Gayana Bot. 66(1): 71-83, 2009

ARTICULO REGULAR

COMPETITIVE EFFECTS OF THE ALIEN INVASIVE CENTAUREA SOLSTITIALISL. ON TWO CHILEAN BACCHARIS SPECIES AT DIFFERENT LIFE-CYCLE STAGES

EFECTOS COMPETITIVOS DE LA ALOCTONA INVASORA CENTAUREA SOLSTITIALIS L. SOBRE DOS ESPECIES CHILENAS DE BACCHARIS EN DIFERENTES ESTADOS DEL CICLO DE VIDA

Susana Gómez-González^1,2, Lohengrin A. Cavieres^1,2, Patricio Torres¹ & Cristian Torres-Díaz¹

¹Departamento de Botánica, Facultad de Ciencias Naturales y Oceanógraficas, Universidad de Concepción, Casilla 160-C, Concepción, Chile

²Instituto de Ecología y Biodiversidad (IEB), Facultad de Ciencias, Universidad de Chile, Casilla 653, Santiago, Chile. sgomez@udec.cl

---

ABSTRACT

Several studies have revealed a variety of mechanisms of invasion of alien plant species. However, little is known on how those mechanisms and their associated effects on native species change across different life-cycle stages. Under controlled conditions, we assessed the interactions between the alien invasive species Centaurea solstitialis L. (Asteraceae) and two pioneer native species to the Chilean matorral; Baccharis linearis (Ruiz et Pav.) Pers. and B. paniculata DC. (Asteraceae). Competitive effects of the invader on natives were evaluated by combining different life-cycle stages: seed-seed, plant-seed, and plant-plant. Seed germination of C. solstitialis was explosive and much faster than that of the native species. The presence of C. solstitialis (individuals or seeds) did not affect negatively the seed germination of the two Baccharis species. However, the presence of C. solstitialis plants significantly decreased the total biomass oí Baccharis plants. Thus, the effect of C. solstitialis on Baccharis species depended on the life-cycle stage at which the interactions occurred. In the Chilean matorral, the early emergence of C. solstitialis could be an important invasion mechanism, enabling established plants to competitively displace late emerging seedlings of Baccharis species. The huge abundance of C. solstitialis in some disturbed matorrals suggests that seedling establishment of these two pioneer species could be limited.

Keywords: Allelopathy, matorral, Mediterranean, seed germination, yellow starthistle.

---

RESUMEN

Varios estudios han revelado una variedad de mecanismos de invasión en las plantas alóctonas. Sin embargo, aún se conoce poco sobre cómo tales mecanismos y sus efectos asociados cambian a través de diferentes estados del ciclo vida. En este estudio evaluamos, bajo condiciones controladas, el resultado de las interacciones competitivas entre la especie invasora Centaurea solstitialis L. (Asteraceae) y dos especies pioneras nativas del matorral chileno; Baccharis linearis (Ruiz et Pav.) Pers. y B. paniculata DC (Asteraceae). Estas interacciones fueron evaluadas combinando diferentes estados del ciclo de vida: semilla-semilla, planta-semilla y planta-planta. La germinación de C. solstitialis fue explosiva, siendo mucho más rápida que la de las especies nativas. La presencia de C. solstitialis (plantas o semillas) no disminuyó la germinación de las especies nativas de Baccharis. Sin embargo, la presencia de plantas establecidas de C. solstitialis disminuyó significativamente la biomasa de las plantas de Baccharis. Entonces, el efecto de C. solstitialis sobre las especies de Baccharis varió en función del estado del ciclo de vida en el cual las interacciones ocurrieron. En el matorral chileno, la emergencia temprana y explosiva de C. solstitialis podría ser un importante mecanismo de invasión, ya que aquellas plantas prontamente establecidas podrían desplazar competitivamente a las plántulas de Baccharis que emergen más tarde. La enorme abundancia de C. solstitialis en algunas zonas de matorral sugiere que el establecimiento de plántulas de estas dos especies pioneras podría estar siendo limitado.

Palabras clave: Alelopatía, matorral, mediterráneo, germinación, abrepuño amarillo.

---

INTRODUCTION

The negative impact of alien invasive plants on native species populations, communities, and ecosystems is widely recognized (Mack et al. 2000), and much research have been done on eluding the mechanisms that allow alien plant species to become dominant in the recipient community (Levine 2000, Levine et al. 2003, Coleman & Levine 2007). However, less investigation has been focused on how different mechanisms interact across different life-cycle stages of invasive and native species to explain the invasion success (e.g. Onetal. 2005, Widmer etal. 2007, Ens & Frens 2008, Muoghalu 2008).

Regarding the effects at the plant neighbourhood scale, many studies have shown that invasive plant species can reduce around 50% of the native plant species' performance in terms of biomass or relative growth rate (for review see Vila & Weiner 2004). Those studies are generally focused on plant-plant interactions, where the negative effects of the invaders on natives are mainly caused by competition for soil resources (Gordon et al 1989, Welker et al. 1991) or by allelochemicals released by the invasive species (Callaway & Aschehoug 2000, Vivanco et al. 2004). In addition, alien invasive plants can also affect native species by inhibiting the establishment of new individuals (i.e., seedlings) by different mechanisms (Yurkonis et al. 2005). For example, some invasive species release allelochemicals that inhibit the seed germination of some native species (i.e., plant-seed interactions, Baise? a/. 2003, Jefferson & Pennacchio 2003, Prati & Bossdorf 2004).

Although seed-seed interactions could play an important role on explaining the alien success, little attention has been paid to interactions between alien and native species at the seed stage. Seed germination and seedling emergence can be affected by the presence of heterospecific seeds in the seed bank (Bergelson & Perry 1989, Lortie & Turkington 2002), where chemicals leached by seeds may be involved (Bergelson & Perry 1989, Murray 1998, Laterra & Bazzalo 1999, Casini & Olivero 2001). Invasion success can also be related to a high germination capability of the invader. Early emergence produces a competitive advantage since early germinated individuals can monopolize resources and attain sufficient biomass for a successful establishment (Miller 1987, Wilson 1988, Verdu & Traveset 2005). Thus, since the importance of the competitive effects of alien plants on native species could differ across their life cycle, a comprehensive understanding of the invasion mechanisms should explore the results of interactions at different life-cycle stages.

Centaurea solstitialis L. (Asteraceae) is an annual or biennial species native to Eurasia that has invaded successfully different regions of the world (Hierro et al. 2006). In Chile, C. solstitialis is distributed across the Mediterranean region of the country (central Chile), and it has been catalogued as a serious weed generating negative impacts on agricultural crops (Matthei 1995). Centaurea solstitialis is generally found on roadsides and old fields, but it has also been observed to grow densely on native shrublands (hereafter the matorral). The Chilean matorral has undergone intense human disturbances since the Spaniard colonization (grazing, burning, clearing, etc.). Disturbed matorrals are savannah-like vegetation where shrub clumps of Acacia caven (Molina) Molina and Baccharis spp. are surrounded by grasslands of weedy native and alien species. In some disturbed matorral stands, C. solstitialis is the dominant species in the herbaceous layer, covering around 40% (Gómez-González, unpublished data). Although the impact of C. solstitialis on native matorral species is unknown, its huge abundance in the herbaceous layer might affect the recruitment of native pioneer shrubs. Field and lab experiments suggest that European forbs can limit the colonization of open areas by Baccharis spp. in the matorral of central Chile (Martinez & Fuentes 1993) and also in shrublands of northern California (Williams etal. 1987, Williams & Hobbs 1989). According to the successional model proposed by Armesto & Pickett (1985) for the Chilean matorral, Baccharis species are key elements for the secondary succession because established individuals in open areas could act as "facilitators" of other late successional species, favouring the regeneration of the vegetation after a disturbance. Hence, it is important to evaluate whether the presence of C. solstitialis negatively affect the seedling emergence and performance of Baccharis species, and to detect the life-cycle stage of the invader at which it has the stronger effects. This information could be useful for future management and restoration of matorral areas currently invaded by C solstitialis.

In this study we take the following questions into consideration: i) which are the competitive effects of C solstitialis on Baccharis species at different life-cycle stages? and ii) in what kind of interaction (seed-seed, plant-seed or plant-plant) is the invader more harmful for Baccharis species? To answer these questions we evaluated under controlled conditions the competitive interactions between C. solstitialis and two Baccharis species: B. linearis (Ruiz et Pav.) Pers. and B. paniculata DC. (Asteraceae). We performed pair-wise experiments to explore: 1) seed-seed effects, 2) plant-seed effects, and 3) plant-plant effects. These experiments involved almost the whole life cycle of C. solstitialis and the early life-cycle stages of Baccharis species, in order to simulate the natural situation in which the invader might affect the seedling establishment of these two pioneer native species.

MATERIALS AND METHODS

Studied species

Centaurea solstitialis is a facultative winter annual species (sometimes biennial or short-lived perennial from a tap root). It produces rosette leaves that lie close to the ground, and erect stems 0.15-2 m in height. Centaurea solstitialis has a large taproot that grows to soil depths of 1 m or more (Sheley et al. 1993). The yellow flowerheads produce two types of achenes (hereafter seeds); most of them (75-90%) have a short pappus but some achenes have no pappus (mainly at the periphery of the flowerhead). Large plants can produce nearly 75,000 seeds with 87.6 to 95.2% of viability (Maddox 1981, Benefield et al. 2001). In C solstitialis infestations, seed density in the soil can range from 3,000 to 10,000 seeds per square meter (Sheley & Larson 1994, DiTomaso et al. 1999), and seedbank longevity can be as long as 10 years (Callihan et al. 1993).
Baccharis linearis andB. paniculata are 0.4-3 m tall dioecious shrubs, densely branched, with 2-10 white flowerheads at the top of the branches. These Baccharis species produce great numbers of achenes, but seed viability is relatively low (30-50%, Gómez-González, personal observation). It has been reported that seed density in the soil can range from 212 to 707 and from 500 to 2,400 seeds per square meter for B. paniculata and B. linearis respectively (Martinez & Fuentes 1993, Gutiérrez et al. 2000). Achenes are wind-dispersed and seedlings tolerate high irradiance and drought, and consequently, they successfully establish in cleared areas of the matorral (Armesto & Pickett 1985). Seed dispersal and seedling emergence occur simultaneously to that of C. solstitialis (late summer and late autumn respectively), so Baccharis spp. are interacting with this invader both at the seed (within the seedbank before rains) and the seedling stages.

Seed collections

During January 2004, we collected seeds of C. solstitialis, B. linearis andi?. paniculata in Quebrada de la Plata (33°29' S; 70°52'W), Province of Santiago, Central Chile. Seed collection was carried out across several disturbed matorral stands which were located along a 5 km trail in Quebrada de la Plata. Seeds of at least 50 plant individuals of each species were collected. Seeds were carried to the laboratory, where we carefully observed all seeds with a binocular microscope (Zeiss). Then, only seeds that looked healthy and filled with an embryo (regarded as viable seeds) were selected for the experiments. All maternal lines were pooled and represented in the seed set used for the experiments.

Experiment 1. Seed-seed interactions

To evaluate possible allelopathic effects at the seed stage, we performed germination trials where seeds of each Baccharis species were germinated under the following conditions: i) 30 seeds in absence of other seeds (control), ii) 30 seeds in presence of 30 conspecific seeds (i.e., 60 seeds of the same species), andiii) 30 seeds in presence of 30 C solstitialis seeds. In addition, 30 seeds of C. solstitialis were germinated in each of the following conditions: i) in absence of other seeds (control) and ii) in presence of conspecific seeds. This last treatment (conspecific interaction) was added to distinguish inter-specific effects from density dependant effects. Each treatment was replicated four times. In all treatments, we placed the seeds on Petri dishes with filter paper and distilled water (2.5 ml aprox.). Seed germination trials were carried out in a growth-chamber with a photoperiod of 16 h light and 8 h darkness and a thermoperiod of 12 h at 10°C and 12 h at 20°C. This thermoperiod simulate the temperature conditions during the fall season, when seedling emergence begins for all these species. Every two days and over a total period of 28 days, emerged seedlings were checked and petri-dishes were randomly re-positioned inside the chamber. All petri-dishes were watered when needed. We considered a seed germinated when cotyledons were visible. For each species and treatment, we calculated the following parameters of germination:

1. The coefficient of velocity of germination CV = 100·S Ni /(S Ni·Ti) with Ni being the number of seeds newly germinated on day i, and Ti the number of days needed for germination, i = 1...28.

2.  The final percentage of germination FG = (S Ni)·100/N with Ni being the number of seeds newly germinated on day i, and N the total number of tested seeds, i= 1...28.

Experiment 2. Plant-seed interactions

To evaluate competitive effects of established individuals on seed germination, we sowed 30 seeds of each Baccharis species in the following conditions: i) absence of plant individuals (control), ii) presence of one C. solstitialis individual, and iii) presence of one conspecific individual. Additionally, we sowed 30 seeds of the invasive species in: i) absence of Baccharis individuals (control), ii) presence of one B. linearis individual, iii) presence of one B. paniculata individual, and iv) presence of one conspecific individual. We established four replicates for each species and treatment. We obtained all plant individuals from the seeds germinated as in the experiment 1. After germination, the seedlings (of the same age) were transplanted in 500 ml pots filled with commercial organic soil (C/N=40, pH 5.0-8.5). After 1 month of growing, we sowed 30 seeds of the corresponding treatment on each pot. Seeds were sown equidistant to the plant shoot at 2 cm depth. We randomly placed the pots within a greenhouse. Every four days, all pots were randomly redistributed in order to avoid any effect of micro-environmental differences inside the greenhouse. Every two days, and over a total period of 28 days, all samples were watered and the number of seedlings emerged were recorded. We considered that a seedling was emerged from the soil when the cotyledons were visible. For each species and treatment, we calculated the final percentage of seedlings emerged and the velocity of emergence as described in the experiment 1 (FG y irrespectively).

Experiment 3. Plant-plant interactions

To evaluate competitive interactions between established individuals, we planted one individual of each Baccharis species in the following conditions: i) absence of other individuals (control), ii) presence of one C. solstitialis individual, and iii) presence of one conspecific individual. Additionally, one C. solstitialis plant was grown: i) in absence of other individuals (control), ii) in presence of one conspecific individual. This last treatment (conspecific interaction) was added to distinguish inter-specific effects from density dependant effects. Each treatment was replicated 10 times.

All plant individuals were obtained from seedlings emerged at the same time and they were grown in pots as described in the experiment 2. After three months of growth, we harvested all plants. At this time, C. solstitialis individuals were adults while Baccharis species were yet saplings. Each individual was separated into roots and shoots and dried in an electric oven at 60°C for 3 days. For each plant individual we calculated the root, shoot and total dry biomass, and the root-shoot ratio. In the case of the intraspecific competition treatments, where two conspecific individuals shared the same pot and individual roots were undistinguishable, the root biomass per individual was estimated as the total root biomass within the pot divided by two.

Data analysis

Statistical analyses were performed using the softwares R2.8.0 (RDevelopment Core Team 2008) and STATISTICA 6.0 (StatSoft, Inc. 2001).

Differences among treatments in the FG were analyzed by means of generalized linear mixed models (GLMM), fitted by the Laplace approximation (Raudenbush et al. 2000, Crawley 2007). In these models, the dependent variable was the FG (data expressed as proportion, binomial errors). Independent variables were the treatment (control, presence of competitor) as fixed factor and the block (Petri dishes and pots for the experiment 1 and 2 respectively) as random factor. Wald-Ztests were used to assess the null hypothesis of no treatment effect (i.e., estimated parameters equal zero). We used Mann-Whitney U tests to explore the effect of the presence of competitor on the velocity of germination (CV) of each species. Additionally, these analyses (GLMM and Mann-Whitney Utests for the FG and the CT respectively) were used to evaluate species-specific differences in the germination capability by taking into account only the control treatments. We analyzed biomass data (experiment 3) with one-way ANOVA and Tukey tests, after the logarithmic transformation of the data. These analyses were unbalanced due to the lose of three samples. P-values < 0.05 were regarded as statistically significant.

RESULTS

Germination capability of the species Centaurea solstitialis showed almost 100% of seed germination in only 4 days (Fig.1). In contrast, Baccharis species showed significantly slower germination (lower CV) than the invader (Mann-Whitney U-test, C. solstitialis vs. B. linearis, P= 0.02, C. solstitialis vs. B.paniculata, P=0.02, Fig. 1). Further, the final germination (FG) of C. solstitialis was significantly greater than that of both Baccharis species (GLMM, C. solstitialis vs. B. paniculata, P < 0.001, C. solstitialis vs. B. linearis, P=0.03, Fig.1).

Seed-seed interactions

Treatments did not affect the final germination of any of the three studied species (FG, Table I, Fig. 2). That is, neither the presence of seeds of the same species nor the presence of heterospecific seeds affected their final germination. Furthermore, there were no significant differences in the velocity of germination between treatments in the most of the cases, except for C. solstitialis seeds, which emerged significantly slower in presence of Baccharis seeds compared to the control (Table I, Fig. 2).

Plant-seed interactions

The presence of a conspecific or heterospecific plant individual did not significantly affect the germination (FG and CV) of the species in the most of the cases (Table II, Fig. 3). Only B. paniculata showed higher CV when seeds emerged in the presence of a C. solstitialis plant compared to the control (Table II, Fig. 3).

Plant-plant interactions

Root, shoot and total biomass of C. solstitialis were not significantly affected by intra- or interspecific competition (Fig. 4a-c). Indeed, compared to the control, the invader reduced its biomass by only 11.3% and 7.8% in presence of B. paniculata andB. linearis respectively. Nevertheless, the root-shoot ratio of the invader did significantly increase due to competition with a conspecific individual as well as with Baccharis spp. (Fig. 4d).

In contrast, the presence of C. solstitialis decreased the root, shoot and total biomass of Baccharis species without changing their root-shoot ratio (Fig. 4a-d). Specifically, the total biomass of B. paniculata and B. linearis was 25% lower in presence of C. solstitialis than without competition (Fig. 4c), although this effect was not significant (Tukey test, B. linearis: alone vs. +C, P=0.06; B. paniculata: alone vs. +C, P=0.06, Fig. 4c).

Figure 1. Comparison among the germination curves of the three studied species (regarding only the control treatment of the experiment 1). Mean values ± SE are shown, n = 4. Different letters denote significant differences in the FG among species (GLMM, Wald Z-test, P < 0.05).

Figura 1. Comparación de las curvas de germinación de las tres especies de estudio (considerando sólo el tratamiento control del experimento 1). Se muestran los valores medios ± EE, n = 4. Letras diferentes denotan diferencias significativas en el FG entre las especies (GLMM, prueba Wald Z-test, P < 0,05).

However, there was a significant biomass reduction (around 40%) in both Baccharis spp. in presence of C. solstitialis compared to when they grew with a conspecific individual (Fig. 4c). It must be noted that Baccharis species showed a tendency to increase their biomass in the presence of a conspecific individual compared to the control (growing alone) (Fig. 4c).

Figure 2. Germination curves of B. paniculata (a), B. linearis (b) and C. solstitialis (c) under the treatments of seed-seed interaction (experiment 1). Mean values ± SE are shown, n = 4. See treatment codes in Table I.

Figura 2. Curvas de germinación de B. paniculata (a), B. linearis (b) y C. solstitialis (c) bajo los tratamientos de interacción entre semillas (experimento 1). Se muestran los valores medios ± EE, n = 4. Ver códigos de los tratamientos en la Tabla I.

Table I. Effect of the presence of hetero- and conspecific seeds (experiment 1) on the final germination (FG) and the velocity of germination (CV) of the studied species. Treatments: Alone: germination in the absence of other seeds; +C: presence of C. solstitialis seeds; +Bp: presence of B. paniculata seeds; +B1: presence of B. linearis seeds. Significant P-values are highlighted in bold (P < 0.05, Wald Z-test for FG and Mann-Whitney C/-test for CV). (3: Estimated coefficient in the GLMM. SE: Standard Error.

Tabla I. Efecto de la presencia de semillas hetero- y coespecíficas (experimento 1) sobre la germinación final (FG) y la velocidad de germinación (CV) de las especies estudiadas. Tratamientos: Alone: germinación en ausencia de otras semillas; +C: presencia de semillas de C. solstitialis; +Bp: presencia de semillas de B. paniculata, +B1: presencia de semillas de B. linearis. Valores de P significativos son marcados en negrita (P < 0,05, prueba Z de Wald para FG y prueba U de Mann-Whitney para CV). (3: Coeficiente estimado en el GLMM. SE: Error Estándar.

Table II. Effect of the presence of hetero- and conspecific individuals (experiment 2) on the final germination (FG) and the velocity of germination (CV) of the studied species. Treatments: Alone: emergence in the absence of plant individuals; +C: presence of a C solstitialis individual; +Bp: presence of a B. paniculata individual; +B1: presence of a B. linearis individual. Significant P-values are highlighted in bold (P < 0.05, Wald Z-test forFG and Mann-Whitney C/-test for CV). (3: Estimated coefficient in the GLMM. SE: Standard Error.

Tabla II. Efecto de la presencia de individuos hetero- y coespecíficos (experimento 2) sobre la germinación final (FG) y la velocidad de germinación (CV) de las especies estudiadas. Tratamientos: Alone: emergencia en ausencia de plantas establecidas; +C: presencia de un individuo de C. solstitialis; +Bp: presencia de un individuo de B. paniculata; +B1: presencia de un individuo de B. linearis. Valores de P significativos son marcados en negrita (P < 0,05, prueba Z de Wald para FG y prueba U de Mann-Whitney para CV). (3: Coeficiente estimado en el GLMM. SE: Error Estándar.

 

Figure 3. Emergence curves of B. paniculata (a), B. linearis (b) and C. solstitialis (c) under the treatments of plant-seed interaction (experiment 2). Mean values ± SE are shown, n = 4. See treatment codes in Table II.

Figura 3. Curvas de emergencia de B. paniculata (a), B. linearis (b) y C. solstitialis (c) bajo los tratamientos de interacción planta-semilla (experimento 2). Se muestran los valores medios ± EE, n = 4. Ver códigos de los tratamientos en la Tabla II.

Figure 4. Effect of plant-plant interaction (experiment 3) on the average root biomass (a), shoot biomass (b), total biomass (c), and root-shoot ratio (d) per plant of the studied species. Treatments: Alone: growing without competition; +C: growing with C. solstitialis; +Bp: growing with B. paniculata; +B1: growing with B. linearis. Mean values ±SE are shown, n = 10. Different letters indicate significant differences among treatments (P < 0.05, Tukey test post- ANO VA).

Figura 4. Efecto de la interacción planta-planta (experimento 3) sobre el promedio de la biomasa de raíces (a), la biomasa de tallos (b), y la razón raíz-tallo (d) por planta de las especies estudiadas. Tratamientos: Alone: creciendo sin competencia; +C: creciendo conC. solstitialis; +Bp: creciendo conB. paniculata; +B1: creciendo con5. linearis. Se muestran los valores medios ± EE, n= 10. Letras diferentes denotan diferencias significativas entre tratamientos (P < 0,05, prueba de Tukey post-ANO VA).

DISCUSSION

Under the controlled conditions of our experiments, we found that C. solstitialis produced different effects on Baccharis species depending on the life-cycle stage at which plant interactions occurred. At the seed stage, while there were no allelopathic effects of the invader on Baccharis spp., C. solstitialis seeds did emerge later in presence of B. linearis seeds. The presence of established C. solstitialis individuals did not affect negatively the seed germination of the natives. Instead of this, seed germination of B. paniculata was accelerated in presence of one C. solstitialis individual. However, established C. solstitialis individuals reduced the performance (biomass) of both native Baccharis species.

Seed germination capability and seed-seed allelopathic effects

Centaurea solstitialis had higher velocity of germination than Baccharis species. After only 4 days, 95% of C. solstitialis seeds emerged, whereas Baccharis species did not reach 10% of germination. These differences in the germination rate could be explained by the differences in their life history, since C. solstitialis is an annual herb and Baccharis species are shrubs. However, it has been found that the germination rate of C. solstitialis is much higher than that of many other matorral species independently of their life history, including weedy natives and alien grasses (Sierra-Almeida & Cavieres, unpublished data).

In C. solstitialis, the early emergence together with its elevated seed production and viability (Maddox 1981, Benefieldeia/. 2001) could be traits associated with its successful invasion in the Chilean matorral and also in other Mediterranean-type ecosystems (Piper 2001). In Mediterranean-type ecosystems early emergence is particularly important because seedlings that emerge earlier can produce an important amount of below-ground biomass before the onset of summer drought (Verdu & Traveset 2005). In the Chilean matorral, most of the species (including C. solstitialis) emerge during the fall after the first rains (Figueroa & Jaksic 2004) and hydric resources for seedlings establishment are available for a very short time. Hence, the massive and fast germination of C. solstitialis could be an advantageous trait for the occupation of available sites and early resources uptake. Indeed, we have observed that this high germination result in extremely dense seedling populations in the matorral. In ecosystems from eastern Washington, Talbott (1987) reported seedling densities approaching 27,000 individuals per square meter, and Piper (2001) has pointed out that C. solstitialis effectively eliminates the emergence or growth of competing vegetation at high densities. Thus, future investigation should be destined to explore whether C. solstitialis is limiting the seedling emergence of Baccharis species in the field. This process is probable to occur in those disturbed matorrals that are highly invaded by this species.

On the other hand, seed germination of C. solstitialis was delayed by the presence of Baccharis seeds. Other studies have also shown allelopathic effects among seeds (Laterra & Bazzalo 1999, Casini & Olivero 2001). This effect could be mediated by allelopathic compounds accumulated in the seed coat. Kuti et al. (1990) showed that some potent phytotoxic compounds (roridins and baccharinoids) are accumulated in the seed coat of several Baccharis species. These allelochemicals interact with the gibberellic acid and thus inhibit the germination of other species (Kuti etal. 1990). Since the seed density of Baccharis species is relatively high in the soil seed bank of the matorral (500 to 2,400 seeds per square meter, Gutiérrez etal. 2000), the explosive emergence of C. solstitialis might be counteracted in some way.

One limitation of our experiment on seed-seed interactions was the fact that we did not test the viability of the non-germinated seeds, and thus the final percentage of germination might have been underestimated in some cases. However, we found high values of germination (over 70%) for all species and treatments, indicating that the studied species do not seem to show seed dormancy as a strategy (Baskin & Baskin 1998). Therefore, it is probably that the most of the remaining non-germinated seeds were not viable.

Plant-plant competitive effects

According to Vila & Weiner (2004), the study of competitive interactions between invasive and native species should include both the effect of the invader on the native and the effect of the native on the invader (native species resistance), because the invasion success results from the balance of both effects. In our experiment, C. solstitialis had negative effects on the biomass of Baccharis species but not vice versa. This competitive superiority could be due to differences in the life history between the invader and Baccharis species. However, Quin et al (2007) reported the same results when they performed greenhouse experiments to evaluate the competitive interactions between C. solstitialis and five herbaceous species natives to California grasslands. Widmer etal. (2007) suggest that C. solstitialis success in California is not fully explained by its life history traits (i.e., prolific seed production, high seed viability, deep-root system, etc.), because those traits are also present in its native range. In fact, Widmer et al. (2007) showed that C. solstitialis has changed the resource allocation in the invaded range, since seeds have larger reserve of starch compared to its native range. As consequence, seedlings are larger in the invaded range and it could give these plants an early competitive advantage against native plants (Widmer et al. 2007). This support the hypothesis that C. solstitialis has evolved in the invaded range, increasing its competitive ability (EIC A hypothesis, Blossey & Notzold 1995). Thus, the negative effects of C. solstitialis onBaccharis species seems not to be caused only by differences in their life history traits, but maybe also by the higher competitive ability that C. solstitialis acquires after evolving in the invaded range.

As suggested by Callaway et al. (2006) and Quin et al. (2007), unlike other invasive Centaurea species, C. solstitialis does not appear to be allelopathic. Our results indicated that the presence of C. solstitialis plants did not affect negatively the seed germination of both Baccharis species, and its effect on established Baccharis plants was not lethal. Instead of allelopathy, our results and the recent evidence indicate that the success of C. solstitialis is mediated by below-ground competition (Quin et al. 2007). When C. solstitialis competed with the Baccharis species the total biomass did not change but the root: shoot ratio increased. In other words, C. solstitialis modified resource allocation in presence of Baccharis species, producing a greater proportion of roots. This is a plastic response that might allow C. solstitialis to be drought tolerant in spite of the presence of competitors (Karcher et al. 2008). Furthermore, it is well known that root growth of C. solstitialis is very fast during the winter and the beginning of spring, reaching depths up to lm (Sheley etal. 1993). In California, Enloeeia/. (2004) found that soils of plant communities dominated by C. solstitialis are significantly drier than those dominated by native grasses. Thus, we suspect that C. solstitialis could out-compete Baccharis seedlings through its ability of reaching the deeper soil layer before the dry season, and also through the plasticity of root: shoot allocation. Light-mediated competition might also play an important role in reducing Baccharis performance. In our experiment, C. solstitialis showed high growth rate and Baccharis plants were rapidly shaded. Shading can be strong in matorral areas dominated by C. solstitialis; because the high seedling density at the onset of the rainy season is added to the presence of old stalks that remain standing from the last summer.

The importance of evaluating the competitive effects of C. solstitialis on seedling emergence and performance of Baccharis species resides in the fact that these species are key elements in the natural recover of the matorral after a disturbance (Armesto & Pickett 1985). Martinez & Fuentes (1993) have shown that some European forbs (e.g., Erodium cicutarium and Trifolium sp.) can limit the colonization of open areas by Baccharis spp. in the matorral of central Chile. The role of alien grasses and forbs in suppressing native shrubland re-establishment has been also reported in California shrublands (Williams et al. 1987, Williams & Hobbs 1989, Eliason & Allen 1997) and more recently in neotropical savannas (Hoffmann & Haridasan 2008). Hence, in those matorral areas in which C. solstitialis is dominant (around 40% cover), the natural succession could be modified if the seedling establishment of Baccharis spp. is limited by competition with the invader (Eliason & Allen 1997). However, more research including field experiments are needed to evaluate this hypothesis.

An unexpected result was that the biomass of both Baccharis species showed certain tendency to increase under conspecific competition compared to control (growing alone). A possible reason for this pattern is that conspecific neighbours could have protected each other from desiccation. Facilitation among conspecific plants of the same age is a kind of interaction that can be relevant for the seedling establishment of plant species in some arid and semiarid ecosystems (Goldberg etal. 2001, Franks 2003). Then, it would be interesting to assess whether seedling survival of these Baccharis species are really being facilitated by conspecifics in the matorral, since they may better resist Centaurea invasion at high densities.

To conclude, our results and available evidence suggest that different mechanisms could be involved together in the invasion success of C solstitialis in recipient communities. At the seed stage, early emergence could be an important invasion mechanism to displace native species by means of site pre-emption, especially in Mediterranean environments where soil resources are available for a very short time. Then, the high competitive ability of established individuals and the plasticity of rootshoot allocation could allow C. solstitialis to monopolize soil resources, reducing their neighbours' growth. Regarding the effects of C. solstitialis on the studied Baccharis species, we propose that competition at the stage of seedling establishment would be the key process in limiting their natural recruitment in the Chilean matorral.

ACKNOWLEDGENTS

We thank anonymous reviewers for their valuable comments and suggestions. Research funded by MECESUPUCO-0214, CONICYT AT-24060008, and Grant No. P05-002 F ICM supporting the Institute of Ecology and Biodiversity (IEB).

REFERENCES

Armesto, J.J. & S.T.A. Pickett. 1985. A mechanistic approach to the study of succession in the Chilean matorral. Revista Chilena de Historia Natural 58: 9-17.

Bais, H., R. Vepachedu, S. Gilroy, R.M. Callaway & J.Vivanco. 2003. Allelopathy and exotic plant invasion: from molecules and genes to species interactions. Science 301: 1377-1380.

Baskin, C.C. & J.M. Baskin. 1998. Seeds. Ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego. 666 pp.

Benefield, C.B., J.M. Ditomaso, G.B. Kyser & A. Tschohl. 2001. Reproductive biology of yellow starthistle: maximizing late-season control. Weed Science 49: 83-90.

Bergelson, J. & R. Perry. 1989. Interspecific competition between seeds: relative planting date and density affect seedling emergence. Ecology 70:1639-1644.

Blossey, B. & R. Notzold. 1995. Evolution of increased competitive ability in invasive non-indigenous plants: a hypothesis. Journal of Ecology 83: 887-889.

Callaway, R.M. & A. Aschehoug. 2000. Invasive plants versus their new and old neighbours: a mechanism for exotic invasion. Science 290: 521-523.

Callaway, R.M., J. Kim & B.E. Mahall. 2006. Defoliation of Centaurea solstitialis stimulates compensatory growth and intensifies negative effects on neighbors. Biological Invasions 8: 1389-1397.

Callihan, R.H., T.S. Prather& F.E.Northam. 1993. Longevity of yellow starthistle (Centaurea solstitialis) achenes in soil. Weed Technology 7: 33-35.

Casini, P. & L. Olivero. 2001. Allelopathic effects of legume cover crops on cogon grass (Imperata brasiliensis Trin.). Allelopathy Journal 8: 189-199.

Coleman, H.M. & J.M. Levine. 2007. Mechanisms underlying the impacts of exotic grasses in a coastal California meadow. Biological Invasions 9: 65-71.

Crawley, M.J. 2007. The R Book. John Wiley & Sons, Ltd., Chichester. viii + 942 pp.

Di Tomaso, J.M., GB. Kyser & M.S. Hastings. 1999. Prescribed burning for control of yellow starthistle [Centaurea solstitialis) and enhanced native plant diversity. Weed Science 47: 233-242.

Eliason, S.A. & E.B. Allen. 1997. Exotic grass competition in suppressing native shrubland re-establishment. Restoration Ecology 5: 245-255.

Enloe, S.F., J.M. Di Tomaso, S.B. Orloff & D.J. Drake. 2004. Soil water dynamics differ among rangeland plant communities dominated by yellow starthistle [Centaurea solstitialis), annual grasses, or perennial grasses. Weed Science 52: 929-935.

Ens, E.-J. & K. French. 2008. Exotic woody invader limits the recruitment of three indigenous plant species. Biological Conservation 141: 590-595.

Figueroa, J.A. & F.M. Jaksic. 2004. Latencia y banco de semillas en plantas de la región mediterránea de Chile central. Revista Chilena Historia Natural 77: 201-215.

Franks, S.J. 2003. Competitive and facilitative interactions within and between two species of coastal dune perennials. Canadian Journal of Botany 81: 330-337.

Goldberg, D.E., R. Turkington, L. Olsvig-Whittaker & A.R. Dyer. 2001. Density dependence in an annual plant community: Variation among life history stages. Ecological Monographs 71:423-446.

Gordon, D.R., J.M. Welker, J.W. Menke & K.J. Rice. 1989. Competition for soil water between annual plants and blue oak (Quercus douglasii) seedlings. Oecologia 79: 533-541.

Gutiérrez, J.R., G Arancio & F.M. Jaksic. 2000. Variation in vegetation and seed bank in a Chilean semi-arid community affected by ENSO 1997. Journal Vegetation Science 11: 641-648.

Hierro, J.L., D. Villarreal, O. Eren, J.M. Graham & R.M. Callaway. 2006. Disturbance facilitates invasion: the effects are stronger abroad than at home. American Naturalist 168: 144-156.

Hoffmann, W. & M. Haridasan. 2008. The invasive grass, Melinis minutiflora, inhibits tree regeneration in a Neotropical savanna. Austral Ecology 33: 29-36.

Jefferson, L.V & M. Pennacchio. 2003. Allelopathic effects of foliage extracts from fourChenopodiaceae species on seed germination. Journal of Arid Environment 55: 275-285.

Karcher, D.E., M.D. Richardson, KHignight & D. Rush. 2008. Drought Tolerance of Tall Fescue Populations Selected for High Root/Shoot Ratios and Summer Survival. Crop Science 48:771-777.

Kuti, J.O., B.B. Jarvts, N. Mokhtarirejali & GA. Bean. 1990. Allelochemical regulation of reproduction and seed-germination of two brazilian Baccharis species by phytotoxic trichothecenes. Journal Chemical Ecology 16: 3441-3453.

Laterra, P. & M.E. Bazzalo. 1999. Seed-to-seed allelopathic effects between two invaders of burned Pampa grasslands. Weed Research 39:297-306.

Levine, J. M. 2000. Species diversity and biological invasions: relating local process to community pattern. Science 288: 852-854.

Levine, J.M., M. Vila, CM. D'antonio, J.S. Dukes, K.Grigulis & S. Lavorel. 2003. Mechanisms underlying the impacts of exotic plant invasions. Proceedings of the Royal Society London B 270: 775-781.

Lortie, C.J. & R. Turkington. 2002. The facilitative effects by seeds and seedlings on emergence from the seed bank of a desert annual plant community. Ecoscience9: 106-111.

Mack, R.N., D. Simberloff, WM. Lonsdale, H. Evans, M. Clout & F.A. Bazzaz. 2000. Biotic invasions: causes, epidemiology, global consequences, and control. Ecological Applications 10: 689-710.

Maddox, D.M. 1981. Introduction, phenology, and density of yellow starthistle in coastal, intercoastal, and central valley situations in California. U.S. Department of Agriculture, Agricultural Research Service, Agricultural Research Results no. ARR-W-20,1-33.

Martínez, E. & E. Fuentes. 1993. Can we extrapolate the California model of grassland-shrub land ecotone? Ecological Applications 3: 417-423.

Matthei, O. 1995. Manual de las malezas que crecen en Chile. Alfabeta Impresores, Santiago. 545 pp.

Miller, T.E. 1987. Effects of emergence time on survival and growth in an early old-field plant community. Oecologia 72: 272-278.

Muoghalu, J.I. 2008. Growth, reproduction and resource allocation of Tithonia diversifolia and Tithonia rotundifolia. Weed Research48: 157-162.

Murray, B.R. 1998. Density-dependent germination and the role of seed leachate. Australian Journal of Ecology 23: 411-418.

Orr, S.P, JA. Rudgers & K. Clay. 2005. Invasive plants can inhibit native tree seedlings: testing potential allelopathic mechanisms. Plant Ecology 181:153-165.

Piper, J.L. 2001. The biological control of yellow starthistle in the western U.S.: four decades of progress. In: Proceedings of the first international knapweed symposium of the twenty-first century, 15-16 March 2001. (Ed. L. Smith), pp. 48-55. Coeur d'Alene, Idaho, USA.

Prati, D. & O. Bossdorf. 2004. Allelopathic inhibition of germination by Alliaria petiolata (Brassicaceae). American Journal of Botany 91: 285-288.

Quin, B., JA. Lau, J. Kopshever, R.M. Callaway, H. McGray, L.G Perry, T.L. Weir, M.W Paschke, J.L. Hierro, J. Yoder, J.M. Vivanco & S. Strauss. 2007. No evidence for root-mediated allelopathy in Centaurea solstitialis, a species in a commonly allelopathic genus. Biological Invasions 9: 897-907.

R Development Core Team. 2008. R: a language and environment for statistical computing. Available from: http://www.rproject.org

Raudenbush, S.W, M-L. Yang & M. Yosef. 2000. Maximum likelihood for generalized linear models with nested random effects via high-order, multivariate Laplace approximation. Journal of Computational and Graphical Statistics 9: 141-157.

Sheley, R.L., L.L. Larson & D.E. Johnson. 1993. Germination and root dynamics of range weeds and forage species. Weed Technology 7:234-237.

Sheley, R.L. & L.L. Larson. 1994. Observation: comparative life-history of cheatgrass and yellow starthistle. Journal of Range Management 47: 450-456.

Statsoft, Inc. 2001. STATISTICA (data analysis software system), version 6. http://www.statsoft.com

Talbott, C.J. 1987. Distribution and ecologic amplitude of selected Centaurea species in eastern Washington. M.S. Thesis, Washington State University, Pullman. 186 p.

Verdú, M. & A. Traveset. 2005. Early emergence enhance plant fitness: a phylogenetically controlled meta-analysis. Ecology 86: 1385-1394.

Vila, M. & J. Weiner. 2004. Are invasive plant species better competitors than native plant species?: evidence from pair-wise experiments. Oikos 105:229-238.

Vivanco, J., H. Bais, F. Stermitz, G. Thelen & R.M. Callaway. 2004. Biogeographical variation in community response to root allelochemistry: novel weapons and exotic invasion. Ecology Letters 7:285-292.

Welker, J.M., D.R. Gordon & K.J. Rice. 1991. Capture and allocation of nitrogen by Quercus douglasii seedlings in competition with annual and perennial grasses. Oecologia 87: 459-466.

WlDMER, T.L., F. GUERMACHE, MY. DOLGOVSKAIA & SY. Reznik. 2007. Enhanced growth and seed properties in introduced vs. native populations of yellow starthistle (Centaurea solstitialis). Weed Science 55: 465-473.

Williams, K. & R.J. Hobbs. 1989. Control of shrub establishment by springtime soilwater availability in an annual grassland. Oecologia 81: 62-66.

Williams, K., R.J. Hobbs & S.P. Hamburg. 1987. Invasion of an annual grassland in northern California by Baccharispilularis spp. consanguínea. Oecologia 72: 461-465.

Wilson, J.B. 1988. The effect of initial advantage on the course of plant competition. Oikos 51: 19-24.

Yurkonis, K.A., S.J. Meiners & B.E. Wachholder. 2005. Invasion impacts diversity through altered community dynamics. Journal of Ecology 93: 1053-1061.

Recibido: 27.10.08

Aceptado: 23.02.09