Oikos 119: 1172–1180, 2010 doi: 10.1111/j.1600-0706.2009.18148.x © 2009 The Authors. Journal compilation © 2010 Oikos Subject Editor: Heikki Setälä. Accepted 13 November 2009

Disturbance influences the outcome of plant–soil biota interactions in the invasive Acacia longifolia and in native species Luís M. Carvalho, Pedro M. Antunes, M. Amélia Martins-Loução and John N. Klironomos L. M. Carvalho ([email protected]), P. M. Antunes, and J. N. Klironomos, Dept of Integrative Biology, Univ. of Guelph, Guelph, Ontario, N1G 2W1, Canada. – M. Amélia Martins-Loução and present address for LMC, Faculdade de Ciências, Centro de Biologia Ambiental, Univ. de Lisboa, Edifício C2, Campo Grande, PT–1749-016 Lisboa, Portugal. PMA also at: Dept of Biology, Algoma Univ., Sault Ste. Marie, Ontario, P6A 2G4, Canada. JNK also at: Biology and Physical Geography Unit, Univ. of British Columbia Okanagan, Kelowna, BC V1V 1V7, Canada.

Interrelated causes of plant invasion have been gaining increasing recognition. However, research on this subject has mainly focused around conceptual models. Here we explore whether plant–soil biota feedbacks and disturbance, two major factors capable of facilitating invasive plants in introduced ranges, interact to preferentially benefit exotics compared to native plants. We investigated the influence of fire disturbance on plant–soil biota interactions for the invasive Acacia longifolia and two dominant natives (Cytisus striatus and Pinus pinaster) in Portuguese dune systems. In the first experiment, we grew exotic and native plants in soil inoculated with soil biota from unburned or recently burned soils collected in an area with small invasion intensity by A. longifolia. Soil biota effects on the exotic legume A. longifolia changed from neutral to positive after fire, whereas the opposite outcome was observed in the native legume C. striatus, and a change from negative to neutral effects after fire occurred in the native P. pinaster. Fire reduced mycorrhizal colonization in all species and rhizobial colonization in C. striatus but not in A. longifolia. In the second experiment, we grew the exotic and native plants with conspecific and heterospecific soil biota from undisturbed soils (area with low invasion intensity by A. longifolia), and from post-fire soils (area affected by a fire ~12 years ago and currently heavily invaded by A. longifolia). The exotic benefited more from post-fire than from undisturbed soil biota, particularly from those associated with natives. Natives did not experience detrimental effects with invasive-associated soil biota. Our results show that fire disturbance affected the functional interactions between soil biota and plants that may benefit more the exotic than some native species. Disturbance may open a window of opportunity that promotes invader success by altering soil enemy and mutualistic impacts.

Understanding the ecological processes that regulate plant invasions is essential to explain why some exotic species become invasive. Interactions between plants and soil biota have recently been shown to contribute to the success of exotic plants in introduced habitats. There is evidence that some invasive plants display less negative, or more positive, plant–soil feedbacks than native plants in their new geographic range, likely due to enemy release (Klironomos 2002, Callaway et al. 2004, Reinhart and Callaway 2004, van der Putten et al. 2005, 2007). Invasive plants may also or otherwise experience novel or stronger positive effects from soil mutualists in introduced ranges (enhanced mutualisms hypothesis; Richardson et al. 2000, Reinhart and Callaway 2006). However, the role of soil biota as facilitators of plant invasions has not been consistently confirmed (Colautti et al. 2004, Levine et al. 2004). The effects of both soil enemies and mutualists on exotic plants will also be probably dependent on the spatial and temporal variation of abiotic conditions (Agrawal et al. 2005). Invasion success of introduced populations may be dependent on the simultaneous action of abiotic and biotic conditions to create a window of opportunity where the 1172

invaders can make advantages in native communities (the invasion opportunity windows hypothesis; Johnstone 1986, Agrawal et al. 2005). Recently, it has been suggested that disturbance, a major factor driving invasion, can interact with plant–soil biota feedbacks, thereby influencing plant invasions (Blumenthal 2005, Kulmatiski and Kardol 2008). The availability of limiting resources may increase following disturbance, which may favour the success of fast-growing exotic species. These species are adapted to high resource availability but invest weakly on defensive mechanisms. For this reason they tend to be highly susceptible to enemies in their native ranges, consequently benefiting significantly from enemy release upon moving to a novel range (resource– enemy release hypothesis, Blumenthal 2005). Besides increasing resource availability, disturbance also has profound direct impacts on soil biological activity. The disturbance-contingent niche creation model (DCNC, Kulmatiski and Kardol 2008) predicts that the suppression of soil mutualistic and pathogenic activity induced by disturbance will benefit the growth of early-successional, fast-growing species relative to late-successional species, and that the subsequent growth of fast-growing exotics in post-disturbance conditions will

result in soil biotic communities that further promote the growth of exotics. The DCNC model integrates disturbance effects with resource availability, enemy release and enhanced mutualisms hypotheses to explain the dominance and persistence of exotic species in introduced sites. These recent theoretical considerations uncover a need for empirical data on interactive effects of plant–soil biota feedbacks and disturbance on the growth of exotic and native plants. Fire is recognized as one of the most important disturbance events affecting the dynamics of plant communities in Mediterranean-climate regions (Moreno and Oechel 1994, Clemente et al. 1996). There is evidence that many exotic plant species invade and dominate burnt areas (D’Antonio and Vitousek 1992, Hobbs and Huenneke 1992). Fire disturbance may favour exotic species by removing competitively superior native species and/or increasing nutrient and light availability (Keeley et al. 2005). But fire can also affect soil pathogenic and mutualistic communities, altering their structure and functioning (Vásquez et al. 1993, Rutigliano et al. 2007), which may influence the direction or magnitude of soil biotic feedbacks on exotic and native species. To our knowledge, it is not known if changes in soil communities induced by fire open a ‘window of opportunity’ to invasion by differently altering antagonistic and mutualistic interactions on exotics relative to natives. The main objective of this study was to determine whether fire disturbance differentially influences plant–soil biota interactions in exotic versus native plant species. We used the fast-growing legume Acacia longifolia, Sydney golden wattle, which is a highly invasive exotic species, particularly after the occurrence of fire disturbance. This species is native to coastal dune systems from southeastern Australia and it has become invasive in several coastal Mediterranean-type ecosystems worldwide, such as in Portugal, Spain, South Africa, United States and New Zealand. In Portugal, many coastal dunes disturbed by fires have been sensitive to the spread and dominance of A. longifolia (Marchante et al. 2003). We also used the legume shrub Cytisus striatus, striated broom or Portuguese broom, and the maritime pine Pinus pinaster, two abundant native species in the Portuguese coastal dune systems typically invaded by A. longifolia (Marchante et al. 2004). Both A. longifolia and C. striatus establish mycorrhizal and N-fixing rhizobial symbioses, whereas P. pinaster establishes mycorrhizal symbioses only. We conducted two experiments. In the first experiment (‘short-term effects of fire’), we investigated whether the growth of the exotic and native plants were differentially affected by short-term fireinduced changes in the soil biota. In this experiment we hypothesized that the exotic A. longifolia would benefit more from the soil biota of fire-disturbed soils than from those of fire-free soils. We base this hypothesis on (1) the suggestion that a general reduction of soil biological activity (as a result of disturbance) benefits fast-growing exotics (Kulmatiski and Kardol 2008), and (2) evidence that native species tend to be generally more dependent on the resident mutualisms than fast-growing invasive species that thrive in post-disturbance conditions (Vogelsang and Bever 2009). In the second experiment (‘long-term effects of fire’), we investigated whether the influence of host soil history (conspecific vs heterospecific soil sources) on net soil biota effects (plant–soil biota feedbacks) changed after invasion by A. longifolia upon a fire

disturbance that occurred 12 years before. Acacia longifolia invasion is known to change soil chemical and microbial processes, which is suggested to generate positive plant–soil feedbacks (Marchante et al. 2008). The subsequent invasion and dominance of a plant such as A. longifolia after fire disturbance may condition soil pathogenic and mutualistic communities consistent with positive feedback favouring their own plant growth (Wolfe and Klironomos 2005, Jordan et al. 2008, Kulmatiski and Kardol 2008). Conversely, it may lead to growth reductions on natives by supporting to a lesser degree their associated mutualistic communities (Vogelsang and Bever 2009) or by accumulating pathogens that are more harmful to them (Eppinga et al. 2006). In this second experiment we hypothesized that the invasive A. longifolia would benefit the most from conspecific soil biota from post-fire soils, whereas the natives would experience the largest growth detrimental effects under these same set of soil biotic conditions.

Material and methods Plant species The invasive Acacia longifolia and the natives Cytisus striatus and Pinus pinaster, which co-occur in the Portuguese coastal dune systems, were selected for this study. Acacia longifolia is a leguminous shrub or tree native to southeastern Australia growing to a height of 10 m. This species is considered one of the most invasive species in the dune ecosystems of Portugal (Marchante et al. 2003). It was introduced in Portugal in the early 1900s to reduce sand erosion and stabilize the dunes. However, in just a few decades this species had spread extensively in some coastal areas of Portugal, particularly after fire disturbance, forming dense stands and seriously threatening the native plant communities. High seed production with high viability and longevity, stimulation of germination by fire, and relative fast growth are factors that have been attributed to the successful establishment of A. longifolia in non-native areas (Crawley 1997). Cytisus striatus is a Portuguese native leguminous shrub which can be 2–3 m tall. These two legume species can establish both N-fixing rhizobial and arbuscular mycorrhizal (AM) symbioses. Pinus pinaster is a coniferous tree growing up to 20–35 m, native to the Mediterranean area including Portugal, and establishing ectomycorrhizal (ECM) symbioses. Soil and seed collection Soils were collected at the end of September 2005 in coastal dune systems halfway the Portuguese coastline. For the first experiment (‘short-term fire effects’), soils were collected in ‘Osso da Baleia’ dune system in a recently intensively burned (August 2005) area (44°24’N, 05°09’W) and an adjacent unburned area (44°28’N, 05°09’W), which has not been affected by fire in decades. Cytisus striatus and P. pinaster dominated these sites with individuals of A. longifolia cooccurring with the native species. In the unburned area soils were collected in the root zone of individuals of each plant species (A. longifolia, C. striatus and P. pinaster), and in the burned area soils were collected in the proximity of burnt 1173

stems of these three plant species. Ten samples from each soil type were collected and mixed. For the second experiment (‘long-term fire effects’), soils were collected in ‘Tocha’ dune system in an area (44°66’N, 05°15’W) that was affected by a fire in 1993 (post-fire soils) and is now highly invaded by A. longifolia (with dense and large monoculture stands, whereas native species are present in very low abundance), and an area (44°68’N, 05°15’W) for which there were no records of fire occurrence (non-fire soils) and where individuals of A. longifolia co-occur with native species, i.e. small invasion intensity. Soils were collected in non-fire and post-fire sites in the root zone of individuals of each of the three plant species. Ten samples from each soil sample were collected and mixed. All soil samples were collected from a depth of 0–20 cm upon removal of the litter layer, and each sample consisted of approximately 2 kg fresh weight. Aseptic precautions (e.g. sterilization with ethanol) were taken during sampling and handling of each soil sample to avoid cross contamination. Soils were left at room temperature for 24 h and then shipped to the Univ. of Guelph, Canada, where the experiments were conducted. Seeds of A. longifolia and C. striatus were collected in the summer of 2005 at ‘Tocha’ sites, and seeds of P. pinaster were purchased from CENASEF, Direcção-Geral dos Recursos Florestais, Portugal (collected from dune systems in the centre of Portugal). Experimental design The first experiment (‘short-term fire effects’) involved a fullfactorial design with two soil inoculum origins (unburned and burned), two soil inoculum treatments (non-sterilized and sterilized) and three growing seedling species (A. longifolia, C. striatus and P. pinaster) with eight replicates per treatment combination. The second experiment (‘long-term fire effects’) had a fully factorial design with two soil inoculum origins (non-fire and post-fire), three sources of soil inoculum (one conspecific and two heterospecifics), two soil inoculum treatments (non-sterilized and sterilized) and three growing seedling species (A. longifolia, C. striatus and P. pinaster) with eight replicates per treatment combination. The two experiments were conducted simultaneously in a greenhouse at the Univ. of Guelph with average temperatures of 25/20°C (day/night) and a 14-h photoperiod. Seeds of all species were surface sterilized with 10% bleach (10 min for A. longifolia and C. striatus, and 30 min for P. pinaster), and then thoroughly rinsed with deionized water. Seeds of A. longifolia and C. striatus were placed in boiling water and left for 18 h until the water was cold, while P. pinaster seeds were left in water at 4°C for four days. Seeds were then germinated in the greenhouse in trays filled with a 1:1 mixture of autoclaved (twice at 121°C for 20 min in two consecutive days) turface and vermiculite. One-half of each soil type were sterilized by triple autoclaving (121°C for 1 h in three consecutive days) and allowed to stabilize for two weeks. One hundred ml soil (either non-sterilized or sterilized) was mixed with autoclaved (twice at 121°C for 1 h in two consecutive days) silica sand at a ratio of 1:9 into 1-liter plastic pots. The first 2-cm layer at the top of the pot consisted of only sand to help preventing cross contamination between 1174

pots. The background soil (sterilized silica sand) represented 90% of the total soil in the pot, and therefore minimizes potential confounding effects of nutrient release (Troelstra et al. 2001). Germinated seedlings (two weeks old) of the three plant species were separately planted in the pots, and grown for four months from January through April 2006. Each experimental unit (pot) was positioned in a completely randomized design on a greenhouse bench. The plants were watered with tap water as necessary and fertilized with 0.25strength Hoagland modified solution (Hoagland and Arnon 1939) with a reduced P concentration (0.1 mM) on a weekly basis. Measurements Plants were harvested at the end of the experiments, and shoots and roots were separated, cleaned and their fresh weights were determined. Roots of A. longifolia and C. striatus were examined for quantification of N-fixing nodules. A 0.3–0.5 g (fresh weight) subsample was removed from each root system and stored in 50% ethanol for later determination of mycorrhizal colonization. Dry weights of the shoot and of the remaining roots were determined after drying at 60°C for five days. The ratio of dry to fresh weight of the remaining roots was used to calculate the dry weight of the root subsample, and then of the entire root sample. The subsamples of stored roots of A. longifolia and C. striatus were cleared in 10% potassium hydroxide, and stained with Chlorazol Black E (Brundrett et al. 1984). Percent root colonization by AM fungi was determined using the magnified intersections method (McGonigle et al. 1990). The subsamples of stored roots of P. pinaster were placed in petri dishes and examined for ECM colonization using a stereo microscope. Percent root colonization by ECM fungi was calculated as a percentage of mycorhizal root tips (total number of mycorrhizal tips/total number of root tips  100). The shoot content of N and P were determined after digestion (Thomas et al. 1967). We calculated the net soil biota effect (i.e. the effect of soil sterilization) on plant dry weight as (XNS–XST)/(XNSXST), where XNS is plant dry weight in non-sterile soil inoculum and XST is the mean plant dry weight in sterile soil inoculum (adapted from Armas et al. 2004), assuring that the net effect can only range from –1 to 1, with positive values indicating positive effects of soil biota on plant dry weight, and negative values indicating negative effects. The net soil biota effect on N and P content was calculated by the same formula replacing plant dry weight by N or P content. We applied the bootstrap procedure to estimate means and errors for net soil biota effects (Efron and Tibshirani 1986). We took a bootstrap sample of the plant biomass values for a non-sterile treatment, and a second bootstrap sample of the corresponding sterile treatment, calculated the net soil biota index, and repeated 999 times (Klironomos 2003) using Excel (Microsoft) and the Excel Add-in from Resampling Stats software (Statistics.com). Statistical analyses In the two experiments, we tested whether estimates of net soil biota effects differed significantly from zero. If 95% of

these bootstrap estimates fell either above or below zero, the estimates were considered significantly different from zero. Data from the first experiment were analyzed with a two-way ANOVA for net soil biota effect on total dry weight, and on N and P content, for AM colonization and number of N-fixing nodules in A. longifolia and C. striatus, and with a t-test for ECM colonization in P. pinaster. Data from the second experiment were also analyzed with a three-way ANOVA for net soil biota effect on plant biomass, AM colonization and number of N-fixing nodules in A. longifolia and C. striatus, and with a two-way ANOVA for ECM colonization in P. pinaster. For mycorrhizal colonization and number of nodules the sterilized treatments were excluded from the analyses since the results were null. Significant differences among soil biota origin and among soil biota source treatment means were determined using t-tests (p  0.05) and among plant species using the Tukey post-hoc comparison test (p  0.05). Prior to statistical analysis, data were transformed (arcsin square root for percentage variables and logarithmic for other variables) as necessary to satisfy the assumptions of the ANOVA (Zar 1984). SPSS 17.0 software (Chicago, IL) was used for all statistical analysis.

Figure 1. Net soil biota effect on plant biomass, calculated as (dry weight in non-sterilized soil – dry weight in sterilized soil)/(dry weight in non-sterilized soil  dry weight in sterilized soil), for A. longifolia, C. striatus and P. pinaster seedlings growing with soil biota from unburned or burned sites. Bars represent means  SE based on 1000 bootstrap estimates. For each plant species, different letters indicate significant differences (p  0.05) among soil biota inoculum sites; significant differences of each of the bars from zero are indicated with asterisks (*** p  0.001).

Results

negative effect found with the soil biota from unburned soils disappeared with the soil biota from burned soils. There were significant effects of soil biota origin (F1,23  37.97, p  0.001) and plant species (F1,23  12.38, p  0.002) on the percentage of root length colonization by AM fungi in A. longifolia and C. striatus. No significant interaction effect was found (F1,23  0.44, p  0.514). There was a drastic reduction in AM colonization with soil biota from burned soils compared to the one from unburned soils (Table 1). In P. pinaster a small, but significant, reduction was found in ECM colonization (t-test, t13  3.37, p  0.005; Table 1). The effect of soil biota origin in the number of N-fixing nodules was different for the two plant species (F1,27  4.48, p  0.044 for soil biota origin effect; F1,27  7.47, p  0.011 for plant species effect; and F1,27  1.23, p  0.278 for origin  plant species interaction effect). For C. striatus it was significantly lower with soil biota from burned soils compared with soil biota from unburned soils, but no significant differences were found in A. longifolia among soil origins (Table 1). The interaction between soil biota origin and plant species, as well as each factor alone, significantly influenced the net soil biota effects on P (F1,24  4.51, p  0.044 for soil biota origin effect; F2,24  8.04, p  0.002 for plant species effect; and

For both experiments in each plant species tested there were no significant differences (p  0.05) in plant biomass among the sterilized soil inocula treatments (data not shown), indicating there were no soil nutritional effects regarding the soil inoculum origin, and therefore, all significant effects were likely due to the soil biota. Experiment 1. Short-term fire effects The net soil biota effect on plant biomass varied among the three plant species (F2,40  10.70, p  0.001 for plant species effect), but in a different way according to the origin of the soil biota (F2,40  7.97, p  0.001 for soil biota origin  plant species interaction effect; no main effect was found for soil biota origin, F1,40  0.08, p  0.391). Soil biota effects varied significantly (Tukey post-hoc test, p  0.05) among plant species with soil biota from unburned soils but not from burned soils. In the exotic A. longifolia, soil biota from unburned soils had a neutral effect on plant biomass, whereas soil biota from burned soils had a significant positive effect (Fig. 1). Conversely, for C. striatus there was a significant positive effect of soil biota from unburned soils and a neutral effect with soil biota from burned soils. For P. pinaster, the

Table 1. Percentage of root colonization by AM fungi for A. longifolia and C. striatus, and ECM fungi for P. pinaster, and number of rhizobial nodules for A. longifolia and C. striatus, in seedlings growing with soil biota inoculum from unburned and burned sites for the experiment 1. Values represent the average with  1 SE. p-value indicate significance of differences (t-test) between unburned and burned origins of soil inoculum. Mycorrhizal colonization (%)

Number of nodules

Seedling species

Unburned

Burned

p

Unburned

Burned

p

Acacia longifolia Cytisus striatus Pinus pinaster

18  5 40  7 86  3

11 93 70  4

0.001 0.001 0.005

12  2 93

10  1 31

0.450 0.042

1175

F2,24  9.92, p  0.001 for origin  plant species interaction effect) and N (F1,24  9.07, p  0.006 for soil biota origin effect; F2,24  4.38, p  0.024 for plant species effect; and F2,24  6.80, p  0.005 for origin  plant species interaction effect). Pinus pinaster showed the major differences in nutrition depending on soil biota origin (unburned or burned) (Fig. 2). For this species the soil biota from unburned soils had negative effects on P and N content, whereas for soil biota from burned soils the effects were positive. In A. longifolia the soil biota had neutral effects on nutrient contents, except for N with inoculum from burned soils for which a positive effect was found. For C. striatus positive effects on nutrient contents were found for both unburned and burned soil origins.

sources (T-test, p  0.05; Fig. 3a). Conversely, the growth of the native species did not depend on fire history regardless of inocula source, except for P. pinaster growing with soil inoculum from A. longifolia (Fig. 3b–c). In conspecific soil, the exotic A. longifolia had neutral net soil biota effects on plant biomass with soil biota from both non-fire and postfire origins. In the two heterospecific soils, A. longifolia experienced neutral effects with soil biota from non-fire soils and

Experiment 2. Long-term fire effcts The three-way ANOVA indicated significant seedling species effects (F2,122  7.72, p  0.001) and a soil origin  species interaction (F2,122  4.40, p  0.014). All other effects were not significant (p  0.05). Analyzing the effects separately by species, A. longifolia experienced larger positive effects growing with soil biota from post-fire than from non-fire sites, although only significant for soil biota from heterospecifics

Figure 2. Net soil biota effect on P and N shoot content, calculated as (shoot element content in non-sterilized soil – shoot element content in sterilized soil)/(shoot element in non-sterilized soil  shoot element in sterilized soil), for A. longifolia, C. striatus and P. pinaster seedlings growing with soil biota from unburned or burned sites. Bars represent means  SE based on 1000 bootstrap estimates. For each plant species, different letters indicate significant differences (p  0.05) among soil biota inoculum sites; significant differences of each of the bars from zero are indicated with asterisks (* p  0.01; ** p  0.01; *** p  0.001).

1176

Figure 3. Net soil biota effect on plant biomass, calculated as (dry weight in non-sterilized soil – dry weight in sterilized soil)/(dry weight in non-sterilized soil  dry weight in sterilized soil), for A. longifolia (a), C. striatus (b) and P. pinaster (c) seedlings growing with soil biota inoculum of conspecific and heterospecific source from non-fire or post-fire sites. Bars represent means  SE based on 1000 bootstrap estimates. For each plant species and rhizospheric soil inoculum source within each graph, different letters indicate significant differences (p  0.05) among sites; significant differences of each of the bars from zero are indicated with asterisks (* p  0.05; ** p  0.01; *** p  0.0001).

significant positive effects with soil biota from post-fire soils (Fig. 3a). Cytisus striatus had positive net soil biota effects that were significantly different from zero when grown with A. longifolia rhizospheric soil biota from both non-fire and post-fire origins, and with conspecific soil biota from the non-fire soils (Fig. 3b). Pinus pinaster had significant positive effects with soil biota from the heterospecific C. striatus of non-fire soils, and effects not significantly different from zero in all other treatment combinations (Fig. 3c). Percent AM fungal colonization in A. longifolia and C. striatus varied depending on the origin (F1,73  11.84, p  0.001) and origin  source interaction (F2,73  2.73, p  0.049) of the soil biota inoculum (F2,73  3.16, p  0.049), and on source  seedling species interaction (F2,73  6.26, p  0.003). All other effects were not significant (p  0.05). There were significant differences between non-fire and post-fire soils for conspecific source in A. longifolia and for A. longifolia heterospecific sources in C. striatus (Table 2). Overall, percent AM fungal colonization was higher with soil inoculum from post-fire than from non-fire soils. In P. pinaster there was a significant origin  source of soil biota interaction in ECM fungal colonization (F2,40  3.60, p  0.036); no significant main effects were observed (p  0.05). There were no significant differences in ECM fungal colonization level among origins of soil inoculum (Table 2). For the number of N-fixing nodules there were significant effects of origin (F1,78  4.35, p  0.040) and source (F2,78  8.31, p  0.001) of the soil biota inoculum, and plant species (F1,78  35.85, p  0.001). No interaction effects were found (p  0.05). Acacia longifolia had significant lower number of nodules with soil inoculum from post-fire than from non-fire soils, and in both two soil origins C. striatus had more root nodules when growing with conspecific than with heterospecific soil inocula (Table 2).

Discussion This study shows that fire disturbance changes the outcome of plant–soil biota interactions in exotic and native plant species. The invasive plant generally experienced higher positive feedbacks with soil biota from disturbed soils compared to feedbacks in non-disturbed soils. The results from the short-

term fire effects experiment support our hypothesis that the invasive plant species experiences greater positive net effects on growth from soil biota impacted by fire than from soil biota of unburned soils (effects changed from neutral to positive after fire). The results from the long-term fire effects experiment also indicated that the invasive plants generally experienced positive feedbacks with soil biota in post-disturbed soils even ~12 years after fire. However, our hypothesis that natives would benefit less from fire disturbed soil biota than the exotic was only supported for one of the native species. Cytisus striatus, which belongs to the same family as A. longifolia, did not benefit from soil biota from recently burned soils (effects changed from positive to neutral after fire). Conversely, in P. pinaster, fire induced a beneficial change in net soil biota effects changing from negative to neutral. Our results for A. longifolia suggest a release from soil enemies after fire. The neutral net effects with soil biota from non-fire soils (two experiments) and from conspecific post-fire soils (‘long-term fire influence’ experiment) suggest an accumulation of enemies as A. longifolia establishes; enemies’ effects which may however be offset by the positive effects of mutualists, particularly N-fixing bacteria. It is possible that introduced plants acquire new enemies or reacquire their natural enemies in time (Mitchell and Power 2003) that may limit their long-term spread and persistence (Nijjer et al. 2007). Reinhart and Callaway (2004) also reported inhibitory soil biota effects for the invasive Acer tree species in introduced ranges. Further studies are needed since we do not know of any study reporting soil generalist or specialist pathogenic effects on A. longifolia in native or introduced ranges, only aboveground insect herbivory effects (Dennill 1987). We hypothesize that, if present, these negative interactions may contribute to constrain the abundance of A. longifolia (biotic resistance from the local soil biota), and a reduction in pathogen load induced by disturbance may be necessary to increase A. longifolia performance. We cannot exclude the hypothesis that a highly effective rhizobial population colonizing A. longifolia after the fire may be also involved in the shift from neutral to positive in soil biota effects, which can be supported by the increase in the net soil biota effects on N content and a non-significant decrease in the number of nodules. This species can be colonized by a large diversity of rhizobia in introduced and native

Table 2. Percentage of root colonization by AM fungi for A. longifolia and C. striatus, and ECM fungi for P. pinaster, and number of rhizobial nodules for A. longifolia and C. striatus, in seedlings growing with soil biota inoculum of conspecific and heterospecific soil sources from non-fire or post-fire sites for the experiment 2. Values represent the average with  1 SE. p-value indicate significance of differences (t-test) between non-fire and post-fire sites. Mycorrhizal colonization (%) Seedling species

Number of nodules

Inoculum source

Non-fire

Post-fire

p

Non-fire

Post-fire

p

A. longifolia

A. longifolia C. striatus P. pinaster

92 37  6 26  3

27  4 40  6 33  8

0.001 0.694 0.423

16  2 14  2 11  1

12  1 71 12  2

0.103 0.006 0.615

C. striatus

A. longifolia C. striatus P. pinaster

36  9 43  9 43  10

66  3 43  7 64  3

0.026 0.839 0.058

10  4 52 32

93 21 21

0.791 0.220 0.552

P. pinaster

A. longifolia C. striatus P. pinaster

85  3 83  3 74  3

70  7 84  4 82  2

0.076 0.791 0.090

1177

ranges varying in their symbiotic effectiveness (Lawrie 1983, Barnet et al. 1985, Rodríguez-Echeverría et al. 2007). The ability to nodulate in post-fire soil conditions may be very important for the invasion of A. longifolia after fire, since the invasibility of this species may be determined by its capacity to nodulate profusely in different types of soils (RodríguezEcheverría et al. 2009). AM fungi do not seem to play a role in the early growth of A. longifolia in burned soil conditions, since there was a large reduction in AM colonization in burned soils without a reduction in plant growth. Conversely, in the native C. striatus, the reduction of soil biota effects on plant biomass after fire accompanied the reduction in P and N content, AM colonization and nodulation. These results suggest that the impact of fire on soil mutualisms affected the seedling establishment of C. striatus. Moreover, they suggest that although both native and exotic legume species may benefit from soil enemy release following fire, A. longifolia experiences higher benefits because is less dependent on the resident mutualistic community, mainly on the mycorrhizal community, than the native C. striatus species. For the native P. pinaster, release of pathogenic effects was probably more significant, since after fire there was an increase in soil biota effects in plant biomass and P and N content, and only a small decrease in ECM colonization. The differences in net effects of soil biota from nonfire sites in P. pinaster growth between the two experiments (negative effects in experiment 1 and neutral effects in experiment 2) seems to indicate some variability among locations in the effects of those potential pathogenic soil communities. Fire can increase nutrient resource availability in the short-term (Gimeno-García et al. 2000), which according to a recent conceptual model developed by Blumenthal (2005) may favours the establishment of fast-growing exotic species also experiencing enemy release. Although our experiments were not designed to directly test whether fire affects exotic and native plant growth through changes in soil nutrient availability, the changes in soil biota density and activity that we observed may be related to alterations in soil nutrient resources after fire. Nevertheless, the fire even is likely to directly affect soil microbial communities. Which groups of soil organisms may be most affected is not known, and this can vary depending on the severity of fire (Neary et al. 1999). It has been reported that fungi are more affected by fire than bacteria (Vásquez et al. 1993, Rutigliano et al. 2007). Even though we did not assess soil community diversity, our results suggest that AM fungal abundance was particularly affected by fire. The large reduction in AM root colonization suggests a decrease in viable AM fungal propagules or loss of propagules associated with post-burning erosion (O’Dea 2007). The results showing higher benefits in the exotic growth with soil biota from post-disturbed than from non-disturbed soils supports the prediction of the DCNC conceptual model that reduction in soil biological activity benefits the growth of fast-growing exotics (Kulmatiski and Kardol 2008). We have also hypothesized that over time since fire the spread and dominance of the exotic would result in positive feedbacks for their own growth and negative to natives (Kulmatiski and Kardol 2008). There is some evidence that invasive plants can induce direct modification on soil communities generating self-facilitative effects on exotics and antagonistic effects on native species (Mummey and Rillig 1178

2006, Jordan et al. 2008, Vogelsang and Bever 2009). However, in our study, we did not find evidence to support this. In the long-term fire effects experiment, in post-fire soils, which corresponds to a highly invaded system, A. longifolia did not experience larger growth benefits when associated with conspecific compared to heterospecific soil biota. As for the natives, growth was unchanged regardless of species origin. Furthermore, root colonization by mycorrhizal and rhizobial mutualists in natives did not decrease with A. longifolia soil biota. Therefore, in contrast to our hypothesis for natives, soil biota from A. longifolia post-fire soils did not have detrimental effects on seedling growth of the native species C. striatus and P. pinaster. The positive effect of soil biota associated with heterospecifics in the growth of A. longifolia several years following fire indicates that natives may have positive effects on A. longifolia establishment mediated through changes to their soil biota. Reinhart and Callaway (2004) have also reported benefits in invasive Acer species induced by the soil biota associated with native species. The effects we found even several years after fire may be explained by changes in soil processes and pools due to the vast spread of A. longifolia that occurred since then. Invaded A. longifolia areas show increases in C and N pools associated with increases in microbial biomass and activity (Marchante et al. 2008), and therefore A. longifolia may ameliorate the conditions for the establishment of their offspring, particularly in soil patches of other species where they can escape accumulation of their own enemies. Our study evaluates the effects of soil biota on individual plant growth, and therefore, we do not know if these effects turn into differences on plant population dynamics promoting the dominance of the exotic over the natives. Nevertheless, our study suggests that plant–soil biota interactions associated with a fire disturbance event may contribute to preferentially benefit the exotic legume A. longifolia over the native legume C. striatus in post-fire communities. The native P. pinaster seems, as A. longifolia, to benefit from soil biota changes following fire as it was observed in the short-term effects experience. However, over time A. longifolia may be able to successfully colonize P. pinaster patches, whereas the opposite seems to be more difficult, as suggested from the differential net soil biota effects observed in the long-term effects experiment. Other factors should also be involved in the spread of A. longifolia, particularly explaining the higher competitive advantage of the exotic against Pinus trees, and other native plants. It has been shown that A. longifolia exhibited higher competitive strength upon successful plant establishment through high shoot elongation rate and more efficient nutrient acquisition (Werner et al. 2008) that accompanied changes in nutrient inputs to soil due to Acacia leaf litter accumulation (Marchante et al. 2008). It is also important to know whether the soil biota effects observed in the invasive ranges differ from those in the native ranges of the exotic species. In a recent study with the invasive species Centaurea solstitialis, Hierro et al. (2006) found that disturbance increased plant abundance and performance more in invasive than in native ranges, and suggested that enemy release may contribute to explain those differences. In this work we employed homogenization of the soil samples within each soil type, which reduced the variance within each soil type treatment. This sample homogenization

approach is common in several studies investigating soil feedback effects in order to reduce the logistics with collecting and handling a large volume of samples, and to minimize the great variance that exists among soil samples in biological, chemical and physical properties. As a result of this pooling, the treatment means should be the same but the variance is reduced. Therefore, the findings of this study should be interpreted with caution. Further investigations accounting for spatial variability by sampling more sites and increasing replication of fire versus nonfire treatments would be needed to support generalizations of patterns found in this study. In conclusion, fire disturbance changes the outcome of plant–soil biota interactions in the invasive A. longifolia and two dominant natives of coastal Portuguese dune systems. The exotic plant experienced positive effects whereas the natives experienced neutral effects from soil biota following fire, and the long-term effects of heterospecific soil biota from natives on the exotic change to positive in soils impacted by fire. Release from soil enemies and/or enhanced rhizobial effectiveness may therefore contribute to increase growth performance of A. longifolia after fire disturbance. Our findings suggest that plant–soil biota interactions can change over time according to the occurrence of disturbances, as it was found for agricultural disturbance (Kardol et al. 2006). The results of our study also highlight the importance of considering interrelated causes in the understanding of plant invasions. Disturbance may open a window of opportunity for invasion by reducing enemy impact and/or increasing mutualistic outcomes on exotics relative to native species. Thus, exotic populations may exploit situations of jointly beneficial biotic and abiotic conditions to broadly spread and dominate over recipient communities. Acknowledgements – The authors wish to thank to Nora Magyara, Ashley Downing and Inês Carvalho for technical assistance, and Ana Júlia Pereira for field information assistance. This work was supported by the Fundação para a Ciência e a Tecnologia (FCT) of Portugal (grants SFRH/BPD/19380/2004 and 33633/2009 to LMC), the Natural Sciences and Engineering Research Council of Canada (to JNK), and in part by a project of the European Science Foundation – TECT and FCT (TECH/0001/2007).

References Agrawal, A. A. et al. 2005. Enemy release? An experiment with congeneric plant pairs and diverse above- and belowground enemies. – Ecology 86: 2979–2989. Armas, C. et al. 2004. Measuring plant interactions: a new comparative index. – Ecology 85: 2682–2686. Barnet, Y. M. et al. 1985. Biological nitrogen fixation and rootnodule bacteria (Rhizobium sp. and Bradyrhizobium sp.) in two rehabilitating sand dune areas planted with Acacia spp. – Austral. J. Bot. 33: 595–610. Blumenthal, D. 2005. Interrelated causes of plant invasion. – Science 310: 243–244. Brundrett, M. C. et al. 1984. A new method for observing the morphology of vesicular–arbuscular mycorrhizae. – Can. J. Bot. 62: 2128–2134. Callaway, R. M. et al. 2004. Soil biota and exotic plant invasion. – Nature 427: 731–733.

Clemente, A. S. et al. 1996. Demographic patterns and productivity of post-fire regeneration in Portuguese Mediterranean maquis. – Int. J. Wildland Fire 6: 5–12. Coulatti, R. I. et al. 2004. Is invasion success explained by the enemy release hypothesis? – Ecol. Lett. 7: 721–733. Crawley, M. J. 1997. Biodiversity. – In: Crawley, M. J. (ed.), Plant ecology. Blackwell, pp. 623–625. D’Antonio, C. M. and Vitousek, P. M. 1992. Biological invasions by exotic grasses, the grass/fire cycle and global change. – Annu. Rev. Ecol. Syst. 23: 63–87. Dennill, G. B. 1987. Establishment of the gall wasp Trichilogaster acaciaelongifoliae (Pteromalidae) for the biological control of Acacia longifolia in South Africa. – Agric. Ecosyst. Environ. 19: 155–168. Efron, B. and Tibshirani, R. 1986. Bootstrap methods for standard errors, confidence intervals, and other measures of statistical accuracy. – Statist. Sci. 1: 54–75. Eppinga, M. B. et al. 2006. Accumulation of local pathogens: a new hypothesis to explain exotic plant invasions. – Oikos 114: 168–176. Gimeno-García, E. et al. 2000. Changes in organic matter, nitrogen, phosphorus and cations in soil as a result of fire and water erosion in a Mediterranean landscape. – Eur. J. Soil Sci. 51: 201–211. Hierro, J. L. et al. 2006. Disturbance facilitates invasion: the effects are stronger abroad than at home. – Am. Nat. 168: 144–156. Hoagland, D. R. and Arnon, D. I. 1939. The water culture method for growing plants without soil. – Calif. Agric. Exp. Stn Circ. 347: 1–32. Hobbs, R. J. and Huenneke, L. F. 1992. Disturbance, diversity, and invasion: implications for conservation. – Conserv. Biol. 6: 324–337. Johnstone, I. M. 1986. Plant invasion windows: a time-based classification of invasion potential. – Brol. Rev. 61:369–394. Jordan, N. R. et al. 2008. Soil modification by invasive plants: effects on native and invasive species of mixed-grass prairies. – Biol. Invas. 10: 177–190. Kardol, P. et al. 2006. Temporal variation in plant-soil feedback controls succession. – Ecol. Lett. 9: 1080–1088. Keeley, J. E. et al. 2005. Alien plant dynamics following fire in Mediterranean-climate California shrublands. – Ecol. Appl. 15: 2109–2125. Klironomos, J. N. 2002. Feedback with soil biota contributes to plant rarity and invasiveness in communities. – Nature 417: 67–70. Klironomos, J. N. 2003. Variation in plant response to native and exotic arbuscular mycorrhizal fungi. – Ecology 84: 2292–2301. Kulmatiski, A. and Kardol, P. 2008. Getting plant-soil feedbacks out of the greenhouse: experimental and conceptual approaches. – In: Lüttige, U. et al. (eds), Progress in botany 69. Springer, pp. 449–472. Lawrie, A. C. 1983. Relationships among rhizobia from native Australian legumes. – Appl. Environ. Microbiol. 45: 1822–1828. Levine, J. M. et al. 2004. A meta-analysis of biotic resistance to exotic plant invasions. – Ecol. Lett. 7: 975–989. Marchante, E. et al. 2008. Short- and long-term impacts of Acacia longifolia invasion on the belowground processes of a Mediterranean coastal dune ecosystem. – Appl. Soil Ecol. 40: 210–217. Marchante, H. et al. 2003. Invasion of the Portuguese dune ecosystems by the exotic species A. longifolia (Andrews) Willd.: effects at the community level. – In: Child, L. E. et al. (eds), Plant invasions: ecological threats and management solutions. Backhuys Publisher, pp. 75–85. Marchante, H. S. et al. 2004. Recovery potential of dune ecosystems invaded by an exotic Acacia species (Acacia longifolia). – Weed Tech. 18: 1427–1433. McGonigle, T. P. et al. 1990. A new method which gives an objective measure of colonisation of roots by vesicular–arbuscular mycorhizal fungi. – New Phytol. 115: 495–501.

1179

Mitchell, C. E. and Power, A. G. 2003. Release of invasive plants from fungal and viral pathogens. – Nature 421: 625–627. Moreno, J. M. and Oechel, W. C. 1994. The role of fire in Mediterranean-type ecosystems. – Springer. Mummey, D. L. and Rillig, M. C. 2006. The invasive plant species Centaurea maculosa alters arbuscular mycorrhizal fungal communities. – Plant Soil 288: 81–90. Neary, D. G. et al. 1999. Fire effects on belowground sustainability: a review and synthesis. – For. Ecol. Manage. 122: 51–71. Nijjer, S. et al. 2007. Negative plant–soil feedbacks may limit persistence of an invasive tree due to rapid accumulation of soil pathogens. – Proc. R. Soc. Lond. B 274: 2621–2627. O’Dea, M. E. 2007. Influence of mycotrophy on native and introduced grass regeneration in a semiarid grassland following burning. – Restor. Ecol. 15: 149–155. Reinhart, K. O. and Callaway, R. M. 2004. Soil biota facilitate exotic Acer invasions in Europe and North America. – Ecol. Appl. 14: 1737–1745. Reinhart, K. O. and Callaway, R. M. 2006. Soil biota and invasive plants. – New Phytol. 170: 445–457. Richardson, D. M. et al. 2000. Plant invasions – the role of mutualism. – Biol. Rev. 75: 65–93. Rodríguez-Echeverría, S. et al. 2007. Genetic diversity of rhizobia associated with Acacia longifolia in two stages of invasion of coastal sand dunes. – Appl. Environ. Microbiol. 73: 5066–5070. Rodríguez-Echeverría, S. et al. 2009. Belowground mutualists and the invasive ability of Acacia longifolia in coastal dunes of Portugal. – Biol. Invas. 11: 651–661. Rutigliano, F. A. et al. 2007. Impact of fire on fungal abundance and microbial efficiency in C assimilation and mineralisa-

1180

tion in a Mediterranean maquis soil. – Biol. Fertil. Soils 44: 377–381. Thomas, R. L. et al. 1967. Comparison of conventional and automated procedures for N, P and K analysis of plant material using a single digestion. – Agron. J. 59: 240–243. Troelstra, S. R. et al. 2001. Interpretation of bioassays in the study of interactions between soil organisms and plants: involvement of nutrient factors. – New Phytol. 150: 697–706. van der Putten, W. H. et al. 2005. Invasive plants and their escape from root herbivory: a worldwide comparison of the rootfeeding nematode communities of the dune grass Ammophila arenaria in natural and introduced ranges. – Biol. Invas. 7: 733–746. van der Putten, W. H. et al. 2007. Soil feedback of exotic savanna grass relates to pathogen absence and mycorrhizal selectivity. – Ecology 88: 978–988. Vásquez, F. J. et al. 1993. Soil microbial populations after wildfire. – FEMS Microbiol. Ecol. 13. 93–104. Vogelsang, K. M. and Bever, J. D. 2009. Mycorrhizal densities decline in association with nonnative plants and contribute to plant invasion. – Ecology 90: 399–407. Werner, C. et al. 2008. Competitive balance between the alien invasive Acacia longifolia and native Mediterranean species. – In: Tokarska-Guzik, B. et al. (eds), Plant invasions: human perception, ecological impacts and management, Backhuys Publishers, pp. 261–275. Wolfe, B. E. and Klironomos, J. N. 2005. Breaking new ground: soil communities and exotic plant invasion. – Bioscience 55: 477–487. Zar, J. H. 1984. Biostatistical analysis. – Prentice Hall.