Allelochemicals of Pinus Halepensis As Drivers of Biodiversity in Mediterranean Open Mosaic Habitats During the Colonization Stage of Secondary Succession

J Chem Ecol DOI 10.1007/s10886-013-0239-6

Allelochemicals of Pinus halepensis as Drivers of Biodiversity in Mediterranean Open Mosaic Habitats During the Colonization Stage of Secondary Succession

Catherine Fernandez & Mathieu Santonja & Raphael Gros & Yogan Monnier & Mathilde Chomel & Virginie Baldy & Anne Bousquet-Mélou

Received: 14 September 2012 /Revised: 30 October 2012 /Accepted: 15 November 2012 # Springer Science+Business Media New York 2013

Abstract The Mediterranean region is recognized as a stimulated. Thus, microorganisms alter the expression of global biodiversity hotspot. However, over the last 50 years allelochemicals released into the ecosystem, which high- or so, the cessation of traditional farming has given way to lights their key role in chemical plant-plant interactions. strong afforestation at the expense of open habitats. Pinus The results of allelopathic experiments conducted in the halepensis Miller, known to synthesize a wide range of lab are consistent with the community patterns observed in secondary metabolites, is a pioneer expansionist species the field. These findings suggest that allelopathy is likely to colonizing abandoned agricultural land that present high shape vegetation composition and participate to the control species richness. Here, laboratory bioassays were used to of biodiversity in Mediterranean open mosaic habitats. study the potential impact of P. halepensis on plant diversity through allelopathy, and the role of microorganisms in these Keywords Allelopathy . Plant-plant interaction . interactions. Germination and growth of 12 target species Community structure . Mediterranean biodiversity . naturally present in fallow farmlands were tested according Secondary succession . Aleppo pine . Soil microorganisms to concentration of aqueous extracts obtained from shoots of young pines (aged about 5 years), with or without the presence of soil microorganisms (autoclaved or natural soil). Under the highest concentrations and autoclaved soil, more Introduction than 80 % of target species were germination and/or growth- inhibited, and only two species were non-sensitive. Under The Mediterranean basin is characterized by a long history more natural conditions (lower extracts concentrations and of farming that has profoundly modified its landscapes since natural soil with microorganisms), only 50 % of species the Neolithic (Blondel and Aronson 1995). This history were still inhibited, one was non-sensitive, and five were pressure has selected a typical biodiversity from open habitats with numerous vital conservation species (Fonderflick Catherine Fernandez and Mathieu Santonja has equal contribution to et al. 2010), which is partly why the incredibly species-rich the work. Mediterranean biomes (Cowling et al. 1996) now count as “biodiversity hotspots” (Myers et al. 2000). Although orig- C. Fernandez (*) : M. Santonja : Y. Monnier : M. Chomel : V. Baldy : A. Bousquet-Mélou inal Mediterranean forest continues to be lost in the south Aix-Marseille Université Institut Méditerranéen de Biodiversité et part of the basin (largely through conversion to agriculture), d’Ecologie (IMBE UMR CNRS IRD 7263), further north, rural flight has significantly lessened the Campus St Charles, Case 4, pressure on ecosystems. This rural exodus has led to the 13331 Marseille Cedex 03, France e-mail: [email protected] afforestation of abandoned agricultural land (Barbero et al. 1990) and a large-scale progression of forest land cover R. Gros (Debussche et al. 2001). Pines generally recolonize aban- Aix-Marseille Université Institut Méditerranéen de Biodiversité et doned land (Richardson et al. 2007). This process, called d’Ecologie (IMBE UMR CNRS IRD 7263), Campus Saint-Jérôme, Case 441, secondary succession, directly threatens a mosaic of habitats 13397 Marseille Cedex 20, France of high patrimonial value by driving a homogenization of J Chem Ecol plant communities and a loss of biodiversity (Chauchard et (Inderjit and Weston 2000; Inderjit 2006). Specific objec- al. 2007; Sheffer 2012). tives were to: i) determine whether allelochemicals of P. Community organization and plant succession are under halepensis affect the germination and seedling growth of the control of biotic processes, particularly plant-plant inter- several species native to Mediterranean open habitats by actions such as resource competition, facilitation, and alle- using 12 target species naturally present in open habitats lopathy (Callaway and Walker 1997). Allelopathy is a typical of land abandonment; ii) evaluate the role of micro- process by which a plant releases biochemicals that influ- organisms in these effects by using natural and autoclaved ence the growth and establishment of other plants (Inderjit soil; and iii) verify that results obtained under lab conditions 2005). Allelopathic compounds are released into the envi- are consistent with community patterns observed in the field ronment through foliar leachates, root exudation, leaf litter, by performing floristic inventories along secondary other-residue decomposition, or volatilization (Rice 1984) succession. and may influence various physiological processes (e.g., photosynthesis, nutrient uptake, cell division, or elonga- tion). Through these detrimental but also beneficial effects Methods and Materials on receptor plants, allelochemicals play important roles in regulating species diversity (Chou 1999). However, if Study Site The study site is located in the Luberon Natural allelochemicals can directly affect plant neighbors, Regional Park, SE France. This site is in secondary succes- allelopathic expression mayinturnbemodifiedbysoil sion, including 5 stages of P. halepensis colonization: fallow microbial communities by influencing the persistence, land without pines, fallow land with a few young pines availability and biological impacts of the allelochemicals (about 5 yr old), young pine forest (about 10 yr old), mid- (for a review, see Cipollini et al. 2012). age forest (about 30 yr old), and old pine forest (>60 yr old). Studying how organisms interact with each other is rele- vant to understanding ecological community organization. Material Collection The study site used for material collec- Research is needed to test the sensitivity of wild plant tion (plants and soil) is fallow land (abandoned farmland, species to allelochemicals, factoring in the potential role of formerly vineyard and almond orchard; 43°45′34.26″N; 5° soil microbial community on the chemical plant-plant inter- 17′57.84″E) that represents early colonization by young P. actions. This is particularly important in Mediterranean halepensis (about 5 yr). plant communities that feature species rich in secondary Of the 50 species naturally present on the site, 12 were metabolites, and for which the ecological relevance of alle- selected: Arabis hirsuta (L.) Scop., Avena barbata Pott ex lopathy remains unclear. Link, Briza maxima L., Dactylis glomerata L., Daucus The long-term aim of our research is to analyze how carota L., Helichrysum stoechas (L.) Moench, Linum stric- allelochemicals of Pinus halepensis Miller influence plant tum L., Reichardia picroides (L.) Roth, Salvia verbenaca L., biodiversity in Mediterranean open mosaic habitats during Sedum sediforme (Jacq.) Pau, Tanacetum corymbosum (L.) the colonization stage of secondary succession. Pinus Sch. Bip., and Trifolium stellatum L. Our selection aimed to halepensis is a circum-Mediterranean species covering 2.5 test different plant families with different life history traits million hectares in the Mediterranean basin (Quezel 2000) (Table 1). Seeds of all species were collected from wild that has expanded massively over the last century, facilitated populations on the study site outside the zone of influence by forest fires and abandonment of farmland (Richardson et of Aleppo pine, then stored in a cold chamber at 5 °C until al. 2007). This pioneer and expansionist species has come to the start of the experiment. Pinus halepensis needles used to dominate the areas of agricultural decline (Gondard et al. prepare the aqueous extracts were collected from 10 young 2003), thus contributing to the homogenization of plant individual Aleppo pines (about 5 yr old) present in the study communities in the Mediterranean area. Moreover, P. site. Soil samples used as bioassay substrate also were halepensis is rich in secondary metabolites (Macchioni et collected outside the zone of influence of Aleppo pine al. 2003; Fernandez et al. 2009) that are thought to play (i.e., at least 10 m from trees), sieved through a 2 mm- a role in plant-plant interactions through allelopathic mesh screen, then stored at room temperature until the start processes (Fernandez et al. 2006, 2008, 2009). of the experiment. This soil was characterized as agricultural In order to test whether P. halepensis allelochemicals Rendzic Leptosol soil according to FAO criteria (FAO 1998) might drive plant biodiversity in Mediterranean open mosa- and as Rendoll according to “Soil Taxonomy” criteria (Soil ic habitats during secondary succession, we performed a lab Survey Staff 1999). experiment consisting of allelopathy bioassays that mimicked natural conditions. These kinds of experiments are Allelopathy Bioassays Water-soluble compounds are prob- an important prerequisite for understanding the scale of ably the most involved in allelopathy (Vyvyan 2002). allelopathic mechanisms in plant community dynamics Previous studies on P. halepensis showed that needle hmEcol Chem J

Table 1 Characteristics of the target species used for the bioassays (Rameau et al. 1989; Gachet et al. 2005)

Target species Families Native range Life Raunkiær plant Grime Flowering Vector of pollen Vector of seed Species characteristics cycles life-forms strategy period dispersion dispersion

Arabis hirsuta (L.) Brassicaceae Eurasian-Mediterranean Biannual Hemicryptophyte SR June–August Entomogamous Anemochory / Heliophilous Mesoxerophilous- Scop. Autochory Calcicolous Avena barbata Pott Poaceae Mediterranean Annual Therophyte SR May–July Anemogamous Zoochory Heliophilous- Xerophilous ex Link Thermophilous Briza maxima L. Poaceae Mediterranean Annual Therophyte S May–June Anemogamous Autochory Heliophilous- Mesoxerophilous Thermophilous Dactylis glomerata L. Poaceae Mediterranean Perennial Hemicryptophyte C / CSR April–August Anemogamous Autochory / Heliophilous Xerophilous Zoochory Daucus carota L. Apiaceae European Biannual Hemicryptophyte SR May–October Entomogamous Autochory / Heliophilous- Mesoxerophilous Zoochory Thermophilous Helichrysum stoechas (L.) Moench Asteraceae West-Mediterranean Perennial Chamaephyte S April–July Autogamous / Anemochory / Heliophilous Xerophilous Entomogamous Barochory Linum strictum L. Linaceae Mediterranean Annual Therophyte SR May–July Entomogamous Autochory Heliophilous- Xerophilous- Thermophilous Calcicolous Reichardia picroides Asteraceae Mediterranean Perennial Therophyte CSR April–July Entomogamous Anemochory / Heliophilous Xerophilous (L.) Roth Barochory Salvia verbenaca L. Lamiaceae Atlantic-Mediterranean Perennial Hemicryptophyte R March–June Autogamous / Zoochory Heliophilous- Calcicolous Entomogamous Thermophilous Sedum sediforme Crassulaceae Mediterranean Perennial Chamaephyte S June–August Entomogamous Anemochory / Heliophilous- Xerophilous- (Jacq.) Pau Autochory Thermophilous Calcicolous Tanacetum corymbosum Asteraceae Continental Perennial Hemicryptophyte SR June–August Entomogamous Anemochory Thermophilous Mesoxerophilous- (L.) Sch.Bip. Mediterranean Calcicolous Trifolium stellatum L. Fabaceae Mediterranean Annual Therophyte S May–July Entomogamous Anemochory Heliophilous Xerophilous J Chem Ecol leachates present higher phenolics diversity and higher after germination (accuracy: 1 mm). We calculated relative quantities of terpenoids than roots (Fernandez et al. 2009), growth using the following formula: but also that needle leachates are the main pathways of allelopathic effects for young pine (Fernandez et al. 2006). GrowthðÞ¼ð % growth parameter in a treatment=growth For these reasons, foliar leachates were used for the bio- parameter in controlÞ100 assays. The stock solution of needle extracts was prepared by soaking 200 g (fresh weight) in 1,000 ml of deionized Chemical Analysis The chemical composition of aqueous water (10 % dry weight, as plant material is 50 % water) for extracts from needles was analyzed using the same proce- 24 h in darkness. Diluted solution (2.5 %) was prepared dure as Fernandez et al. (2009). Polar compounds (fatty from stock solution. Extracts were stored at 4 °C until acids, fatty diacids, simple phenols, acetophenones, pheno- starting the experiment. In order to prevent degradation of lic acids, and cinnamic acids), and less polar compounds compounds, new extracts were prepared once a week with (monoterpenes and sesquiterpenes) were quantified using fresh material. Note that for these experiments, the 10 % GC-MS instruments. A specific SIM (Selected Ion aqueous extracts of needles were diluted to an osmotic Monitoring) method was developed to analyze polar com- pressure significantly lower than 0.5 atm, as suggested by pounds by determination of major fragments and retention Anderson and Loucks (1966). time of injected authentic reference standards (Sigma- In order to assess the role of natural soil microbial Aldrich®). communities in shaping allelopathic effects (Inderjit 2006; A SCAN method was developed for less polar com- Kaur et al. 2009), we compared natural and autoclaved soil pounds analyzed. The identification was done by com- bioassays. Sterilization consisted in autoclaving soil for two parison of MS spectra to those of authentic reference cycles of 1 h (24 h apart) at 121 °C in order to eliminate a standards (Sigma-Aldrich®). Database searches in the fraction of the microbial community (Alef and Nannipieri NIST 2008 mass spectral library were conducted for 1995; Trevors 1996). tentatively identified components. Retention indexes of Bioassays were conducted in Petri dishes with 50.0 g compounds were determined relative to Wisconsin Diesel (± 0.1 g) of soil (autoclaved or natural) corresponding to a Range Hydrocarbons injection (Interchim, Montlucon, thickness of 0.5 to 0.6 mm to allow smooth radicle France) and tentatively confirmed by comparison with those development. Each Petri dish was sown with 25 seeds of expected in the literature (Adams 2007). each target species that were watered every 2 d with 5 ml of deionized water (control) or with needle aqueous extract Soil Microbial Activity Basal respiration (BR) was mea- (2.5 % or 10.0 %). While the bioassays performed with sured to assess the ecophysiological state of soil microbial autoclaved soil and 10 % needles aqueous extract allowed communities in the allelopathy bioassay soils. Ten grams us to observe the allelopathic potential of P. halepensis, the (dry weight equivalent) of fresh soil were placed in 117 ml bioassays performed with natural soil and 2.5 % aqueous glass jars then pre-incubated for 4 d at 22 °C to allow extract more realistically mimic the natural conditions. Four microbial respiration to restart. The glass jars were then replicates were performed for each treatment (species x soil closed with hermetic rubber septa and incubated for 4 h at x extract dose). Bioassays were conducted under natural 22 °C. After incubation, 1 ml of air sampled in the head- photoperiod and controlled temperature (20.5 °C±1 °C) space with a syringe was injected into a gas chromatograph until 5 d after germination (see below). (ChrompackCHROM 3-CP 9001) to analyze CO2 produc- We studied two sets of plant traits reflecting germination tion. The gas chromatograph was equipped with a thermal and seedling growth. Relative germination was calculated conductivity detector and a packed column (Porapack). − by using the following formula: Carrier gas (helium) flow was regulated at 60 ml.h 1. Ambient CO2 concentrations were subtracted from sampled GerminationðÞ¼ð % number of germinated seeds in a treatment= CO2 concentrations, and the values were adjusted to 22 °C according to Ideal Gas Laws using Q10=2. Microbial bio- Þ number of germinated seeds in control 100 mass (MB) was estimated using substrate-induced respiration (SIR) rates (Anderson and Domsch 1978). Ten grams and speed of germination was calculated by using the (dry weight equivalent) of subsamples were placed in Kotowski velocity coefficient (Mazliak 1982), Cv=100 117 ml glass jars and amended with powdered glucose − (ΣNi / ΣNiTi), where Ni is number of seeds germinated at (1,000 μgCg1 soil). After incubation (1 h, 22 °C), 1 ml time i, and Ti is number of days since the start of the of air sampled in the headspace with a syringe was injected experiment. into the gas chromatograph to analyze CO2 production. SIR To estimate seedling growth, length of root and shoot was rates were converted into MB using equations given by measured for each individual, at one time point, i.e., 5 d Beare et al. (1990). J Chem Ecol

Vegetation Analysis Floristic inventories were completed of community structure in (i) fallow land without pines, (ii) during spring and summer of 2009, taking into account a fallow land with few young pines, (iii) young pine forest, P. halepensis recovery gradient. All 5 colonization stages (iv) mid-aged forest. Old forests were not considered, as it were sampled. Sample plot size was 25 m2, and we ran 5 was rare to find the same species in at least two sites. With replicates. The 12 species observed for each plot were each these new traits, we framed an underlying Euclidean assigned an abundance-dominance value on an ordinal scale distance matrix on traits of sensitivity and contribution to of classes derived from the Braun-Blanquet method (1932): community structure similarity within successional stages 0=absent, 1=cover<1 %, 2=cover<5 %, 3=cover [5– (species*traits). We then performed a distance-based multi- 25 %], 4=cover [25–50 %], 5=cover [50–75 %], 6=cover variate multiple regression analysis (distLM) on traits of [75–100 %]. For statistical analysis, cover-abundance clas- contributiontosimilarityof community structure within ses were converted to percentage midpoints (respectively 0, successional stages, with sensitivity traits as predictor 0.5, 2.5, 15, 37.5, 62.5, 87.5). variables. Multivariate analysis were performed using Primer software v6 (Primer-E Ltd, UK). Statistical Analyses Treatments were compared using a Chi-squared test for germination rates followed by a post hoc test using Bonferroni adjustment (Scherrer 1984). Results Differences in plant growth (root, shoot) according to ssex- tract dose and soil type were tested using multi-way and Effects on Seed Germination Except for three target species one-way ANOVA (on significant interactions) followed by (B. maxima, D. carota, and T. corymbosum), Pinus extracts a post hoc Tukey test. Normality and homocedasticity had an effect on germination rate (Fig. 1). Two species were were previously tested by Shapiro-Wilk and Bartlett tests, stimulated by extracts (A. barbata and S. sediforme) while 7 respectively. Statgraphics1 (Version 4.1) was used for these were inhibited at least on autoclaved soil (A. hirsuta, D. statistical analyses. glomerata, H. stoechas, L. strictum, R. picroides, S. verbe- Differences in soil microbial activity (BR, MB) accord- naca and T. stellatum). The inhibition observed in auto- ing to extract dose and soil type were tested using a Kruskal- claved soil, however, was not reproduced in natural soil Wallis test followed by a post hoc Student-Newman-Keuls for 4 species: A. hirsuta, D. glomerata, H. stoechas, and T. test. stellatum. Moreover, inhibition was stronger at high extract Sensitivity index (phenotypic response to allelochemi- dose (on autoclaved soil), and in one case even completely cals) was assessed using results obtained with natural soil blocked germination (R. picroides on autoclaved soil with a and the 2.5 % extract dose for each trait (e.g., shoot growth, 10.0 % extract dose). germination rate) and each species: (treatment mean value— Pinus extracts also had an effect on germination velocity control mean value) / maximal mean value (control or (Table 2), except for A. barbata, B. maxima, and D. glom- treatment). The indices obtained were used to run hierarchi- erata. We observed either a slowdown (A. hirsuta, D. car- cal clustering on an underlying dissimilarity matrix of sen- ota, L. strictum, R. picroides, S. verbenaca, S. sediforme, sitivity (species*plant trait) to group species as a function of and T. stellatum) or an increase (H. stoechas and T. corym- sensitivity. Analysis of Similarity (ANOSIM) Global R sta- bosum at least on natural soil) in germination velocity. As tistic was used to test for differences among groups of was the case for germination rate, the effects were modulat- species sensitivity resulting from cluster analysis. ed by soil sterilization but also by extract dose. Following this first step, from a second underlying Braun- Blanquet dissimilarity matrix of relative abundance data Effects on Seedling Growth Shoot growth generally is im- (species*quadrat), we ran cluster analysis to discriminate pacted by allelochemicals, but microorganisms present in the main groups representative of successional stage. In a natural soil may offset the negative effect generally ob- third step, we attempted to connect species sensitivity to served (Fig. 2). Eight species (A. hirsuta, D. carota, D. allelopathy with the results of the floristic inventories. For glomerata, L. strictum, R. picroides, S. verbenaca, T. cor- this, we calculated the contribution of each species to suc- ymbosum, and T. stellatum) showed inhibited shoot growth cessional variation in community structure using Similarity on both soil types, with a stronger effect for the higher-dose Percentage (SIMPER) routines. This procedure calculated extract. For three species (B. maxima, H. stoechas, and S. Bray-Curtis dissimilarity values among replicates between sediforme), the inhibition of shoot growth observed on and within successional stages in the entire dataset of spe- autoclaved soil was not reproduced in natural soil, and in cies relative abundance. The average within-successional some cases shoot growth was stimulated by the lower con- stages similarities then were broken down into separate centration of extract. contributions from each taxon. Four traits were calculated Root growth universally followed the same pattern (Fig. 3): for each species, i.e., the relative contribution to similarity 5 species (D. carota, L. strictum, R. picroides, S. verbenaca, J Chem Ecol

Arabis hirsuta Avena barbata Briza maxima 140 140 140 2 = 11.92** 2 = 3.44 2 = 7.25** 2 = 7.46** 2 = 0.63 2 = 3.38 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 a a a b b c a b c

Germination (%) 20 20 20 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Dactylis glomerata Daucus carota Helichrysum stoechas 140 140 140 2 = 19.77*** 2 = 0.62 2 = 0.75 2 = 3.97 2 = 10.38** 2 = 2.43 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 a ab b a b b 20 Germination (%) 20 20 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Linum strictum Reichardia picroides Salvia verbenaca 140 140 140 2 = 29.02*** 2 = 6.41* 2 = 24.39*** 2 = 13.00** 2 >1000*** 2 = 20.76*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 a b c a ab b c a b b Germination (%) 20 20 20 a b a b c a b c 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Sedum sediforme 2 = 19.15*** Tanacetum corymbosum 2 = 0.36 Trifolium stellatum 140 140 140 2 = 1.92 2 = 0.96 2 = 20.77*** 2 = 5.25 120 120 120 100 100 100 80 80 80 60 60 60 40 a b b 40 40 20 b Germination rate (%) 20 20 a b 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Control 2.5% Extract 10.0% Extract

Fig. 1 Relative Germination of the 12 target species tested (± standard germination rate are also given (chi-squared value and P: * significant error) as a function of Pinus halepensis needles extract dose (high vs at 0.05; ** significant at 0.01; ***: significant at 0.001; N=100 for low) and soil quality (sterilized vs. natural). Chi-squared tests on each treatment) and T. corymbosum) showed inhibited root growth on both to the results obtained with autoclaved soil and high concen- soil types, with a stronger effect for the higher-dose tration of extract, and (ii) the sensitivity in the conditions closer extract. For three species (A. hirsuta, D. glomerata, and T. to natural ones, corresponding to the results obtained with corymbosum), the inhibition of root growth observed on au- natural soil (including microorganisms) and low concentration toclaved soil was not reproduced on natural soil. In terms of of extract. shoot growth, H. stoechas and S. sediforme were inhibited Focusing on potential sensitivity, 10 species inhibited with autoclaved soil and stimulated with natural soil only at germination and/or growth in the presence of pine extracts, the lower-dose extract. Finally, two species (A. barbata and B. but on a sensitivity gradient: 7 species were strongly germi- maxima)showedstimulatedrootgrowthonbothsoiltypes. nation and growth-inhibited (A. hirsuta, D. glomerata, H. These results allow us to group species according to their stoechas, L. strictum, S. verbenaca, and T. stellatum) and sensitivity to extracts: (i) the potential sensitivity corresponding one was totally inhibited (R. picroides germination was J Chem Ecol

Table 2 Velocity coefficient of the 12 target species as a function of Pinus halepensis needles extract dose (Control, 2.5 %, 10.0 %) and soil quality (sterilized, natural). Different letters indicate significant differences between treatments at P<0.05

Sterilized soil Natural soil

Control 2.5 % extract 10.0 % extract Control 2.5 % extract 10.0 % extract

Arabis hirsuta 16.23b 11.81ab 11.17a 16.40b 13.76ab 12.58a Avena barbata 29.16a 29.53a 31.72a 24.64a 25.57a 28.08a Briza maxima 28.48a 30.17a 31.84a 28.42a 29.81a 28.72a Dactylis glomerata 9.30a 9.44a 10.65a 9.00a 9.76a 10.072a Daucus carota 9.59b 8.02ab 4.34a 11.64b 7.51ab 5.23a Helichrysum stoechas 8.88a 14.50ab 16.15b 9.81a 10.79ab 12.81b Linum strictum 10.50b 7.565ab 7.10a 18.66b 11.27ab 9.38a Reichardia picroides 19.59b 10.74a 21.42b 12.26ab 8.98a Salvia verbenaca 34.17b 28.22ab 15.56a 38.08b 31.33ab 29.83a Sedum sediforme 7.63b 6.28a 6.67ab 12.11b 7.39ab 6.35a Tanacetum corymbosum 7.24a 6.23a 5.43a 6.08a 9.29b 7.86ab Trifolium stellatum 12.10b 8.4a 8.10a 10.45b 7.86ab 7.13a

suppressed by the high-dose extract), while 3 species were sesquiterpenes, 1 diterpene, and 1 oxygenated diterpene, 7 only seedling growth-inhibited (D. carota, S. sediforme, and fatty acids, 8 phenolic compounds, and 2 others (Table 3). T. corymbosum). Finally, two species showed stimulated Mixtures predominantly consisted of phenolics and fatty responses: B. maxima for root growth (but accompanied acids. by an inhibition of shoot growth), and A. barbata, which was the only species that also showed stimulated germina- Soil Microbial Activity Basal respiration and microbial bio- tion rate. mass were strongly decreased by the sterilization process Under the more natural conditions (low extract concen- (Kruskal-Wallis tests, P<0.05), whatever the extract dose. tration and in presence of a soil microbial community), of BR and MB decreased significantly in bioassays using the the 7 species strongly inhibited during bioassays with higher-dose extract on natural soil, whereas MB increased autoclaved soil and high extract concentration, only 3 with extract dose on autoclaved soil (Kruskal-Wallis tests, P (L. strictum, R. picroides, and S. verbenaca) were inhibited <0.05; Fig. 4). for all traits measured. Daucus carota, D. glomerata, and T. stellatum were inhibited but only for growth, whereas Vegetation Analysis The abundance of the 12 species A. barbata and S. sediforme were stimulated for both chosen for the bioassays decreased along the secondary germination and seedling growth, B. maxima and succession with land closure due to pine colonization H. stoechas were stimulated for seedling growth only, and (Table 4). Moreover, finer-grained analysis of early col- A. hirsuta and T. corumbosum became non-sensitive. onization stages (fallow, fallow with few young pines, The cluster analysis run with a multi-trait approach that young pinewood) revealed that the abundance of 4 uses all sensitivity indices revealed 3 different groups (data species decreased once pines were present (D. carota, not shown) representing 3 response types (ANOSIM, P< L.strictum,R.picroidesand S. verbenaca;Table4). 0.05; Fig. 5). Group 1 pools species that are highly sensitive Three other species showed a decrease in abundance (R. picroides, S. verbenaca, T. stellatum), group 2 pools from fallow to young pinewoods (A.hirsuta,T.corym- species with moderate sensitivity (L. strictum, D. carota S. bosum and T. stellatum). Other species abundances did sediforme, A. hirsuta), group 3 pools non-sensitive species not vary significantly among the first three stages of (A. barbata, B. maxima, H. stoechas, D. glomerata, T. succession (Table 4). Cluster analysis on abundance data corymbosum). Germination velocity emerged as an impor- discriminated two groups of species: one characteristic tant sensitivity trait. of open habitats (fallow, fallow with few young pines) and one characteristic of woodlands (all types of forest Chemical Analysis Aqueous extracts of young Pinus nee- pine; ANOSIM, P<0.05). The distLM analysis allowed dles contain numerous terpenoids and phenolic com- us to link the species more contributive to open habitats pounds (identified in Table 3): 8 monoterpenes, 12 or woodlands with their specific sensitivity to allelopa- oxygenated monoterpenes, 4 sesquiterpenes, 3 oxygenated thy. The first axis explains 68.3 % of the variability J Chem Ecol

Arabis hirsuta Avena barbata Briza maxima 140 140 140 F = 15.56*** F = 8.18*** F = 2.25 F = 1.44 F = 10.85*** F = 3.65* 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40

Shoot growth (%) 20 20 20 b a a b ab a a a a a a a b a ab a b b 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Dactylis glomerata Daucus carota Helichrysum stoechas 140 140 140 F = 13.28*** F = 21.56*** F = 7.29** F = 14.78*** F = 15.01*** F = 8.02*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40

Shoot growth (%) 20 20 b b a c b a 20 b b a b a a b b a ab b a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Linum strictum Reichardia picroides Salvia verbenaca 140 140 140 F = 15.01*** F = 8.02*** F = 97.50*** F = 16.78*** F = 254.30*** F = 204.30*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40

Shoot growth (%) 20 b a a b b a 20 b a b a a 20 c b a c b a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Sedum sediforme Tanacetum corymbosum Trifolium stellatum 140 140 140 F = 77.79 *** F = 40.48*** F = 19.81*** F = 15.86*** F = 34.27*** F = 25.45*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 20 Shoot length (%) 20 20 c b a b c a c b a b b a c b a c b a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Control 2.5% Extract 10.0% Extract

Fig. 2 Relative shoot growth of the 12 target species tested (± standard significant at 0.001) along with Tukey’s tests on differences between error) as a function of Pinus halepensis needles extract dose (high vs. doses for each soil type. Values sharing the same letter are not different low) and soil quality (sterilized vs. natural). ANOVA test results are at α=5 % given (F value and P: * significant at 0.05; ** significant at 0.01; ***: along a sensitivity gradient with more sensitive species Discussion in the positive part and fewer sensitive species in the negative part. Moreover, the most sensitive species are The results of our study confirm that extracts of P. halepen- all contributive to open habitats, whereas the species sis affect the germination and the growth of most target mostly contributive to woodland communities all species. Extracts of P. halepensis needles present high alle- belonged to the group of less allelopathy-sensitive spe- lopathic potential, as both reference species (Avena sativa, cies (Fig. 5) except for T. corymbosum.Finally,germi- Lactuca sativa, and Lemna minor) and wild species (Cistus nation velocity was the only trait that varied from other salvifolius, Cynodon dactylon, Festuca arundinacea, Linum traits in terms of response to allelopathy, thus well- strictum, and several species from this study) are sensitive to discriminating the more sensitive species. Pinus allelochemicals (Nektarios et al. 2005; Fernandez et J Chem Ecol

Arabis hirsuta Avena barbata Briza maxima 140 140 140 F = 6.57*** F = 0.97 F = 4.55* F = 21.51*** F = 21.12*** F = 5.30** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40

Root growth (%) 20 b a a a a a 20 a ab b a b c 20 a b c a ab b 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Dactylis glomerata Daucus carota Helichrysum stoechas 140 140 140 F = 84.93*** F = 1.48 F = 11.09 *** F = 15.37*** F = 123.51*** F = 60.39*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 Root growth (%) 20 c b a a a a 20 b b a b b a 20 c b a b c a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Linum strictum Reichardia picroides Salvia verbenaca 140 140 140 F = 170.00*** F = 77.97*** F = 184.40*** F = 29.76*** F = 229.65*** F = 52.67*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 Root growth (%) 20 c b a c b a 20 b b a 20 c b a c b a a a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Sedum sediforme Tanacetum corymbosum Trifolium stellatum F = 2.04 140 140 140 F = 120.45*** F = 66.26*** F = 16.78*** F = 96.27*** F = 103.51*** 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 Root length (%) 20 c b a b c a 20 c b a a a a 20 c b b a a a 0 0 0 Autoclaved soil Natural soil Autoclaved soil Natural soil Autoclaved soil Natural soil

Control 2.5% Extract 10.0% Extract

Fig. 3 Relative root growth of the 12 target species tested (± standard significant at 0.001) along with Tukey’s tests on differences between error) as a function of Pinus halepensis needles extract dose (high vs. doses for each soil type. Values sharing the same letter are not different low) and soil quality (sterilized vs. natural). ANOVA test results are at α=5 % given (F value and P: * significant at 0.05; ** significant at 0.01; ***: al. 2006; Valera-Burgos et al. 2012). Pinus halepensis nee- but terpenoid concentration were higher. Genetic variation dles contain numerous phenolic and terpenoid compounds may allow local adaptation, driving increased allocation to (Robles et al. 2003; Ormeño et al. 2008). Aqueous extracts allelopathy in competitive environments and decreased of young (aged ~10 years) P. halepensis trees living in allocation to allelopathy when the costs exceed benefits abandoned farmland have been analyzed previously (Meiners et al. 2012). Our extracts, as in Fernandez et al. (Fernandez et al. 2009), revealing a complex mixture (2009), showed a predominance of phenolics, including belonging to different functional phytochemical groups gallic, 4-hydrobenzoic, p-coumaric and caffeic acids, which (monoterpenes, oxygenated monoterpenes, sesquiterpenes, are known phytotoxins that act as plant growth regulators oxygenated sesquiterpenes, fatty acids, phenolic com- (Williams and Hoagland 1982; Kuiters 1989). For example, pounds). Our extracts contained the same chemical groups gallic, 4-hydrobenzoic, and p-coumaric acid can reduce J Chem Ecol

Table 3 Chemical compounds found in extracts of Pinus halepensis Table 3 (continued) needle from fallow land (mean concentration ± standard deviation) in ng/ml. RI: Retention Index (retention indexes of fatty acids and phe- RI Common name Mean (± SE) nolics are those of methyl derivatives due to the extraction/derivatiza- tion method employed) 1384 4-Hydroxybenzoic acid* 92.63 (16.73) 1482 Vanillin* 56.71 (11.05) RI Common name Mean (± SE) 1602 Protocatechuic acid* 627.27 (114.78) Monoterpenes 1731 Gallic acid* 635.16 (70.44) 970 α-Pinene* 9.68 (0.14) 1576 Acetovanillone* 148.29 (19.97) 994 Sabinene° 2.95 (0.06) 1681 p-coumaric acid* 1841.18 (488.21) 996 β-Pinene* 1.69 (0.03) 1894 Caffeic acid* 117.92 (75.94) 1007 β-Myrcene 3.53 (0.08) Others 1008 3-Carene* 3.99 (0.14) 987 Benzaldehyde* 73.33 (11.42) 1026 α-Terpinene* 0.07 (0.01) 1386 Cinnamic acid* 20.47 (2.08) 1033 p-Cymene* 0.10 (0.02) Total monoterpenes 427.78 (10.02) 1084 δ-Terpinene* 1.63 (0.14) Total sesquiterpenes 82.63 (7.34) Oxygenated monoterpenes Total diterpenes 20.17 (3.95) 1071 trans-p-Menth-2-en-1-ol° 2.24 (0.19) Total fatty acids 2132.58 (358.96) 1076 cis-Linalool oxide° 8.07 (0.23) Total phenolics 3640.25 (657.71) 1092 trans-Linalool oxide° 4.19 (0.20) *compared to authentic standards; °tentatively identified 1098 trans-Sabinene hydrate° 4.39 (0.86) 1100 Linalool* 28.22 (1.10) 1114 cis-Sabinene hydrate° 14.45 (2.49) plant water use and/or inhibit seed germination/root elonga- 1141 Camphor* 3.83 (0.20) tion (Blum and Gerig 2006; Reigosa and Pazos-Malvido 1175 4-Terpineol° 222.91 (5.71) 2007). Caffeic acid may be involved in physiological and α 1189 -Terpineol* 111.48 (2.44) biochemical changes in plants, including effects that lead to 1191 p-Cymen-8-ol 0.94 (0.18) strongly inhibited root growth (Gallet 1994; Barkosky et al. 1258 Piperitone ° 2.93 (0.28) 2000). Moreover, as a rule, allelopathic interactions are not 1285 Bornyl acetate* 0.49 (0.10) due to a single compound but rather to a pool of several Sesquiterpenes allelochemicals (as in our extracts) acting synergistically to 1415 β-Caryophyllene* 28.25 (3.94) inhibit or stimulate growth (Reigosa et al. 1999). 1449 α-Caryophyllene° 6.65 (1.02) There are difficulties in relating laboratory bioassays to 1498 α-Muurolene° 1.33 (017) allelopathic interactions in the field. These are related to 1529 δ-Cadinene° 2.65 (0.42) slow release, low concentrations, or degradation of the Oxygenated sesquiterpenes toxic agent. Despite these limitations, laboratory bioassays 1548 Elemol* 16.16 (0.33) are useful for screening allelopathic potential. They are 1577 Caryophyllene oxide° 22.11 (1.66) particularly useful when natural target species and natural 1610 Humulene epoxide II ° 5.48 (0.35) soil are used in bioaasays, thus making them more eco- Diterpenes logically meaningful, as in this study. The effects ob- 1929 Cembrene° 7.41 (1.45) served in the bioassays in this study varied depending Oxygenated diterpenes on target species and on presence of the soil microbial 2055 Thunbergol° 12.76 (2.50) community. Moreover, degree of inhibition was related to Fatty acids extract dose. With the high-dose extract and autoclaved 1041 Succinic acid* 1273.56 (256.13) soil, more than 80 % of target species were germination 1324 Capric acid* 25.71 (5.57) and/or growth-inhibited and only two species—the 1472 Citric acid* 351.25 (78.93) Poaceae A. barbata and B. maxima—were non-sensitive. 1524 Lauric acid* 57.21 (10.23) El-Khawas and Shehata (2005) showed that Eudicots are 1557 Azelaic acid* 48.67 (7.36) more sensitive than Monocots to the allelochemicals of 1935 Palmitic acid* 159.84 (36.07) Acacia nilotica and Eucalyptus rostrata. Alrababah et al. 2159 Stearic acid* 216.34 (37.46) (2009) reported a similar sensitivity gradient between Phenolics Fabaceae and Poaceae to allelochemicals of P. halepensis. 1195 Salicylic acid* 121.09 (26.54) Delayed germination may have important biological implications, particularly under a Mediterranean climate J Chem Ecol

0.030 4.000 F = 2.68 F = 6.46* F = 10.97** F = 14.59** 0.025 3.000 0.020

0.015 2.000

0.010 1.000 0.005 a a a b b a a a b b b a Microbial biomass (µg Cmic/g) 0.000 0.000

Basal respiration(µg C-Co2/h/g)/g) Autoclaved soil Natural soil Autoclaved soil Natural soil

Control 2.5% Extract 10.0% Extract

Fig. 4 Mean basal respiration and biomass of the microbial commu- significant at 0.05; ** significant at 0.01; ***: significant at 0.001) nity during the bioassays (± standard error) as a function of Pinus along with Tukey’s tests on differences between doses for each soil halepensis needles extract dose (high vs. low) and soil quality (steril- type. Values sharing the same letter are not different at α=5 % ized vs. natural). ANOVA test results are given (F value and P:* where early-emerging species could be more competitive for even become positive (mainly for low-dose extracts). It seems access to resources (Herranz et al. 2006). Effects on shoot clear that the microbial community present in natural soil and root growth are equally capable of shaping not just affects the toxicity of allelochemicals, as has been shown also competitiveness for access to resources but also seedling in other bioassays (Schmidt 1990; Inderjit 2005;Kauretal. viability. The successful establishment of a species in the 2009). Indeed, microbial activity can cause both quantitative Mediterranean region is dependent largely on a well- and qualitative variation in chemicals present in the soil envi- developed root system for efficient capture of resources ronment through degradation and transformation (Inderjit and (Lloret et al. 1999; Green et al. 2005), particularly when Weiner 2001), and influence the persistence, availability and water uptake is a limiting factor. Inhibition of root system biological activity of allelochemicals (Blum and Shafer 1988; development can strongly diminish the performance of more Inderjit and Weston 2000; Meiners et al. 2012). Microbial sensitive plants, making them less competitive and, more activity may prevent compounds from building up to importantly, less tolerant to drought (Herranz et al. 2006). phytotoxic levels in natural soils (Schmidt and Ley 1999). Nevertheless, autoclaved soil bioassays can overestimate the The natural soil used in this study contained more microbial ability of allelochemicals to influence the growth of neighbor- biomass (and therefore microorganisms) than the autoclaved ing plants. With natural soil and low extract concentrations, soil. Soil microbial biomass increased when needle extracts only 50 % of species were inhibited (and only 25 % for were added to the autoclaved soil. Heterotrophic microorgan- germination and growth simultaneously). This indicates that isms, especially fungi, typically use phenolic compounds as a the effects of allelochemicals are less inhibitory, disappear, or source of organic carbon (Blum and Shafer 1988). It seems

Table 4 Cover-abundance percentages, calculated using a modified Braun-Blanquet method (1932), for each species along secondary succession stages from fallow land without Pinus to old Pinus forest. Different letters indicate significant differences between successional stages at P<0.05

Fallow land Fallow land with Young Pinus Mid-aged Pinus Old Pinus forest young Pinus (±5 yrs) forest (±10 yrs) forest (±30 yrs) (>60 yrs)

Arabis hirsuta 5.63a 2.50ab 0.70b 0.20b 0.00b Avena barbata 2.00ab 5.00a 2.10ab 0.70b 0.20b Briza maxima 5.13abc 7.50ab 10.00a 0.70bc 0.10c Dactylis glomerata 8.75a 7.50a 4.60ab 0.20b 0.10b Daucus carota 8.75a 2.50b 0.80b 0.10b 0.00b Helichrysum stoechas 1.50a 1.3a0b 0.80abc 0.30bc 0.00c Linum strictum 2.00a 0.90b 0.30bc 0.10c 0.00c Reichardia picroides 11.88a 2.10b 0.30b 0.00b 0.00b Salvia verbenaca 20.63a 7.50b 0.00c 0.00c 0.00c Sedum sediforme 5.13a 4.60a 1.20b 0.10b 0.0b Tanacetum corymbosum 2.00a 1.70ab 0.80bc 0.20c 0.10c Trifolium stellatum 2.50a 2.10a 0.20b 0.10b 0.00b J Chem Ecol

Fig. 5 Graphical representation Resemblance: D1 Euclidean distance of the results of the distLM 0.5 (distance based analysis on a linear model) on a dbRDA1, 2 plot (distance based redondancy G. Velocity analysis). The different groups discriminate according to T. corymbosum sensitivity to allelochemicals D. glomerata S. verbenaca T. stellatum H. stoechas (bioassays results) and 6% of total variation) . B. maxima R. picroides colonization stages (floristic A. barbata inventory results): three groups 0 Root growth for sensitivity to A. hirsuta allelochemicals (high Total growth Shoot growth sensitivity, moderate sensitivity,

,6% of fitted, 19 Germination L. strictum non-sensitivity) and two groups . D. carota for the colonization stages (open habitats, woodlands), i.e., S. sediforme F Fallow; FP fallows with few young pines; YP young pine dbRDA2 (1 forest (10 years); MF mid-aged -0.5 -0.5 0 0.5 1.0 dbRDA1 (68,3% of fitted, 68,3% of total variation) likely that microorganisms present in the extracts may have in some species but stimulates others (Fernandez et al. colonized sterile soil and used the secondary metabolites for 2006; Alrababah et al. 2009). Floristic inventories per- their growth. These microorganisms almost certainly included formed in the different stages of Pinus colonization epiphytic bacteria and fungi present in the phyllosphere. correlated with results from the lab bioassays, i.e., were Indeed, Mediterranean perennial plants, show high bacterial consistent with community patterns observed in the colonization in their phyllosphere, even if species rich in terpe- field. noids and phenolics (known antimicrobials) are characterized In the last stage of succession, the abundances of all 12 by scarce presence of bacteria (Karamanoli et al. 2000, 2005). species chosen for bioassays decreased strongly. Light avail- However, with natural soil and the high-dose extract, microbial ability in old pine forest can constitute a limiting factor for biomass decreased, probably due to stronger toxicity of allelo- the development of all these species. Focusing on the first chemicals to indigenous soil microorganisms compared to stages of colonization, where light is not limiting, 4 species phyllospheric microorganisms. showed a strong decrease in abundance once pines are Our results suggest that the plant leaf-colonizing bacteria present (D. carota, L. strictum, R. picroides and S. verbe- could be better adapted to such allelochemicals but the naca). Bioassays confirmed that these 4 species were highly introduced bacteria seem not able to colonize their new soil sensitive to extracts. The abundance of two other species environment while competing with indigenous microbes. decreased from fallow to young pinewoods (A. hirsuta, and This decrease in microbial biomass could in turn explain T. stellatum), and these species were germination or growth- the lower plant growth at this concentration, as the P. hale- inhibited or non-sensitive to extracts. Our multi-trait analy- pensis allelochemicals may have accumulated to high toxic sis highlighted strong links between medium/high sensitiv- levels. This point may confirm that allelochemicals are ity and open habitat vegetation characteristics: all the likely to influence plant growth not only directly but also sensitive species became scarce in earlier stages of pine indirectly through altered microbial associations (Meiners et colonization despite the fact that these heliophilous species al. 2012). In parallel, microbial degradation of compounds had enough light for growth (Kato-Noguchi et al. 2009). may generate new equally-toxic compounds. Note that Therefore, it is likely that the limitation of these species many cinnamic acids (e.g., ferulic acid, p-coumaric acid) expansion is at least partly due to the allelochemicals emit- that undergo microbial degradation could produce benzoic ted by P. halepensis. The non-sensitive species persisted in acids (e.g., vanillic acid, p-hydroxybenzoic acid; Inderjit woodlands. As has been demonstrated by Callaway et al. and Weiner 2001). (2005), the production of allelochemicals generates a con- Even though the allelopathic effect of Aleppo pine de- sistent selective pressure on resident species, and so geno- creased in the bioassays that mimicked natural conditions, types that are less sensitive should be favored in the half of the species tested still were negatively influ- community and able to maintain their position along the enced by Pinus extracts. This result underlines the eco- succession. Finally, the disappearance of some species in logical importance of the allelopathic features of this the older stage may be more confidently attributed to com- pine that hampers or blocks establishment and growth petition for resources (light, nutrients and/or water). J Chem Ecol

The link between sensitivity and abundance during suc- BRAUN-BLANQUET, J. 1932. Plant sociology. The study of plant com- cessional stages correlates with the differential sensitivity of munities. McGraw-Hill Eds, New York. CALLAWAY, R. M. and WALKER, L. R. 1997. Competition and facilita- species to phytotoxic interactions. Plants of Mediterranean tion: a synthetic approach to interactions in plant communities. open habitats are highly sensitive to Pinus extracts. In an Ecology 78:1958–1965. evolutionary context, it is generally supposed that plant CALLAWAY, R. M., RIDENOUR, W. M., LABOSKI, T., WEIR, T., and communities unexposed to allelochemicals may remain VIVANCO, J. M. 2005. Natural selection for resistance to the allelopathic effects of invasive plants. J. Ecol. 93:576–583. more sensitive to their effects (Meiners et al. 2012). CHAUCHARD,S.,CARCAILLET,C.,andGUIBAL, F. 2007. Patterns Mediterranean open mosaic habitats have been shaped by of land-use abandonment control tree-recruitment and forest a long history of agricultural or pasture practices since back dynamics in Mediterranean mountains. Ecosystems 10:936– to the Neolithic, and it is likely that during this long period, 948. CHOU, C. H. 1999. Roles of allelopathy in plant biodiversity and the open-habitats species have had little exposure to Pinus sustainable agriculture. Crit. Rev. Plant Sci. 18:609–636. allelochemicals. This point should be studied further as the CIPOLLINI, D., RIGSBY, C. M., and BARTO, E. K. 2012. Microbes as evolutionary history of allelopathic sensitivity and the re- targets and mediators of allelopathy in plants. J. Chem. Ecol. – cipient soil and plant community takes on increasing signif- 38:714 727. COWLING, R. M., RUNDEL,P.W.,LAMONT, B. B., ARROYO, M. K., and icance (Inderjit et al. 2011). ARIANOUTSOU, M. 1996. Plant diversity in Mediterranean climate regions. Trends Ecol. Evol. 11:362–366. Acknowledgements This study was funded by the CNRS [French DEBUSSCHE, M., DEBUSSCHE, G., and LEPART, J. 2001. Changes in the national centre for scientific research] within the framework of the Zone vegetation of Quercus pubescens woodland after cessation of “ ” Atelier Arrière-pays Méditerranéen . We are grateful to the staff of the coppicing and grazing. J. Veg. Sci. 12:81–92. Luberon Natural Regional Park. We would also like to thank Stéphane EL-KHAWAS, S. A. and SHEHATA, M. M. 2005. The allelopathic poten- Greff (IMBE) for his contribution to the chemical analyses, Sylvie tialities of Acacia nilotica and Eucalyptus rostrata on monocot Dupouyet (IMBE) for her help with the bioassays, and A-T-T (Scientific (Zea mays L.) and dicot (Phaseolus vulgaris L.) plants. and technical translation) for proofreading the draft manuscript. Biotechnology 42:3–34. FAO 1998. World reference base for soil resources. International Society of Soil Science, Rome. References FERNANDEZ,C.,LELONG, B., VILA,B.,MÉVY,J.P.,ROBLES,C., GREFF,S.,DUPOUYET,S.,andBOUSQUET-MÉLOU,A.2006. Potential allelopathic effect of Pinus halepensis in the sec- ADAMS, R. P. 2007. Identification of essential oil components by gas ondary succession: an experimental approach. Chemoecology chromatography / mass spectrometry, 4th ed. Allured Publishing 16:97–105. Corporation, Carol Stream. FERNANDEZ,C.,VOIRIOT, S., MEVY, J. P., VILA,B.,ORMEÑO,E., ALEF, K. and NANNIPIERI, P. 1995. Methods in applied soil microbiol- DUPOUYET, S., and BOUSQUET-MELOU, A. 2008. Regeneration ogy and biochemistry. Academic, London. failure of Pinus halepensis Mill.: the role of autotoxicity and some ALRABABAH, M., TADROS, M. J., SAMARAH, N. H., and GHOSHEH,H. abiotic environmental parameters. Forest Ecol Manag. 255:2928– 2009. Allelopathic effects of Pinus halepensis and Quercus coc- 2936. cifera on the germination of Mediterranean crop seeds. New FERNANDEZ, C., MONNIER, Y., ORMEÑO, E., BALDY, V., GREFF, S., Forest 38:261–272. PASQUALINI, V., MÉVY, J. P., and BOUSQUET-MELOU, A. 2009. ANDERSON, J. P. E. and DOMSCH, K. H. 1978. A physiological method Variations in allelochemical composition of leachates of different for the quantitative measurement of microbial biomass in soil. organs and maturity stages of Pinus halepensis. J. Chem. Ecol. Soil Biol. Biochem. 10:215–221. 35:970–979. ANDERSON, R. C. and LOUCKS, O. L. 1966. Osmotic pressure influ- FONDERFLICK, J., LEPART, J., CAPLAT, P., DEBUSSCHE, M., and MARTY, ence in germination tests for antibiosis. Science 152:771–773. P. 2010. Managing agricultural change for biodiversity conserva- BARBERO, M., BONIN, G., LOISEL, R., and QUEZEL, P. 1990. Changes tion in a Mediterranean upland. Biol. Conserv. 43:737–746. and disturbances of forest ecosystems caused by human activities GACHET, S., VELA, E., and TATONI, T. 2005. BASECO: a floristic and in the western part of the Mediterranean basin. Vegetatio 87:151– ecological database of Mediterranean french flora. Biodivers. 173. Conserv. 14:1023–1034. BARKOSKY, R. R., EINHELLIG, F. A., and BUTLER, J. L. 2000. Caffeic GALLET, C. 1994. Allelopathic potential in bilberry-spruce forests: acid-induced changes in plant-water relationship and photosynthesis influence of phenolic compounds on spruce seedlings. J. Chem. in leafy spurge Euphorbia esula. J. Chem. Ecol. 26:2095–2109. Ecol. 20:1009–1024. BEARE, M. H., NEELY, C. L., COLEMAN, D. C., and HARGROVE,W.L. GONDARD, H.,ROMANE, F.,ARONSON, J., and SHATER, Z. 2003. Impact 1990. A substrate-induced respiration (SIR) method for measure- of soil surface disturbances on functional group diversity after ment of fungal and bacterial biomass on plant residues. Soil Biol. clear-cutting in Allepo pine (Pinus halepensis) forests in southern Biochem. 22:585–594. France. Forest Ecol. Manag. 180:165–174. BLONDEL,J.andARONSON, J. 1995. Biodiversity and ecosystem GREEN, J. J., BADDELEY, J. A., CORTINA, J., and WATSON, C. A. 2005. function in the Mediterranean basin: human and non-human Root development in Mediterranean shrub Pistacia lentiscus as determinants. In biodiversity and ecosystem function in affected by nursery treatments. J. Arid. Environ. 61:1–12. Mediterranean-type ecosystems. Ecol. Stud. 109:44–119. HERRANZ, J. M., FERRandIS, P., COPETE, M. A., DURO, E. M., and BLUM, U. and GERIG, T. M. 2006. Interrelationships between p-cou- ZALACAIN, A. 2006. Effect of allelopathic compounds produced maric acid, evapotranspiration, soil water content and leaf expan- by Cistus ladanifer on germination of 20 Mediterranean taxa. sion. J. Chem. Ecol. 32:1817–1834. Plant Ecol. 184:259–272. BLUM, U. and SHAFER, S. R. 1988. Microbial populations and phenolic INDERJIT 2005. Soil microorganisms: an important determinant of acids in soils. Soil Biol. Biochem. 20:793–800. allelopathic activity. Plant Soil 274:227–236. J Chem Ecol

INDERJIT 2006. Experimental complexities in evaluating the allelopath- QUEZEL, P. 2000. Taxonomy and biogeography of Mediterranean pines ic activities in laboratory bioassays: a case study. Soil Biol. (Pinus halepensis and P. brutia), PP.1–12, in G. NE’EMAN and L. Biochem. 38:256–262. TRABAUD (EDS.), Ecology, biogeography and management of INDERJIT and WEINER, J. 2001. Plant allelochemical interference or soil Pinus halepensis and P. brutia forest ecosystems in the Mediter- chemical ecology? Perspect. Plant. Ecol. Evol. Syst. 4:3–12. ranean basin. Backhuys publishers, Leiden. INDERJIT and WESTON, L. A. 2000. Are laboratory bioassays for RAMEAU, J. C., MANSION, D., and DUME, G. 1989. Flore forestière allelopathy suitable for prediction of field responses? J. Chem. française (guide écologique illustré), Tome 1 : Plaines et Collines. Ecol. 26:2111–2118. Institut pour le développement forestier. IDF Publications. INDERJIT,WARDLE, D. A., KARBAN, R., and CALLAWAY, R. M. 2011. REIGOSA, M. J. and PAZOS-MALVIDO, E. 2007. Phytotoxic effects of 21 The ecosystem and evolutionary contexts of allelopathy. Trends plant secondary metabolites on Arabidopsis thaliana germination Ecol. Evol. 26:655–662. and root growth. J. Chem. Ecol. 33:1456–1466. KARAMANOLI,K.,VOKOU,D.,MENKISSOGLU, U., and CONSTANTINIDOU, REIGOSA, M. J., SANCHEZ-MOREIRAS, A., and GONZALEZ, L. 1999. H. I. 2000. Bacterial colonization of phyllosphere of Ecophysiological approach in allelopathy. Crit. Rev. Plant Sci. Mediterranean aromatic plants. J. Chem. Ecol. 26:2035–2048. 18:577–608. KARAMANOLI, K., MENKISSOGLU-SPIROUDI, U., BOSABALIDIS, A. M., RICE, E. L. 1984. Allelopathy. Academic, USA. VOKOU, D., and CONSTANTINIDOU, H. I. A. 2005. Bacterial col- RICHARDSON, D. M., RUNDEL, P. W., JACKSON, S. T., TESKEY, R. O., onization of the phyllosphere of nineteen plant species and anti- ARONSON,J.,BYTNEROWICZ,A.,WINGFIELD,M.J.,and microbial activity of their leaf secondary metabolites against leaf PROCHES, S. 2007. Human impacts in pine forests: past, present associated bacteria. Chemoecology 15:59–67. and future. Annu. Rev. Ecol. Evol. Syst. 38:275–297. KATO-NOGUCHI, H., FUSHIMI, Y., and SHIGEMORI, H. 2009. An alle- ROBLES,C.,GREFF,S.,PASQUALINI, V., GARZINO,S.,BOUSQUET- lopathic substance in red pine needles (Pinus densiflora). J. Plant MELOU, A., FERNandEZ, C., KORBOULEWSKY, N., and BONIN, Physiol. 166:442–446. G. 2003. Phenols and flavonoids in Aleppo pine needles as bio- KAUR, H., KAUR, R., KAUR, S., BALDWIN, I. T., and INDERJIT 2009. indicators of air pollution. J. Environ. Qual. 32:2265–2271. Taking ecological function seriously: soil microbial communities can SCHERRER, B. 1984. Biostatistique. Gaëtan Morin publishers, obviate allelopathic effects of released metabolites. PLoS One 4:e4700. Chicoutimi. KUITERS, A. T. 1989. Effects of phenolic acids on germination and SCHMIDT, S. K. 1990. Ecological implication of destruction of juglone early growth of herbaceous woodland plants. J. Chem. Ecol. (5-hydroxy-1,4-napthoquinone) by soil bacteria. J. Chem. Ecol. 15:467–479. 16:3547–3549. LLORET, F., CASANOVAS, C., and PEÑUELAS, J. 1999. Seedling survival SCHMIDT,S.K.andLEY, R. E. 1999. Microbial competition and of Mediterranean shrubland species in relation to root:shoot ratio, bioavailability limit the expression of allelochemicals in natural seed size and water and nitrogen use. Funct. Ecol. 13:210–216. soils, pp. 339–351, in Inderjit, K. M. M. Dakshini, and C. L. Foy MACCHIONI, F., CIONI, P. L., FLAMINI, G., MORELLI, I., MACCIONI, S., (eds.), Principles and practices in plant ecology: allelochemical and ANSALDI, M. 2003. Chemical composition of essential oils from interactions. CRC Press, Boca Raton. needles, branches and cones of Pinus pinea, P. halepensis, P. pinaster SHEFFER, E. 2012. A review of the development of Mediterranean and P. nigra from central Italy. Flavour Frag J. 18:139–143. pine-oak ecosystems after land abandonment and afforestation: MAZLIAK, P. 1982. Physiologie végétale: croissance et développement. are they novel ecosystems? Ann. For. Sci. 69:429–443. Hermann Publ, Paris. SOIL SURVEY STAFF. 1999. Soil Taxonomy. 2nd edition. USDA Natural MEINERS, S. J., KONG, C. H., LADWIG, L. M., PISULA, N. L., and Resources Conservation Service Agriculture Handbook 436. U.S. LANG, K. A. 2012. Developing an ecological context for allelop- Government Printing Office, Washington, D.C. athy. Plant Ecol. 213:1221–1227. TREVORS, J. T. 1996. Sterilization and inhibition of microbial activity MYERS, N., MITTERMEIER, N. A., MITTERMEIER, C. G., DA FONSECA, in soil. J. Microbiol. Meth. 26:53–59. G. A. B., and KENT, J. 2000. Biodiversity hotspots for conserva- VALERA-BURGOS, J., DIAZ-BARRADAS, M. C., and ZUNZUNEGUI,M. tion priorities. Nature 403:853–858. 2012. Effects of Pinus pinea litter on seed germination and NEKTARIOS,P.A.,ECONOMOU,G.,andAVGOULAS, C. 2005. seedling performance of three Mediterranean shrub species. Allelopathic effects of Pinus halepensis needles on turfgrasses Plant Growth Regul. 66:285–292. and biosensor plants. Hortscience 40:246–250. VYVYAN, J. R. 2002. Allelochemicals as lead for new herbicides and ORMEÑO,E.,BALDY,V.,BALLINI,C.,andFERNANDEZ,C.2008. agrochemicals. Tetrahedron 58:1631–1649. Production and diversity of volatile terpenes from plants on cal- WILLIAMS, R. D. and HOAGLAND, R. E. 1982. The effect of naturally- careous and siliceous soils: effect of soil nutrients. J. Chem. Ecol. occurring phenolic compounds on seed germination. Weed Sci. 34:1219–1229. 30:206–212.