OUP user menu

Changes in soil diversity and global activities following invasions of the exotic invasive plant, Amaranthus viridis L., decrease the growth of native sahelian Acacia species

Arsene Sanon, Thierry Béguiristain, Aurelie Cébron, Jacques Berthelin, Ibrahima Ndoye, Corinne Leyval, Samba Sylla, Robin Duponnois
DOI: http://dx.doi.org/10.1111/j.1574-6941.2009.00740.x 118-131 First published online: 3 September 2009


The objectives of this study were to determine whether the invasive plant Amaranthus viridis influenced soil microbial and chemical properties and to assess the consequences of these modifications on native plant growth. The experiment was conducted in Senegal at two sites: one invaded by A. viridis and the other covered by other plant species. Soil nutrient contents as well as microbial community density, diversity and functions were measured. Additionally, five sahelian Acacia species were grown in (1) soil disinfected or not collected from both sites, (2) uninvaded soil exposed to an A. viridis plant aqueous extract and (3) soil collected from invaded and uninvaded sites and inoculated or not with the arbuscular mycorrhizal (AM) fungus Glomus intraradices. The results showed that the invasion of A. viridis increased soil nutrient availability, bacterial abundance and microbial activities. In contrast, AM fungi and rhizobial development and the growth of Acacia species were severely reduced in A. viridis-invaded soil. Amaranthus viridis aqueous extract also exhibited an inhibitory effect on rhizobial growth, indicating an antibacterial activity of this plant extract. However, the inoculation of G. intraradices was highly beneficial to the growth and nodulation of Acacia species. These results highlight the role of AM symbiosis in the processes involved in plant coexistence and in ecosystem management programs that target preservation of native plant diversity.

  • exotic plant invasion
  • Amaranthus viridis
  • soil microbial community
  • mycorrhizae


Varied and considerable impacts on native flora and fauna can result from exotic plant species that invade natural ecosystems and public lands (Adair & Groves, 1998). Exotic plants may become invasive in their introduction to a geographical area through a number of biological processes such as higher performance in their introduced ranges (Thébaud & Simberloff, 2001), release from their native specialized antagonists (Mitchell & Power, 2003), direct chemical interference (allelopathic effect) with native plant species (Callaway & Ridenour, 2004) and resistance variability of native plant communities to invasion (Levine & D'Antonio, 1999). These disturbances of the native plant ecosystem resulting from exotic plant invasion modify succession, dominance, community structure and composition and vegetation dynamics (del Moral & Muller, 1970). It has also been suggested that exotic plants could disrupt mutualistic associations within native microbial communities (Richardson et al., 2000; Callaway & Ridenour, 2004; Stinson et al., 2006). Terrestrial ecosystems' functioning and stability are mainly determined by plant-specific richness and composition, and accumulating evidences indicated that plant species dynamics are strongly interlinked with soil microbiota (Grayston & Campbell, 1996; Hooper & Vitousek, 1997; Hart et al., 2003; Kisa et al., 2007). Therefore, one of the main success ways of exotic invasive plant species might result from these exogenous organisms-mediated modifications in the structure and activities of soil microbial communities (Kourtev et al., 2003; Wolfe & Klironomos, 2005; Batten et al., 2006; Mummey & Rillig, 2006; Stinson et al., 2006).

Among soil microbial communities, arbuscular mycorrhizal (AM) fungi form an important component of the sustainable soil–plant system (Bruno et al., 2003; Johansson et al., 2004; Finlay, 2008). AM symbiosis influences plant development (Schreiner et al., 2003; Duponnois et al., 2005) as well as plant community diversity (van der Heijden et al., 1998b; O'Connor et al., 2002) and could affect relationships between plants (van der Heijden et al., 2003). Another important group of mutualists are the nitrogen-fixing bacterial associates of legumes called rhizobia (Sprent, 2001). Rhizobia induce root nodules and fix atmospheric nitrogen into ammonium that is transferred to their leguminous hosts. This rhizobial symbiosis is important because nitrogen is one of the main nutrients that limits plant productivity in natural ecosystems, especially under tropical environmental conditions. More recently, it has been demonstrated that symbiotic interactions between legumes and rhizobia contribute to plant productivity, plant community structure and acquisition of limiting resources in legume-rich grassland communities (van der Heijden et al., 2006).

Amaranthus viridis L. (Amaranthaceae) is an annual weed native from Central America. This plant is now largely widespread in all warm parts of the world including regions between 30°S and 30°N (Le Bourgeois & Merlier, 1995). Amaranthaceae are characterized by their nitrophily and the C4 photosynthesis pathway. These physiological traits are advantageous in man-made habitats, under conditions of fertilization and/or irrigation (Maillet & Lopez-Garcia, 2000). Furthermore, A. viridis and other species of Amaranthus are also characterized by an extended period of germination, rapid growth and high rates of seed production (Bensch et al., 2003) that might facilitate plant invasiveness (Rejmanek & Richardson, 1996). In Pakistan, where A. viridis is considered as an invasive plant species, Hussain et al. (2003) observed that aqueous extracts from shoots, rain leachate, litter and root exudation significantly reduced the germination and seedling growth of pearl millet (Pennisetum americanum), wheat (Triticum aestivum) and corn (Zea mays). Moreover, they also found that the shoot extract of this exotic plant retarded the development of meristematic cells of test species underlying the allelopathic potential of A. viridis against crops. Of the 50 species of the genus Amaranthus, nine of them, including A. viridis, have already been considered invasive and noxious in the United States and Canada (USDA Plant Database). In Senegal, A. viridis was reported in agrosystems, and increasingly invaded fallow lands, areas of pasture and domestic waste deposit areas. Its growth is positively correlated to soil fertility (e.g. organic matter and nitrogen contents) (Le Bourgeois & Merlier, 1995). Amaranthus viridis, by invading large areas, compromised native plant (grass-, shrub- and woodland) growth. The mechanisms underlying A. viridis' capacity to enter and proliferate within sahelian native plant communities have not yet been investigated. Among native threatened plants, sahelian Acacia species represent keystone trees in sub-Saharan arid and semi-arid areas. Indeed, they are particularly adapted to these regions' harsh climatic conditions and some of them prevent wind and rain erosion, control sand dunes, are sources of wood and provide fodder for browsing livestock. These legume trees have high economic value, and their rhizobial and mycorrhizal symbiosis improves soil fertility (Dreyfus & Dommergues, 1981; Duponnois et al., 2001).

The purpose of this study was to examine the potential for soil-mediated effects resulting from A. viridis establishment, with an emphasis on effects mediated via symbiotic soil microorganisms such as AM fungi and rhizobia, to affect the growth of several sahelian Acacia species [Acacia raddiana thorns, Acacia senegal (L.) Willd. Syn., Acacia seyal Del., Acacia nilotica Willd. ex Del. and Faidherbia albida (Del.) A. Chev], the most representative Acacia species in West Africa. We hypothesized that soil microbiota will respond to A. viridis by changing in both structure and global activities. We studied the modifications on the growth of Acacia species and their mycorrhizal and rhizobial status. We further hypothesized that an enhancement of AM propagule density (through controlled AM fungal inoculation of Acacia species) would minimize the effect of this invasive plant species.

Materials and methods

Experimental ecosystem area and soil sampling

The experimental area was located in the IRD experimental station of Bel Air (14°43′N, 17°26′W) (Dakar, Senegal). The climate is semi-arid with a long dry season (6–8 months), a mean annual temperature of 24 °C and a mean annual rainfall of 300 mm (Gassama-Dia et al., 2003). The plant cover was sparse and degraded. The vegetation is composed of grasses (Alysicarpus ovalifolius L., Fabaceae; Boerhavia diffusa L., Nyctaginaceae; Commelina forskalaei Vahl, Commelinaceae, Eragrostis tremula L., Poaceae, etc.) and tree species (Acacia spp., Mimosaceae; Leucaena spp., Mimosaceae; Balanites aegypticum, Zygophyllaceae), and A. viridis was highly abundant. The average soil physicochemical characteristics were as follows: pH (H2O) 8.2, clay 5.2%, fine silt 1.8%, coarse silt 2.0%, fine sand 66.5%, coarse sand 24.5%, carbon 0.36%, total nitrogen 0.04% and soluble phosphorus 39.6 mg kg−1. A total of 10 sampling (1 × 1 m quadrat) plots were distributed in the studied area according to the presence or not of A. viridis plants (five quadrats with a history of A. viridis and five quadrats without a history of A. viridis, but covered with grasses). Then the soil (0–15-cm depth) of each quadrat was collected, mixed and sieved (mesh size, <2 mm) to remove plant root materials and kept at 4 °C before further analysis. All soil samples were characterized by measuring the total carbon, total nitrogen, total phosphorous and soluble phosphorous contents (Olsen et al., 1954; Aubert, 1978) in the Laboratoire des Moyens Analytiques (LAMA) [Certifié International Standard for Organization 9001, version 2000, Dakar, Senegal; US Imago (Unité de Service Instrumentations, Moyens Analytiques, Observatoires en Géophysique et Océanographie), Institut de Recherche pour le Développement (http://www.lama.ird.sn)].

Measurement of soil microbial global activities

The total microbial activity in soil samples was measured using the fluorescein diacetate [3′, 6′,-diacetylfluorescein (FDA)] hydrolysis assay according to the method of Alef (1998) in which the fluorescein released was assayed colorimetrically at 490 nm, after 1 h of soil incubation. The total microbial activity was expressed as milligrams of product corrected for background fluorescence per hour and per gram of soil. Dehydrogenase activity was measured by readings at A490 nm following the method of Prin et al. (1989) and Schinner et al. (1996) with iodo nitrotetrazolium (INT) as an artificial electron acceptor to form INT-formazan (INTF).

Assessment of soil microbial structure

Bacterial community 16S rRNA gene copy number quantification

Real-time PCR assays were used to quantify the 16S rRNA gene copy number in soil samples according to Cébron et al. (2008). Total DNA from 0.5 g of soil samples were extracted using a bead-beating protocol (Cébron et al., 2008). Samples were mixed with glass beads, 800 μL of extraction buffer [100 mM Tris, 100 mM EDTA, 100 mM NaCl, 1% (w/v) polyvinylpolypyrrolidone, 2% (w/v) sodium dodecyl sulfate, pH 8.0] and 40 μL of 6% CTAB in 5 mM CaCl2. After bead beating on a horizontal grinder Retsch (Roucaire Instruments Scientifiques, France), DNA was extracted using phenol/chloroform/isoamyl alcohol (25/24/1) and washed twice with chloroform/isoamyl alcohol (24/1). DNA precipitation was carried out using isopropanol and then resuspended in 100 μL Tris-HCl buffer (10 mM, pH 8) and stored at −20 °C.

A real-time PCR experiment was conducted in an iCycler iQ (Bio-Rad) associated with iCycler optical system interface software (version 2.3; Bio-Rad). Real-time PCR were performed in 25-μL reaction volumes containing 1 × iQ SYBR Green Supermix (Bio-Rad), 0.4 μM of each primers (968f and 1401r), 0.9 μg μL−1 bovine serum albumin (BSA), 0.5 μL dimethyl sulfoxide (DMSO), 0.1 μL of T4 bacteriophage gene 32 product (QBiogene) and 1 μL of template DNA or distilled water (negative control). Ten times dilution series of a plasmid standard with a known concentration from 108 to 101 target gene copies per microliter were used for a quantification calibration curve. The following temperature profiles were used for rRNA gene amplifications: step one, heating to 95 °C (5 min), followed by 50 cycles of four steps of 30 s of denaturation at 95 °C, 30 s at the primer-specific annealing temperature (56 °C) and 30 s of elongation at 72 °C, and the SYBR Green I signal intensities were measured during a 10-s step at 80 °C. The final step consisted of 7 min at 72 °C. Then a melting curve analysis was performed as a final step by measuring the SYBR Green I signal intensities during a 0.5 °C temperature increment every 10 s from 51 to 95 °C.

Temporal thermal gradient gel electrophoresis (TTGE) analysis

The eubacterial primer set, 968f-GC [with the GC clamp described by Muyzer et al. (1993)] and 1401r, was used to amplify a 475-bp fragment of the 16S rRNA gene (Felske et al., 1998; Heuer et al., 1999). For each amplification reaction, 1 μL of crude DNA extract was added to 49 μL of PCR mix consisting of 50 mM buffer, 3 mM MgCl2, 0.2 mM dNTPs, 3 mg mL−1 of BSA, 1.5 μL of DMSO, 2 μM of (each) primer and 5 U of Taq polymerase (FastStart; Roche Diagnostic). DNA was amplified in an iCycler (Bio-Rad) with the following amplification program: 94 °C for 5 min (one cycle); 94 °C for 40 s, 56 °C for 30 s and 72 °C for 40 s (38 cycles); and 72 °C for 5 min (one cycle). PCR products, stained with ethidium bromide, were visualized under UV light after electrophoresis in a 1% (w/v) agarose gel to verify the size and quantity of amplified PCR products.

16S rRNA gene fingerprints of the bacterial community present in soil samples were performed by TTGE using the Dcode universal mutation detection system (Bio-Rad Laboratories) under the conditions modified from Corgié (2004). This molecular method has already been proven to give sufficient and reproducible information about modifications in the soil microbial community structure (Muyzer & Smalla, 1998) and has been used in various environmental samples (Halos et al., 2006). The polyacrylamide gels [6% (w/v) acrylamide, 0.21% (w/v) bisacrylamide, 7 M urea, 1.25 × Tris-acetate-EDTA and 0.2% (v/v) glycerol] were allowed to polymerize for 1 h. DNA samples (10 μL) were separated by electrophoresis in 1.25 × Tris-acetate-EDTA at a constant voltage (100 V), with a temperature gradient from 57 to 67 °C (temperature increment of 2 °C h−1). After electrophoresis, gels were stained with ethidium bromide and enumerated under UV light.

Quantification of band intensity was performed using Gel Doc (Biorad) coupled to quantity one software. For each profile, the intensity of the bands was summed and the relative intensity of each band was calculated in order to assess the abundance of the species in the initial sample (Marschner & Baumann, 2003).

Assessment of AM fungus community structure

AM hyphal length was measured on membrane filters according to Jakobsen et al. (1992). The total hyphal length was estimated using the Gridline intersect method (Hanssen et al., 1974). The AM fungi hyphae were distinguished from hyphae of other soil fungi following the morphological criteria described by Nicolson (1959). AM fungal spores were also extracted from soil samples by wet sieving and decanting, followed by sucrose centrifugation (Gerdemann & Nicolson, 1963). After centrifugation, the supernatant was poured through a 50-mm sieve and rinsed with tap water. Spores were counted under a stereomicroscope and grouped according to morphological characteristics. The relative abundance of each fungal type was calculated per 100 g of dry soil. Spore size and color were assessed in water under a stereomicroscope (Olympus SZ H10 research stereomicroscope) whereas spore wall structures and other attributes were observed on permanent slides prepared according to Azcon-Aguilar et al. (2003) under a microscope connected to a computer with a digital image analysis software. Morphotype classification to the genus level, and when possible to the species, was mainly based on morphological features such as color, size, wall structure and hyphal attachment (Morton & Benny, 1990; INVAM, 1997). The relative abundance of each fungal species in each treatment was calculated.

Greenhouse experiments

Assessment of A. viridis mycorrhizal dependence

The AM fungus Glomus intraradices, Schenk and Smith (DAOM 181602, Ottawa Agricultural Herbarium) was propagated on maize (Z. mays L.) for 12 weeks under greenhouse conditions in calcined clay (Plenchette et al., 1996). Maize plants were uprooted, gently washed and colonized roots were hand cut into 1–3-mm-long pieces, bearing around 250 vesicles cm−1, each considered as one propagule (Plenchette, 2000). To obtain a logarithmic scale of inoculum density (0, 3, 30 and 100 propagules per 100 g soil), AM root pieces were counted under a dissecting microscope and, for each inoculum rate, the number was adjusted to 100 root pieces per 100 g of soil with nonmycorrhizal maize roots, prepared as above. Root pieces were then thoroughly mixed with a disinfected sandy soil (120 °C, 60 min) whose physicochemical characteristics were as follows: pH (H2O) 5.3, clay (%) 3.6, fine silt (%) 0.0, coarse silt (%) 0.8, fine sand (%) 55.5, coarse sand (%) 39.4, carbon (%) 0.17, nitrogen (%) 0.02, total phosphorous (mg kg−1) 39 and Olsen phosphorous (mg kg−1) 4.8.

Seeds of A. viridis collected from the experimental area were surface-sterilized with 1% NaOCl for 15 min and rinsed with demineralized water. They were pre germinated for 2 days in Petri dishes on a humid filter paper at 25 ° C in the dark. The germinating seeds were used when rootlets were 1–2 cm long. For each inoculum density, plastic pots (5.5 cm diameter; 6 cm high) were filled with 100 g of soil containing the required number of AM propagules, and one pre germinated seed of A. viridis was planted per pot. Pots were arranged in a randomized complete block design with six replicates per treatment. They were placed under greenhouse conditions (30 ° C day, 20 ° C night, 10-h photoperiod) and watered daily with deionized water without adding nutrients. After 3 months of culture, the plants were harvested, and the oven-dried weight (1 week at 65 ° C) of the shoot was recorded. The root systems were gently washed, cleared and stained according to the method of Phillips & Hayman (1970). About 50 1-cm root pieces were observed per plant under a microscope (magnification, × 250). The extent of mycorrhizal colonization was expressed as (the number of mycorrhizal root pieces)/(total number of observed root pieces) × 100. The remaining roots were oven dried (1 week at 65 °C) and weighed.

Effect of soil origins on Acacia seedling growth

Seeds of Sahelian Acacia species, A. raddiana, A. senegal, A. seyal, A. nilotica and F. albida were surface sterilized with 95% concentrated sulfuric acid for 15 min (A. senegal), 30 min (A. seyal, F. albida), 60 min (A. raddiana) and 120 min (A. nilotica). They were pre germinated for 2 days in Petri dishes on a humid filter paper at 25 ° C in the dark. Soils from the same origin were pooled in the lab to obtain two types of soil. One part of the soil collected from A. viridis -invaded area or -uninvaded area was disinfected by autoclaving (120 ° C, 60 min) and the other part was not (undisinfected) to perform four treatments: (1) soil with a history of A. viridis, (2) disinfected soil with a history of A. viridis, (3) soil without a history of A. viridis and (4) disinfected soil without a history of A. viridis. For each soil treatment and each origin, plastic pots (4-cm diameter, 20-cm height) were filled with 250 g of soil and one pre germinated seed of the tested Acacia species was planted per pot. Pots were arranged in a randomized complete block design with six replicates per treatment under the same greenhouse conditions as described previously. After 5 months of culture, the plants were harvested and the oven-dried weight (1 week at 65 ° C) of the shoots was measured. Their entire root systems were washed under tap water. On each plant, the extent of AM colonization was assessed (Phillips & Hayman, 1970) and the total dry weight of root nodule per plant (1 week at 65 °C) was determined. Then the roots were oven dried (1 week at 65 °C) and weighed.

Effect of A. viridis plant extract on Acacia seedling growth

Amaranthus viridis plant extract was prepared by grinding 10 g fresh weight of the whole plant material in 100 mL of deionized water in a Waring Blender and filtering through Whatman no. 1 paper with a Buchner funnel. A 100-mL aqueous extract (or 100-mL distilled water for the control treatment) was added to plastic pots filled with 250 g of soil without a history of A. viridis. After 1 week of soil exposure to the extract, one germinating seed of each Acacia species was planted per pot with six replicates per treatment. Pots were arranged in a randomized complete block design under the same greenhouse conditions as before. After 8 months of culture, the plants were harvested, shoot and root biomasses were determined and the total dry weight of root nodule per plant (1 week at 65 ° C) and percent mycorrhizal colonization of roots were assayed.

Effect of G. intraradices inoculation on Acacia seedling growth planted in soil invaded by A. viridis or uninvaded by A. viridis

One germinating seed of each Acacia species was planted per pot filled with 250 g of the following soil treatments: (1) soil invaded by A. viridis, (2) soil invaded by A. viridis and inoculated with G. intraradices, (3) soil uninvaded by A. viridis and (4) soil uninvaded by A. viridis and inoculated with G. intraradices. The inoculum of G. intraradices was produced as described previously. The AM inoculation consisted of adding 1 g of fresh maize root (mycorrhizal, or not, for the control without fungus) to a hole (1 cm × 5 cm) made in each pot. The experimental design and the environmental conditions of the culture in the greenhouse were the same as described before. After an 8-month culture, plants were harvested, dried at 65 ° C for 1 week, weighed to determine biomass and the root nodulation was calculated. Root AM colonization was determined according to Phillips & Hayman (1970).

Plant-trapping tests

Pregerminated seeds (1–2-cm-long rootlets) of Acacia species were transferred under aseptic conditions into Gibson tubes (Gibson, 1963) containing a sterile Jensen nitrogen-free medium (Vincent, 1970). The tubes were placed in a growth chamber under controlled conditions (12 h day length at 28 °C, 40 000 lx, 75% relative humidity and 20 °C at nights). After 1 week of growth, 1 mL of the stirred soil solution was added to each tube. A soil suspension was obtained by adding 10 g of each soil sample to 90 mL of sterile saline buffer pH 7 (NaCl 0.15 M, KH2PO4 0.002 M, Na2HPO4 0.004 M) for 1 h. Soil treatments were as follows: (1) soil invaded by A. viridis, (2) soil uninvaded by A. viridis and (3) soil uninvaded by A. viridis and moistened by the extract of whole plants of A. viridis at the same rate as described before. Four replicates were made for each soil. Uninoculated plants were used as controls. Plants were observed for nodule formation for 6 weeks after inoculation, and fresh nodules were collected, numbered and oven dried.

Test of antibacterial activity

The aqueous extract of whole plants of A. viridis was tested for antibacterial activity by the diffusion technique on solid media against a seeded culture of selected strains of nitrogen-fixing bacteria (Table 1) (Vincent & Vincent, 1944). Sterilized filter paper disks (diameter 13 mm) were saturated with 0.1 mL of an A. viridis aqueous extract. One hundred milliliters of a 48-h liquid culture of rhizobia was used in seeding the Petri dishes. After inoculation, dishes were usually allowed to dry for a few hours before the test disks were added. The test plates were routinely examined and the zones of inhibition were recorded after 1 week of incubation at 25 °C in the dark.

View this table:

Effect of fresh extract of Amaranthus viridis whole plant on nitrogen-fixing bacteria

Rhizobial strainsGenusHost plantCountry of collectionEffect of fresh extract of A. viridis whole plant
ORS 1080RhizobiumA. raddianaSénégal+*
ORS 1081RhizobiumA. raddianaSénégal+
ORS 1082RhizobiumA. raddianaSénégal+
ORS 1083RhizobiumA. raddianaSénégal+
ORS 1084RhizobiumA. raddianaSénégal+
ORS 1140RhizobiumF. albidaSénégal+
ORS 1141RhizobiumF. albidaSénégal+
ORS 1142RhizobiumF. albidaSénégal+
ORS 1143RhizobiumF. albidaSénégal+
ORS 1146RhizobiumF. albidaSénégal+
ORS 1302RhizobiumA. seyalBurkina Faso+
ORS 1317RhizobiumA. seyalMauritanie+
ORS 1320RhizobiumA. seyalMauritanie+
ORS 3453MesorhizobiumA. seyalSénégal+
ORS 3454MesorhizobiumA. seyalSénégal+
ORS 3160RhizobiumA. niloticaSénégal+
ORS 3161RhizobiumA. niloticaSénégal+
ORS 3162RhizobiumA. niloticaSénégal+
ORS 3163RhizobiumA. niloticaSénégal+
ORS 3164RhizobiumA. niloticaSénégal+
ORS 3416MesorhizobiumA. senegalSénégal+
ORS 3417MesorhizobiumA. senegalSénégal+
ORS 3418MesorhizobiumA. senegalSénégal+
ORS 3419MesorhizobiumA. senegalSénégal+
ORS 3420MesorhizobiumA. senegalSénégal+
  • * The zone of bacterial growth inhibition around the filter paper disks moistened with the fresh extract of Amaranthus viridis whole plant.

Statistical analysis

All data were subjected to a one-way anova and the mean values were compared using Newman–Keul's test (P<0.05). For percentage mycorrhizal infection, data were transformed by Arc sin√x. Community similarities between TTGE profiles, based on the relative band intensity, were analyzed by principal component analysis (Wikström et al., 1999) with ade-4 software (http://pbil.univ-lyon1.fr/ADE-4/) (Thioulouse et al., 1997).


Chemical and microbial soil characteristics

Soil nitrogen, carbon, total phosphorus and soluble phosphorus contents were significantly higher in the soil collected from the A. viridis-invaded area than in the soil sampled from the uninvaded area (Table 2). The same positive effects of A. viridis were observed on the total microbial activity (FDA) and dehydrogenase activity (Table 2).

View this table:

Chemical characteristics and enzymatic activities of soils invaded or not by Amaranthus viridis plants

Uninvaded by A. viridisInvaded by A. viridis
Total nitrogen (%)0.11 (0.003)* a0.20 (0.006) b
Total carbon (%)1.21 (0.037) a2.46 (0.067) b
Total phosphorus (mg kg−1)625.3 (5.24) a1675.7 (322.48) b
Soluble phosphorus (mg kg−1)107.6 (19.57) a211.6 (5.03) b
16S rRNA gene copy number per gram of soil (× 107)4.26 a33.73 b
Total microbial activity (mg of hydrolyzed fluorescein diacetate h−1 g−1 of soil)8.94 (0.56) a18.04 (1.01) b
Dehydrogenase activity (mg INTF day−1 g−1 of soil)5.03 (0.55) a29.44 (1.96) b
  • * SE.

  • Data in the same line followed by the same letter are not significantly different according to the Newman–Keul test (P<0.05).

The total bacterial community structures from both soils were assessed by TTGE analysis of the 16S rRNA gene (Fig. 1). The TTGE gel displayed numerous bands (mean number, 14) of various intensities that resulted from differences between the 16S rRNA gene sequences of different bacterial species. The two soils shared most of the TTGE bands, suggesting that a common bacterial population colonized these soils. However, one bacterial species seemed to be more represented in soils invaded by A. viridis. Ordination plots generated by principal component analysis of band intensity data (representation of the first two axes, about 75% of the total inertia) clearly separated the soils on the basis of their origin (Fig. 2). The four replicates of soil collected from invaded sites presented a very close group in comparison with soils sampled from uninvaded sites, indicative of a strong disturbance in recipient soil bacterial populations and, therefore, their homogenization. Modifications occurring in the bacterial community structure were accompanied by shifts in the community density in soil samples. The data from the quantification of 16S rRNA gene copy number indicated significant stimulation of bacterial population in the soil invaded by A. viridis. The 16S rRNA gene copy number averaged 33.7 × 107 in soil collected from invaded sites and about 4.3 × 107 in uninvaded soil (Table 2).


16S rRNA gene-TTGE patterns of the total bacterial communities from the soils invaded or not by Amaranthus viridis plants.


Principal component analysis of the band intensity data of soils invaded or not with Amaranthus viridis.

The total number of AM spores as well as the length of external hyphae found in the uninvaded soil were significantly higher than that recorded in the invaded soil (Table 3). Two genera (Glomus and Scutellospora) and four species (Glomus constrictum, G. intraradices, Scutellospora armeniaca and Scutellospora claroideum) were present in both soil origins, but their abundance was significantly higher in the uninvaded soil (Table 3).

View this table:

Influence of Amaranthus viridis on the structure of AM fungal communities and hyphal length

Uninvaded by A. viridisInvaded by A. viridis
AM fungal species (number of AM spores per 100 g of soil)
Glomus constrictum290.2 (23.4)* b104.2 (6.1) a
Glomus intraradices16.1 (3.1) b5.2 (1.1) a
Scutellospora armeniaca31.5 (2.6) b13.7 (2.1) a
Scutellospora claroideum12.7 (1.9) b6.7 (1.7) a
Total number of AM spores per 100 g of soil350.3 (26.2) b129.7 (8.4) a
Hyphal length (m g−1 of soil)3.21 (0.32) b0.83 (0.08) a
  • * SE.

  • Data in the same line followed by the same letter are not significantly different according to the Newman–Keul test (P<0.05).

Mycorrhizae (vesicles) were observed on A. viridis root systems only at the highest inoculum rates (30 and 100 AM propagules per 100 g of soil) (Table 4). Plant growth was negatively linked with the rates of AM inoculation (r2=0.34, P=0.0028).

View this table:

Mycorrhizal dependency assessment of Amaranthus viridis plants on a disinfected soil inoculated with the AM fungus Glomus intraradices

Number of mycorrhizal root fragments per 100 g of soil
Shoot biomass (mg dry weight)322.9 b274.3 a206.1 a141.5 a
Root biomass (mg dry weight)270.2 b171.2 ab129.7 ab119.8 a
Mycorrhizal colonization (%)0 a0 a4.3 b7.5 c
  • Data in the same line followed by the same letter are not significantly different according to the Newman–Keul test (P<0.05).

Growth response of Acacia species in soils collected from uninvaded and invaded soils

The shoot and root growth, root nodule biomass and AM colonization of all the tested Acacia species were significantly decreased when plants were cultivated in the soil invaded by A. viridis than in the uninvaded soil (Table 5). In both sterilized soils invaded by A. viridis and uninvaded soils, no significant difference in the shoot and root biomasses of A. raddiana, A. senegal and F. albida was observed while these biomasses were significantly lower for other Acacia species after culturing in invaded soil (Table 5). Moreover, reductions in growth of A. raddiana, A. senegal and F. albida plants when cultured in the unsterilized soil, but invaded by A. viridis, were similar to those observed when seedlings were grown in sterilized soil from both sites invaded or not by A. viridis (Table 5).

View this table:

Growth, mycorrhizal colonization and nodule biomass of Acacia species in Amaranthus viridis-invaded and -uninvaded soils sterilized or not

Uninvaded by A. viridisInvaded by A. viridis
A. raddiana
Shoot biomass (mg dry weight)128.3 b165.1 c111.7 ab100.0 a
Root biomass (mg dry weight)88.3 ab96.7 b51.7 a44.7 a
Mycorrhizal colonization (%)20.8 b9.7 a
Nodule biomass (mg dry weight)5.5 b0.0 a
A. senegal
Shoot biomass (mg dry weight)205.0 a230.1 b201.7 a206.7 a
Root biomass (mg dry weight)210.1 a223.3 a195.0 a181.7 a
Mycorrhizal colonization (%)75.5 b33.1 a
Nodule biomass (mg dry weight)8.8 b2.3 a
F. albida
Shoot biomass (mg dry weight)391.7 a485.1 b306.7 a385.1 a
Root biomass (mg dry weight)500.1 ab580.0 b461.8 a461.9 a
Mycorrhizal colonization (%)83.7 b29.8 a
Nodule biomass (mg dry weight)17.5 b6.9 a
A. seyal
Shoot biomass (mg dry weight)1083.3 c1850.1 d816.7 a1366.7 b
Root biomass (mg dry weight)1636.7 b2066.7 c1333.3 a1633.3 b
Mycorrhizal colonization (%)38.7 b11.8 a
Nodule biomass (mg dry weight)116.6 b72.5 a
A. nilotica
Shoot biomass (mg dry weight)1953.3 b2541.7 c1513.3 a2201.7 b
Root biomass (mg dry weight)760.1 b1006.7 c611.7 a930.1 c
Mycorrhizal colonization (%)51.8 b28.3 a
Nodule biomass (mg dry weight)112.8 b69.7 a
  • Data in the same line followed by the same letter are not significantly different according to the Newman-Keul test (P<0.05).

Extracts of whole plants of A. viridis significantly decreased the shoot growth of A. senegal, F. albida, A. seyal and A. nilotica and the root growth of A. raddiana and A. nilotica (Table 6). For all the Acacia species, mycorrhizal colonization and root nodule biomass were significantly lower in the soil inoculated with A. viridis aqueous extract (Table 6).

View this table:

Growth, mycorrhizal colonization and nodule biomass of Acacia species planted in Amaranthus viridis-uninvaded soil inoculated or not (control) with fresh extract of A. viridis whole plant

Without A. viridis plant extractWith A. viridis plant extract
A. raddiana
Shoot biomass (mg dry weight)1474.0 a1232.1 a
Root biomass (mg dry weight)758.2 b586.1 a
Mycorrhizal colonization (%)67.2 b30.6 a
Nodule biomass (mg dry weight)40.0 b29.2 a
A. senegal
Shoot biomass (mg dry weight)1010.2 b764.3 a
Root biomass (mg dry weight)942.1 a824.2 a
Mycorrhizal colonization (%)91.6 b31.4 a
Nodule biomass (mg dry weight)40.1 b14.9 a
F. albida
Shoot biomass (mg dry weight)1450.2 b990.3 a
Root biomass (mg dry weight)1250.0 a1007.1 a
Mycorrhizal colonization (%)92.4 b32.1 a
Nodule biomass (mg dry weight)52.6 b21.8 a
A. seyal
Shoot biomass (mg dry weight)2640.1 b1724.3 a
Root biomass (mg dry weight)2680.2 a2206.1 a
Mycorrhizal colonization (%)69.8 b27.6 a
Nodule biomass (mg dry weight)137.9 b46.5 a
A. nilotica
Shoot biomass (mg dry weight)3674.1 b3122.0 a
Root biomass (mg dry weight)882.0 b738.2 a
Mycorrhizal colonization (%)57.4 b19.8 a
Nodule biomass (mg dry weight)130.4 b87.9 a
  • Data in the same line followed by the same letter are not significantly different according to the Newman-Keul test (P<0.05).

For both soils invaded by A. viridis and uninvaded by A. viridis, G. intraradices inoculation had significantly improved shoot and root growth of all Acacia species (Table 7), with higher growth stimulation in the soil invaded by A. viridis (Table 7). AM inoculation had also improved root nodule biomass in both soils (Table 7).

View this table:

Growth, mycorrhizal colonization and nodule biomass of Acacia species planted in Amaranthus viridis-invaded or -uninvaded soils and inoculated (with Gi) or not with Glomus intraradices (without Gi)

Uninvaded by A. viridisInvaded by A. viridis
With GiWithout GiWith GiWithout Gi
A. raddiana
Shoot biomass (mg dry weight)1255.1 c988.3 b1431.7 c720.2 a
Root biomass (mg dry weight)1130.2 c976.7 b1190.0 c670.1 a
Mycorrhizal colonization (%)73.3 d8.2 b56.1 c2.1 a
Nodule biomass (mg dry weight)40.5 c24.5 b35.6 c13.9 a
A. senegal
Shoot biomass (mg dry weight)1123.3 d638.3 b863.3 c481.7 a
Root biomass (mg dry weight)1205.0 c943.3 b1163.3 c685.1 a
Mycorrhizal colonization (%)49.2 d17.8 b28.2 c4.3 a
Nodule biomass (mg dry weight)68.1 d14.7 b29.7 c0.0 a
F. albida
Shoot biomass (mg dry weight)1293.3 c828.3 b1073.3 c570.0 a
Root biomass (mg dry weight)1346.7 c1140.0 b1518.3 c898.3 a
Mycorrhizal colonization (%)51.2 c15.8 b53.3 c5.3 a
Nodule biomass (mg dry weight)28.6 c18.4 b17.7 ab12.5 a
A. seyal
Shoot biomass (mg dry weight)2583.3 c1891.7 b2268.3 c1457.7 a
Root biomass (mg dry weight)3233.3 c2515.0 b2710.0 c1498.3 a
Mycorrhizal colonization (%)74.2 b14.2 a85.1 c10.8 a
Nodule biomass (mg dry weight)92.7 d13.5 b21.4 c0.0 a
A. nilotica
Shoot biomass (mg dry weight)3320.0 b2728.3 a2945.1 b2535.1 a
Root biomass (mg dry weight)738.3 a671.7 a760.1 a686.7 a
Mycorrhizal colonization (%)36.7 b3.2 a25.3 b1.8 a
Nodule biomass (mg dry weight)29.1 c10.1 b18.6 bc2.2 a
  • Data in the same line followed by the same letter are not significantly different according to the Newman–Keul test (P<0.05).

Nodulation assessment and antibacterial activity of A. viridis aqueous extract

The number of nodule and their total biomass per plant recorded on each Acacia species were significantly lower in the invaded soil than in the uninvaded soil (Table 8). This negative effect was found when the A. viridis aqueous extract was mixed with the uninvaded soil (Table 8). In addition, the A. viridis aqueous extract exhibited an inhibitory effect against all the rhizobial strains tested (Table 1).

View this table:

Nodulation of Acacia species in vitro conditions with soil suspensions from Amaranthus viridis-invaded or -uninvaded soils and from the uninvaded soil mixed with fresh extract of A. viridis whole plant

Acacia speciesSoil suspension origins
A. viridis- uninvaded soilA. viridis- invaded soilA. viridis- uninvaded soil +A. viridis extract
A. raddiana
Number of nodule per plant7.7 b0.0 a1.7 a
Total biomass of nodule per plant (mg dry weight)5.4 b0.0 a0.3 a
A. senegal
Number of nodule per plant5.7 b3.7 ab1.2 a
Total biomass of nodule per plant4.1 b1.6 ab0.4 a
F. albida
Number of nodule per plant50.5 c21.6 b1.6 a
Total biomass of nodule per plant20.8 b9.8 a8.2 a
A. seyal
Number of nodule per plant11.0 b0.6 a4.2 ab
Total biomass of nodule per plant18.2 b9.8 a8.2 a
A. nilotica
Number of nodule per plant2.4 a0.0 a0.4 a
Total biomass of nodule per plant3.3 b0.0 a0.1 a
  • Data in the same line followed by the same letter are not significantly different according to the Newman-Keul test (P<0.05).


Our results clearly show that the weed plant species, A. viridis, which is usually considered as a nonmycorrhizal plant, exerts a positive effect on soil nutrient content by increasing carbon, nitrogen and phosphorous concentrations and microbial activities (FDA and dehydrogenase) whereas it negatively affects the growth of the Acacia species by altering the development of rhizobia and AM communities. Moreover, AM inoculation was beneficial to the growth of Acacia species irrespective of whether the natural population of AM fungi in the soil was impoverished or not.

In our study, we observed differences between the average soil physicochemical characteristics in the IRD experimental station and those measured in the quadrant sampled for the experimental survey (with and without A. viridis). These differences might be due to the fact that in the study, soils have been sampled only in patches colonized by vegetation that is able to strongly mediate nutrient cycling and availability. By contrast, the determination of the global physicochemical characteristics of the site was assayed on soil samples representative of the site, including bulk soil patches on which nutrient cycling might be fairly reduced.

Importantly, it has been reported in many studies that plant invasions were associated with elevated or fluctuating resource levels (Davis et al., 2000; Daehler, 2003; Ehrenfeld, 2003). Invasive species grew better than native species at a higher nutrient concentration, but not at lower nutrient availability, where native vegetation is more competitive (Daehler, 2003). Studies have demonstrated that high nutrient-demanding invasive species can generate their own nutrient-rich sites, thus possibly promoting their own invasion (Vitousek et al., 1987; Ehrenfeld et al., 2001). Exotic plant species can cause an increase in soil pH as well as organic carbon content and nitrification rates, providing more available nitrate (Callaway, 1995; Kourtev et al., 1998, Kourtev et al., 1999, Kourtev et al., 2003; Ehrenfeld et al., 2001) and also phosphorus content (Ehrenfeld, 2003). The changes in the amounts and availability of nutrients may be attributed to concurrent changes in plant biomass or litter nutrient concentration with invasion (Ehrenfeld, 2003; Kao-Kniffin & Balser, 2008). Data of the present study well support this invasive plant strategy.

In addition, our results indicate that the presence of the invasive plant A. viridis has induced severe disturbances in soil microbial communities, the primary mediators of soil nutrient cycling. It has been well established that the structure and functional diversity of soil bacterial communities were mainly dependent on the aboveground plant composition (Grayston et al., 2001). Our data are in accordance with these previous studies because a higher global fertility is observed under A. viridis. It also seems that the invasive plant exerts a selective effect on some components of these bacterial communities, which may vary in abundance, and thus altering links between native aboveground and belowground communities by ways including the timing, quality, quantity and spatial structure of plant-derived soil inputs (Wolfe & Klironomos, 2005). A shift in the composition or the abundance of particular members of the microbial community has also been suggested to alter nutrient pools and fluxes (Balser & Firestone, 2005).

From the present study, we found a higher bacterial abundance and bacterial activity in the soil collected under A. viridis, but, in contrast, a decrease in AM spore abundance and hyphal length. This invasive plant effect has already been reported by Kourtev et al. (2002), who found, in soil collected from the invasive Japanese barbery, Barberis thunbergii, an overall decrease in fungal abundance and a conversion to a microbial community dominated by bacteria. Moreover, some authors have found that increased soil fertility shifted the microbial community structure, with a noticeable decrease in fungi and increase in bacteria (Pennanen et al., 1999; Bradley et al., 2006).

Because it belongs to the presumed nonmycorrhizal family Amaranthaceae (Tester et al., 1987), A. viridis has usually been considered as a nonmycorrhizal plant species. However, it has been recently reported that, under natural conditions, AM associations were found in A. viridis (Muthukumar et al., 2006). Our results corroborate these observations, but at a higher G. intraradices inoculum rate. In addition, the presence of AM structures in the roots of A. viridis was linked to a depressive effect on the plant growth. Reports of negative plant growth responses are common (Bougher et al., 1990). Mycorrhizal associations are detrimental (parasitic) to plants when the net costs are higher than net benefits (Johnson et al., 1997). Costs of mycorrhizae are traditionally expressed in terms of carbon (photosynthate) allocated to the fungus (Fitter, 1991). It has been hypothesized that AM fungal parasitism could result from (1) developmental factors, (2) environmental factors and (3) genotypic factors (Johnson et al., 1997). From the present study, the parasitic effect of G. intraradices inoculation could result from the first and the third hypothesis, but not from the ‘environmental factors’ hypothesis because the soil used in this experiment had a low fertility (in particular, low phosphorous and nitrogen contents). For instance, it has been reported that AM symbiosis can depress seedling growth at the first stages of seedling development (Koide, 1985). Because A. viridis plant growth was depressed by AM fungal inoculation and had a low affinity for AM establishment, it could be a good competitor in areas with low densities of AM fungi (Grime et al., 1987; Hartnett et al., 1993). This reduced dependence on mycorrhizal fungi can be viewed as part of a life-history strategy that is successful in disturbed environments (Francis & Read, 1995).

Our results also indicated that Acacia species nodulation was significantly reduced when these plants were grown in A. viridis-invaded soil. This negative effect of the invasive plant on the symbiotic microorganisms (AM fungi and rhizobia) could result from abiotic factors. Indeed, higher phosphorus and nitrogen contents in soil are known to inhibit the development of the fungal and rhizobial symbionts (Vincent, 1970; Smith & Read, 2008). However, A. viridis could also interfere with AM and rhizobia colonization of Acacia species root systems and slow their growth in invaded soil. These inhibitions were similar to those recorded for most of the Acacia species in sterilized soil from both A. viridis-invaded and A. viridis-free sites. This result strongly suggested that the invasive plant reduced Acacia growth through a microbially mediated effect and that this depletion did not only result from soil differences or direct allelopathy against plants. It has been demonstrated recently that an invasive plant, Alliaria petiolata (garlic mustard), suppressed native plant growth by disrupting AM symbiotic relationships through root exudation of antifungal compounds (Stinson et al., 2006). These authors showed that garlic mustard inhibited AM formation in native tree species, more particularly by reducing the germination rates of native AM spores (Stinson et al., 2006). In an earlier study, Vaughn & Berhow (1999) also observed that phytochemicals could have direct effects on plant growth through allelopathy as well as indirect effects via disruption of AM fungi. In our experimental work, the suppressive effect of A. viridis on Acacia species nodulation and root AM establishment was also found after the addition of an A. viridis plant aqueous extract to uninvaded soil. In addition, this plant extract exhibited a strong antibiotic effect against all the rhizobial strains tested (rhizobia isolated from the root systems of the targeted Acacia species). Hence, these results clearly demonstrated that A. viridis disrupts the formation of AM associations and nodulation, probably through phytochemical inhibition. Inhibition effects on nitrogen-fixing bacteria have been observed previously in pioneer plants such as Amaranthus retroflexus and suggested that this mechanism may be important in competition and succession among plants (Rice, 1964).

Hence, abundant soil mutualists' declines may initiate a positive feedback to maintain exotic plants in their introduced area (Vogelsang et al., 2004). With the successful establishment of exotic species, re-establishment of soil mutualists may be slowed, thereby impeding native plant species that exhibit high dependencies on these microorganisms. Therefore, in systems where native plants have strong mutualistic relationships with soil symbionts such as rhizobia or AM fungi, disturbances that disrupt these microbial symbiotic relationships could facilitate the establishment of exotic species having a low dependence on these microorganisms for their growth and survival.

The inoculation of G. intraradices was highly beneficial to the nodulation as well as the growth of Acacia species in both soil origins. Promoting the effect of AM inoculation on plant nodulation could probably result from close synergistic interactions between mycorrhizal fungi and rhizobia (Cornet & Diem, 1982; Founoune et al., 2002). In this regard, it has been demonstrated that mycorrhizal infection helps nodule formation and functioning under stress conditions (Azcon et al., 1988; André, 2005). Moreover, our result is in accordance with other studies, from which it has been established that this fungal isolate was very effective on the growth of plant species (Duponnois & Plenchette, 2003; Villenave et al., 2003; Duponnois et al., 2005). Importantly, it is also well established that AM fungi could affect plant community structure by enhancing the growth of the stronger mycorrhizal plant species (Gange et al., 1993; van der Heijden et al., 1998a). In the case of this study on A. viridis-invaded soil, our results provide evidence that an increase in soil AM propagule density could mediate the invasive plant antagonistic effect on Acacia species. It has been observed that microorganisms can act as allelochemical mediators, inactivating or metabolizing toxic compounds. Previous findings suggested that AM fungi, associated with their mycorrhizosphere microbial communities, could protect seedlings from allelopathy and other phytochemical compounds (Pellissier & Souto, 1999; Blum et al., 2000; Renne et al., 2004).

In conclusion, our results confirm the high importance of biological mechanism by which an invasive plant can alter native communities by disrupting the development of AM fungi and rhizobia. They also highlight the key role of AM fungi in mediating plant coexistence and the necessity to manage AM community in soil (i.e. controlled mycorrhization of Acacia tree species) to improve successful re-establishment of native species. However, further research must be undertaken to identify allelopathic processes (e.g. compounds produced by the invasive plant) and therefore to better understand the mechanisms of soil mutualistic community degradation.


  • Editor: Philippe Lemanceau


View Abstract