 56. Google Scholar CrossRef Search ADS Bartolomé J Plaixats J Fanlo Ret al. ( 2005) Conservation of isolated Atlantic heathlands in the Mediterranean region: effects of land-use changes in the Montseny biosphere reserve (Spain). Biol Conserv 122: 81– 8. Google Scholar CrossRef Search ADS Bates D Maechler M Bolker Bet al. ( 2014) lme4: Linear Mixed-Effects Models Using Eigen and S4. R package version 1.1–7 . http://CRAN.R-project.org/package=lme4 (10 January 2017, date last accessed). Bazzaz FA( 1979) The physiological ecology of plant succession. Annu Rev Ecol Syst 10: 351– 71. Google Scholar CrossRef Search ADS Bertness MD Callaway R( 1994) Positive interactions in communities. Trends Ecol Evol 9: 191– 3. Google Scholar CrossRef Search ADS PubMed Bertol N Paniw M Ojeda F( 2015) Effective prey attraction in the rare Drosophyllum lusitanicum, a flypaper-trap carnivorous plant. Am J Bot 102: 689– 94. Google Scholar CrossRef Search ADS PubMed Brewer JS( 2003) Why don’t carnivorous pitcher plants compete with non-carnivorous plants for nutrients? Ecology 84: 451– 62. Google Scholar CrossRef Search ADS Bullock JM( 2009) A long‐term study of the roles of competition and facilitation in the establishment of an invasive pine following heathland fires. J Ecol 97: 646– 56. Google Scholar CrossRef Search ADS Calvo L Tárrega R De Luis E( 2002) The dynamics of Mediterranean shrubs species over 12 years following perturbations. Plant Ecol 160: 25– 42. Google Scholar CrossRef Search ADS Callaway RM( 1995) Positive interactions among plants. Bot Rev 61: 306– 49. Google Scholar CrossRef Search ADS Callaway RM( 2007) Positive Interactions and Interdependence in Plant Communities . Berlin: Springer. Céspedes B Torres I Pérez Bet al. ( 2014) Burning season does not affect post‐fire regeneration but fire alters the balance of the dominant species in a seeder‐dominated Mediterranean shrubland. Appl Veg Sci 17: 711– 25. Google Scholar CrossRef Search ADS Correia E Freitas H( 2002) Drosophyllum lusitanicum, an endangered West Mediterranean endemic carnivorous plant: threats and its ability to control available resources. Bot J Linn Soc 140: 383– 90. Google Scholar CrossRef Search ADS Dickie IA Schnitzer SA Reich PBet al. ( 2005) Spatially disjunct effects of co-occurring competition and facilitation. Ecol Lett 8: 1191– 200. Google Scholar CrossRef Search ADS PubMed Donovan LA Richards JH( 2000) Juvenile shrubs show differences in stress tolerance, but no competition or facilitation, along a stress gradient. J Ecol 88: 1– 16. Google Scholar CrossRef Search ADS Garrido B Hampe A Marañón Tet al. ( 2003) Regional differences in land use affect population performance of the threatened insectivorous plant Drosophyllum lusitanicum (Droseraceae). Divers Distrib 9: 335– 50. Google Scholar CrossRef Search ADS Gómez-Aparicio L Zamora R Gómez JMet al. ( 2004) Applying plant facilitation to forest restoration in Mediterranean ecosystems: a meta-analysis of the use of shrubs as nurse plants. Ecol Appl 14: 1128– 38. Google Scholar CrossRef Search ADS He Q Bertness MD Altieri AH( 2013) Global shifts towards positive species interactions with increasing environmental stress. Ecol Lett 16: 695– 706. Google Scholar CrossRef Search ADS PubMed Holmgren M Scheffer M( 2010) Strong facilitation in mild environments: the stress gradient hypothesis revisited. J Ecol 98: 1269– 75. Google Scholar CrossRef Search ADS Holmgren M Scheffer M Huston MA( 1997) The interplay of facilitation and competition in plant communities. Ecology 78: 1966– 75. Google Scholar CrossRef Search ADS Hothorn T Bretz F Westfall P( 2008) Simultaneous inference in general parametric models. Biom J 50: 346– 63. Google Scholar CrossRef Search ADS PubMed Keeley JE Bond WJ Bradstock RAet al. ( 2012) Fire in Mediterranean Ecosystem: Ecology, Evolution and Management . Cambridge: Cambridge University Press. Maestre FT Bautista S Cortina J( 2003) Positive, negative, and net effects in grass-shrub interactions in Mediterranean semiarid grasslands. Ecology 84: 3186– 97. Google Scholar CrossRef Search ADS Maestre FT Callaway RM Valladares Fet al. ( 2009) Refining the stress‐gradient hypothesis for competition and facilitation in plant communities. J Ecol 97: 199– 205. Google Scholar CrossRef Search ADS Maestre FT Valladares F Reynolds JF( 2005) Is the change of plant–plant interactions with abiotic stress predictable? A meta‐analysis of field results in arid environments. J Ecol 93: 748– 57. Google Scholar CrossRef Search ADS Mateos E Santos X Pujade-Villar J( 2011) Taxonomic and functional responses to fire and post-fire management of a Mediterranean Hymenoptera community. Environ Manage 48: 1000– 12. Google Scholar CrossRef Search ADS PubMed Moreno JM Zuazua E Pérez Bet al. ( 2011) Rainfall patterns after fire differentially affect the recruitment of three Mediterranean shrubs. Biogeosciences 8: 3721– 32. Google Scholar CrossRef Search ADS Menges ES Quintana-Ascencio PF( 2004) Population viability with fire in Eryngium cuneifolium: deciphering a decade of demographic data. Ecol Monogr 74: 79– 99. Google Scholar CrossRef Search ADS Müller J Deil U( 2001) Ecology and structure of Drosophyllum lusitanicum (L.) Link populations in the south-western of the Iberian Peninsula. Acta Bot Malacit 26: 47– 68. Ojeda F Marañón T Arroyo J( 1996) Postfire regeneration of a Mediterranean heathland in southern Spain. Int J Wildland Fire 6: 191– 8. Google Scholar CrossRef Search ADS Ojeda F Pausas JG Verdú M( 2010) Soil shapes community structure through fire. Oecologia 163: 729– 35. Google Scholar CrossRef Search ADS PubMed Onset Computer Corporation( 2013) HOBO Data Loggers . Madrid. http://www.onsetcomp.com/products/data-loggers/u23-001 (10 February 2015, date last accessed). Ortega-Olivencia A Carrasco JP Devesa JA( 1995) Floral and reproductive biology of Drosophyllum lusitanicum (L.) Link (Droseraceae). Bot J Linn Soc 118: 331– 51. Paniw M Quintana-Ascencio PF Ojeda Fet al. ( 2016) Accounting for uncertainty in dormant life stages in stochastic demographic models. Oikos , 10.1111/oik.03696. Paniw M Quintana-Ascencio PF Ojeda Fet al. ( 2017) Interacting livestock and fire may both threaten and increase viability of a fire-adapted Mediterranean carnivorous plant. J Appl Ecol , 10.1111/1365-2664.12872. Paniw M Salguero-Gómez R Ojeda F( 2015) Local-scale disturbances can benefit an endangered, fire-adapted plant species in Western Mediterranean heathlands in the absence of fire. Biol Conserv 187: 74– 81. Google Scholar CrossRef Search ADS Peñuelas J Prieto P Beier Cet al. ( 2007) Response of plant species richness and primary productivity in shrublands along a north–south gradient in Europe to seven years of experimental warming and drought: reductions in primary productivity in the heat and drought year of 2003. Glob Chang Biol 13: 2563– 81. Google Scholar CrossRef Search ADS Potts SG Vulliamy B Dafni A Ne’eman Get al. ( 2003) Response of plant‐pollinator communities to fire: changes in diversity, abundance and floral reward structure. Oikos 101: 103– 12. Google Scholar CrossRef Search ADS Quintana-Ascencio PF Menges ES Weekley CV( 2003) A fire-explicit population viability analysis of Hypericum cumulicola in Florida rosemary scrub. Conserv Biol 17: 433– 49. Google Scholar CrossRef Search ADS Salces-Castellano A Paniw M Casimiro-Soriguer Ret al. ( 2016) Attract them anyway: benefits of large, showy flowers in a highly autogamous, carnivorous plant species. AoB PLANTS 8: plw017. Schulze W Schulze ED Schulze Iet al. ( 2001) Quantification of insect nitrogen utilization by the venus fly trap Dionaea muscipula catching prey with highly different isotope signatures. J Exp Bot 52: 1041– 9. Google Scholar CrossRef Search ADS PubMed Sedláková I Chytrý M( 1999) Regeneration patterns in a Central European dry heathland: effects of burning, sod-cutting and cutting. Plant Ecol 143: 77– 87. Google Scholar CrossRef Search ADS Soliveres S Maestre FT Berdugo Met al. ( 2015) A missing link between facilitation and plant species coexistence: nurses beneﬁt generally rare species more than common ones. J Ecol 103: 1183– 9. Google Scholar CrossRef Search ADS Verdú M Rey PJ Alcantara JMet al. ( 2009) Phylogenetic signatures of facilitation and competition in successional communities. J Ecol 97: 1171– 80. Google Scholar CrossRef Search ADS Ver Hoef JM Boveng PL( 2007) Quasi-Poisson vs. negative binomial regression: how should we model overdispersed count data? Ecology 88: 2766– 72. Google Scholar CrossRef Search ADS PubMed Vilà M Sardans J( 1999) Plant competition in Mediterranean‐type vegetation. J Veg Sci 10: 281– 94. Google Scholar CrossRef Search ADS Vilà M Terradas J( 1995). Effects of competition and disturbance on the resprouting performance of the Mediterranean shrub Erica multiflora L. (Ericaceae). Am J Bot 82: 1241– 8. Google Scholar CrossRef Search ADS Vuong HQ( 1989) Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica 57: 307– 33. Google Scholar CrossRef Search ADS Wright A Schnitzer SA Reich PB( 2014) Living close to your neighbors: the importance of both competition and facilitation in plant communities. Ecology 95: 2213– 23. Google Scholar CrossRef Search ADS PubMed Yates CJ Ladd PG( 2010) Using population viability analysis to predict the effect of ﬁre on the extinction risk of an endangered shrub Verticordia ﬁmbrilepis subsp. ﬁmbrilepis in a fragmented landscape. Plant Ecol 211: 305– 19. Google Scholar CrossRef Search ADS Zeppel MJ Harrison SP Adams HDet al. ( 2015) Drought and resprouting plants. New Phytol 2: 583– 9. Google Scholar CrossRef Search ADS © The Author(s) 2017. Published by Oxford University Press on behalf of the Institute of Botany, Chinese Academy of Sciences and the Botanical Society of China. All rights reserved. For permissions, please email: email@example.com
Journal of Plant Ecology – Oxford University Press
Published: Jun 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera