Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
1. The ability of exotic plant species to establish and expand in new areas may be enhanced by a relatively high ability to acquire soil nutrients. To test this hypothesis, we predicted that the capacity for nutrient acquisition would be higher in seedlings of exotic species than in seedlings of native congeners.
2. We selected the five exotic species that had recently increased in abundance in a riverine habitat in the Netherlands and that had a native congener that was common in the same habitat. We grew seedlings of each of these ten species singly in pots of soil from this habitat in a glasshouse. After two months, we measured the final dry mass and N and P content of each plant and components of microbial biomass and nutrient mineralization in the soil. We also measured these soil characteristics in pots that had been left unplanted.
3. Exotic and native congeners did not differ consistently in the uptake of N or P or in effects on components of soil mineralization. Within a genus, values of these measurements were sometimes higher, sometimes lower and sometimes similar to the exotic when compared with the native species.
4. Depending upon the statistical analysis used, biomarker-based biomass of arbuscular mycorrhizal fungi was generally higher in soil planted with exotic than with native species. Most measures of microbial biomass and soil mineralization were higher in pots that had been planted with plants than in pots with no plant.
5.Synthesis. Our results do not suggest that invasive, exotic plant species generally possess greater capacity for nutrient acquisition during the early establishment than native species do.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Over many years, humans have, both intentionally and accidentally, introduced a large number of plant species into non-native habitats across much of the biosphere (Hodkinson & Thompson 1997; Vitousek et al. 1997). Only a small number of these introduced exotic plant species are able to establish in their new region, and only a subset of those species will spread into new areas (Williamson & Fitter 1996; Richardson et al. 2000). Several exotic plant species have been shown to increase nutrient cycling after establishment (e.g. Duda et al. 2003; Chapuis-Lardy et al. 2006; Sharma & Raghubanshi 2009), especially when natives were displaced by plant species from different functional groups (Ehrenfeld 2003; Liao et al. 2008). The impact of exotic plant species on ecosystem processes is mostly studied via comparisons of invaded and non-invaded sites. This approach has the advantage in that it includes long-term changes in litter and nutrient dynamics, but this does not indicate whether altered nutrient dynamics play a role during the establishment of exotics.
Exotic plant species may be expected to exploit soil nutrients better than natives do, given that their above-ground nutrient concentrations are often higher and resource use strategies are often more efficient than those of native species present in the invaded range (e.g. Agrawal et al. 2005; Funk & Vitousek 2007). Furthermore, exotic plant species are often observed to be released from natural enemies such as pathogens and herbivores (Keane & Crawley 2002; Klironomos 2002). For example, two recent studies observed that exotic plant species that invaded a riverine habitat experienced less negative effects from soil biota (Van Grunsven et al. 2007; Engelkes et al. 2008). In addition, exotic plant species, which could be released from enemies in their new range, have been observed to allocate more of their resources to plant production instead of defence compared with the species from the native range (Blair & Wolfe 2004; Zou et al. 2006; Feng et al. 2009), possibly as a result of increased uptake of soil nutrients in the new range (Zou et al. 2006).
Here we compare nutrient acquisition strategies of exotic plant species and phylogenetically related native congeners during their establishment phase to test the hypothesis that invasive exotic plants have relatively high nutrient acquisition during early establishment. We specifically predicted that exotic plant species would show greater uptake of N and P than native congeners. We further predicted that exotics had lower amounts of available phosphorus, greater fluxes of N, P and biomass of microbes in the rhizosphere than natives. We also expected that the uptake of N would be correlated with fluxes of N in soil and that the uptake of P would be correlated with both P mineralization and AMF biomass. We tested these predictions in a glasshouse experiment using five exotic and five phylogenetically related native plant species all co-occurring in the same habitat. This approach enabled us to control for phylogenetic differences between native and exotic plant species (Harvey & Purvis 1991), as well as for habitat-specific aspects, including physical, chemical and biological soil characteristics.
Materials and methods
Five exotic plant species and their five native congeners, all members of the Asteraceae, were selected for the experiment with the following selection criteria also used by Engelkes et al. (2008): (i) exotic plant species increased in frequency in the second half of the 20th century; (ii) a native plant species within the same genus had to be present in the Netherlands; (iii) the native plant species is abundant in the Netherlands; (iv) the native and exotic plant species within a genus co-occur within the same habitat; (v) seeds that can be germinated had to be present for both native and exotic plant species; and (vi) all plant pairs had to be present in the same habitat. There were five pairs that fulfilled these criteria (Table 1). They all occur in the Geldersche Poort region (51°52′ N; 6°02′ E) along the river Waal (Dirkse, Hochstenbach & Reijerse 2007).
Table 1. Exotic and native plant species selected for the experiment. Plant species names follow Van der Meijden (2005). Bold plant names are exotic plant species
Most of the plant seeds were collected along the river Waal (Artemisia biennis, Artemisia vulgaris, Bidens tripartita, Senecio inaequidens, Senecio vulgaris and Tragopogon pratensis subsp. pratensis). Seeds of Tragopogon dubius were collected in the centre of the Netherlands (52°16′ N; 5°35′ E), whereas seeds of Bidens frondosa, Centaurea jacea and Centaurea stoebe were obtained from small seed companies that collect seeds locally in the Netherlands or Germany.
Seeds were surface-sterilized in a 0.5% sodium hypochlorite solution, germinated on glass beads, moistened with demineralized water and placed in a germination cabinet. As not all seeds germinated at the same time, seedlings were placed in a 4 °C climate chamber with 10 h light per 24 h after germination until the start of the experiment. Seedlings from B. frondosa and B. tripartita were kept in sterilized soil after germination, because these seedlings did not survive on glass beads.
Soil sampling and preparing
In September 2007, soil was collected from five locations in the nature reserve Millingerwaard (51°52′ N; 6°00′ E), which is part of the Geldersche Poort region in the Netherlands. All plant species used for our experiment occur in this region and most of them co-occur in the Millingerwaard. This nature reserve is located along the river Waal, which is the southernmost branch of the river Rhine. The Millingerwaard, as other nature reserves in this region, is grazed by horses and cattle and flooded periodically during high water discharge of the river. After sampling, the soil was homogenized, sieved through a 10-mm mesh, and stored outdoors for 2.5 weeks. The sandy soil (90.5% sand, 9.47% silt and 0.09% clay) contained 14% moisture (w/w), had a pHwater of 7.94, 1.97% C and a C:N ratio of 20.3.
Experimental design and set up
In our experimental design, we had five exotic and five phylogenetically related native plant species. There were 10 replicate pots for each plant species and 10 control pots without plants. The control pots had been set up at the same time as the pots with plants and were included to determine the effects of plant presence on soil nutrient cycling, soil nutrient pools and microbial community structure. Before planting the seedlings, 3-L pots were filled with moist field soil equivalent to 2500 g dry soil per pot. During the experiment, all soil were kept at 50% water holding capacity, which is equivalent to17.7% moisture based on dry soil. Each pot was planted with seedlings from one of the five exotic plant species or their native congeners. Three seedlings were planted per pot. We planted two individuals per pot for both native and exotic plant species within Tragopogon and Bidens pairs, because seedling availability was limited. The plants were grown in a climate-controlled glasshouse with a day temperature of 21 °C (±2 °C) and a night temperature of 16 °C (±1 °C). There was 16 h daylight in the greenhouse and artificial light was supplemented up to 225 μmol m−2 s−1 when natural daylight fell below this level. The experiment was started on 16 October 2007 and the plants were harvested on 10 December 2007. At harvest, soil, roots and shoots were separated.
Plant biomass, nutrient content and soil analyses
The harvested plant material was dried at 70 °C to determine shoot and root biomass. The dried roots and shoots were ground and analysed for total C, N and P content. Total C and N were determined in 3 mg of plant material using an NC analyser (Thermo flash EA 1112, Thermo Scientific, Milan, Italy), which uses flash combustion in conjunction with an analytical column and a thermal detector to determine percentages of N and C in the plant samples. Total P was measured following digestion with a mixture of H2SO4-Se and salicylic acid (Novozamsky et al. 1984). Briefly, digestion was performed by incubating a mixture of 300 mg plant material and 2.5 mL of a mixture of H2SO4-Se and salicylic acid for 3 h at 100 °C, 20 min at 150 °C and finally at 325 °C until liquid was transparent. Then, 48.3 mL Milli-q water was added and the mixture was filtered. Phosphorus in the filtrate was coloured with Molybdate reagens, and was measured against a reference line on a plate reader (bio-tek synergy HT-1, biotek, Winooski, USA) at 720 nm. With these data, the total nutrients in the plant were calculated for N and P separately using the following formula: Total N and P = (shoot concentration × shoot biomass) + (root concentration × root biomass).
Soil sub-samples were stored in plastic bags at 4 °C prior to analysis. Soil water content was determined as weight loss after overnight incubation at 105 °C. All soils reached constant mass after overnight incubation. This dry weight was used to calculate soil nutrient concentrations and mineralization activities.
Potential ammonification activity was determined using the arginine ammonification assay according to Alef & Kleiner (1987) with minor modifications. For each pot, two replicate samples of fresh soil (equivalent to 1 g dry weight) were spiked with 0.25 mL arginine solution (4.8 mg arginine per mL milli-q water), and one control was left unspiked. After 3 h of incubation at 25 °C, the reaction was stopped by adding 4 mL 2 M KCl. Ammonium was extracted by shaking for 15 min. Arginine was also added to control for background colour. Samples were centrifuged at 5250 × g for 10 min and dilutions (1:1 with water) were stored at −20 °C until analyses. Potential nitrifying activity was measured as accumulation of nitrite and nitrate over a 3-h period in ammonium-enriched soil slurries as described by Bodelier et al. (1998). N-NH4+ and N-NO3− concentrations were measured on a Technicon TrAAcs 800 autoanalyser (Technicon Instruments Corp., Tarrytown, NY, USA).
Alkaline phosphomonoesterase is produced by soil microorganisms. This extra-cellular enzyme releases phosphate from phosphomonoesters in soil organic matter. The phosphate released from soil organic matter by alkaline phosphomonoesterase can be taken up by plants. Enzyme activity was measured according to Schinner et al. (1995) with the following modifications: two samples and one control were centrifuged at 5250 × g for 10 min and the supernatant was diluted five times before the released p-nitrophenol was measured on a plate reader (biotek synergy HT-1, biotek, Winooski, VT, USA) at 405 nm.
Potential plant-available inorganic P was estimated via extraction of soil subsamples in 0.5 M NaHCO3 (Polsen) (Troelstra, Wagenaar & De Boer 1990). To measure Polsen, soil sub-samples were dried at a maximum of 40 °C for 1 week after harvest.
Phospholipid derived fatty acids (PLFA’s) were extracted from freeze-dried soil samples with a three-step extraction protocol according to Boschker (2004). This consisted of a Bligh and Dyer total lipid extraction, a fractioning of the total lipids on a silicic-acid column with chloroform, acetone and methanol, and finally mild-alkaline derivation to release the fatty acid methyl esters (FAMEs). Neutral lipids were collected in the chloroform fraction, and phospholipids were collected in the methanol fraction. The FAMEs were analysed using gas chromatography–flame ionization detector (GC–FID) on a Focus GC (Thermo Scientific, Bremen, Germany) with a Zebron ZB5 column (dimensions: 60 m, 0.32 mm, 0.25 μm; Phenomenex, Torrance, CA, USA). Peak areas were calculated relative to an internal standard (19:0). In total, 28 different PLFA’s were detected and identified based on retention time comparisons with FAME standards. The following PLFA’s were chosen to represent bacterial biomass: i15:0, a15:0, 15:0, i16:0, 16:1ω9, 16:1ω7t, i17:0, a17:0, 17:0, cy17:0, 18:1ω7 and cy19:0 (Frostegård & Bååth 1996); total fungal biomass: 18:2ω6 (Federle 1986); actinomycetes: 10Me16:0, 10Me17:0, 10Me18:0 (Frostegård, Tunlid & Bååth 1993). Fungal: bacterial ratios were based on 18:2ω6:bacterial PLFA ratio (Frostegård & Bååth 1996). The PLFA-based AMF biomarker 16:1ω5 was not a good indicator for arbuscular mycorrhizal fungal biomass, because this biomarker is also present in bacteria (Olsson 1999; Hedlund 2002). Therefore, the neutral lipid fatty acid (NLFA) 16:1ω5 was used as an AMF biomarker (Hedlund 2002). Ergosterol was also used to measure fungal biomass. This is a specific fungal biomarker in the cell wall of fungi, but not in AMF (Olsson et al. 2003). This was extracted from frozen soil using an alkaline-extraction method and measured on a HPLC (Dionex, Germering, Germany) equipped with a C 18 reverse-phase column (Alltech, Deerfield, IL, USA) and a UV-detector set at 282 nm (De Ridder-Duine et al. 2006).
Fixed-effect models with origin, genus and their interaction were performed using Statistica version 8 (Statsoft, Inc. (2007), Tulsa, OK, USA), following the approach of previous studies with similar experimental design (e.g. Agrawal et al. 2005). Planned comparisons were used to test for differences between species within genera. Genus was considered fixed, because plant species were selected non-randomly following the list of criteria described above, and we included all genera that met these criteria. A consequence of this approach is that the inference space for the statistical test is limited to the genera and plant species included in the experiment and does not extend to other genera that were not included in our experiment. As experimental designs similar to ours are sometimes analysed differently, we also fitted models with genus as random factor (using the type III anova implementation of PROC MIXED in sas for windows 9.1., the SAS institute, Cary, NC, USA). A drawback of this approach is that it is underpowered with only few levels (five in our study) of the random factor included in the study (Bart, Fligner & Notz 1998). Therefore, the random factor approach is presented in the Table S1 in Supporting Information, and is briefly discussed in the result section.
Data were transformed when necessary to meet the assumptions for anova. The following transformations were used: y2 for total plant N uptake; ln(y) for total plant P uptake, potential nitrifying activity and fungal biomass (PLFA biomarker); y(−1/2) for nitrate concentration in soil and biomass of actinomycetes in soil; and y(−1/4) for fungal:bacterial ratio in soil. Control pots and planted pots were analysed with planned comparisons after a one-way anova at pot level (see Tables S2 and S3). Pearson correlations between plant nutrient uptake and soil nutrient mineralization measurements were performed. For each table of results presented in the manuscript, we set a table-wide false discovery rate (FDR) threshold for significance at FDR = 0.05 to control for inflation of type I error due to multiple testing (Benjamini & Hochberg 1995; Verhoeven, Simonsen & McIntyre 2005). FDR is typically more powerful than Bonferroni approaches and has fewer false negative results (=type II errors) (Verhoeven, Simonsen & McIntyre 2005). The FDR method evaluates ranked P-values against a rank-specific significance threshold, and results in a list of significant P-values of which a pre-determined proportion are false discoveries. Thus, at FDR = 0.05, 5% of the significant results are expected to be false discoveries (=type I error).
Plant nutrient uptake
There was no consistent difference in nutrient uptake between exotic and native congeners (Table 2; Fig. 1). Therefore, the results did not support the hypothesis that exotic plant species exploit soil nutrient pools better than congeneric native plant species. There was also no significant origin effect for total C content in plant species if analysed with genus as a random factor (Table S1), and lower C content in exotic plant species was only observed within three genera (Fig. S1). However, total C content was lower in exotic than native congeners if analysed with genus as a fixed factor (Table 2 and Fig. S1).
Table 2. Comparison of genus and origin differences for nutrient acquisition, microbial activity and microbial biomass using a two-way fixed-factor anova. Bold P-values indicate that they are significant at a false discovery rate threshold of P = 0.05 (see Materials and methods)
Genus × Origin
*d.f.den is denominator degrees of freedom (d.f.). numerator d.f.: genus = 4; origin = 1; genus × origin = 4.
Total N uptake
Total P uptake
Total C in plant
Potential nitrifying activity
Fungal biomass (PLFA)
Fungal biomass (ergosterol)
Fungi : bacteria (PLFA)
Soil nutrient fluxes
No origin effects were observed for N and P fluxes if genus was considered a random factor and multiple testing correction was performed (Table S1). However, alkaline phosphomonoesterase activity was significantly higher in soil from exotic plant species than in soil from native congeners if tested with genus as a random effect without correction for multiple tests or with genus as a fixed effect (Table 2 and Fig. 2). The significant effect of origin was mainly due to differences within two genera. Origin did not affect concentration of plant-available P (Polsen; Table 2 and Fig. S2) or fluxes of N as measured by arginine ammonification and potential nitrification (see Table 2 and Fig. 2). Pots with plants had higher ammonification and lower nitrification than unplanted pots and generally had lower rates of alkaline phophomonoesterase (Table S2).
Microbial biomass in soil
Origin did not affect mass of soil microbes when genus was treated as a random factor (Table S1). However, PLFA-based fungal biomass and NLFA-based AMF biomass were on average higher in soil of exotic plant species than in that of native plant species if the data were analysed with a fixed-factor approach (Table 2, Fig. 3 and Fig. S3). Microbial biomass tended to be higher in pots with plants than in unplanted pots (Table S3).
Correlations between plant nutrient uptake and soil nutrient fluxes and AMF
Plant N uptake did not correlate with soil arginine ammonification or nitrifying activity (P > 0.05). Plant P uptake did not correlate with NLFA-based AMF biomass or plant-available P concentrations. However, alkaline phosphomomoesterase activity correlated negatively with P uptake (r = −0.48, P < 0.001).
Overall, our results do not support the hypothesis that exotic plant species are generally better able to exploit soil nutrient pools during their establishment than native plant species. Instead, in our study, rhizosphere nutrient fluxes and plant N and P uptake were higher, lower or did not differ between exotic and native plant species within the studied genera. These results are consistent with a study of effects of two exotic and one native plant species in the US on short-term N mineralization (Kourtev, Ehrenfeld & Haggblom 2003).
Plant presence increased arginine ammonification and microbial biomass, but decreased nitrification rates in comparison with control pots. This plant-induced stimulation of N-mineralizing microbes and simultaneous suppression of nitrifying bacteria points at an increased N flux from soil organic matter, whereas soil conditions remain N-limited. Under such conditions plants are often better competitors for N-NH4+ than nitrifying bacteria (Verhagen et al. 1994). However, plant N uptake did not correlate with potential soil N mineralization, possibly because the range of arginine ammonification rates between plant species was too narrow. Furthermore, plant species may have used additional strategies for N uptake to the ones we have measured here. For example, the turnover of soil microbes, e.g. via soil fauna, may have released N for plant uptake when they were grazing on rhizosphere microbes (Clarholm 1985; Ingham et al. 1985; Bardgett & Chan 1999; Bonkowski 2004).
Phosphorous uptake differed between some species within genera. However, it is not clear which processes caused this difference, because a higher P uptake did not correlate with a decrease in potential plant-available inorganic P pools (Polsen) or higher organic P mineralization as indicated by alkaline phosphomonoesterase activity. Interestingly, native plant species inhibited alkaline phosphomonoesterase more than exotic plant species did, at least when analysed with genus as a fixed effect. Alkaline phosphomonoesterase is excreted by soil microbes in situations with low P availability (Bardgett 2005). However, our soil was not P-limited, because the concentrations of potential plant-available P (Polsen) in our study is considered to be moderately high (26–45 mg P per kg soil; Sharpley & Withers 1994). Furthermore, the low N:P ratios in plant shoots (range N:P: 3.1 ± 0.3–8.0 ± 0.3; data not shown) indicated that our plant species were N-limited rather than P-limited (Koerselman & Meuleman 1996).
Exotic plant species in our study had on average more NLFA-based AMF biomass than congeneric native plant species when analysed with genus as a fixed effect, except for the genus Artemisia, where an opposite pattern was observed. In other studies, exotic plant species have been observed to increase (e.g. Walling & Zabinski 2004; Batten et al. 2006) or decrease AMF-abundance (e.g. Mummey & Rillig 2006; Vogelsang & Bever 2009). Those studies used a variety of different techniques to quantify AMF-abundance. Here we measured AMF with the NLFA-biomarker, which is present in storage lipids in external hyphae (Olsson 1999; Hedlund 2002). Therefore, exotic plant species in our study might have a higher carbon allocation to AMF hyphae than native plant species, but without a higher uptake of P. It is possible that a reduction in the dependence on AMF of the exotic plant species studied here (Seifert, Bever & Maron 2009) might have contributed to the observed results (Pringle et al. 2009) and made the AMF more parasitic to the exotic plants (Bever 2002). Interestingly, our results suggest that there was no origin effect for saprotrophic fungal biomass when using ergosterol (Fig. 3). It could be that fungal biomass in soil needs more time to respond, because a study on exotic Solidago gigantea showed that it takes more than one growing season to increase fungal biomass (Scharfy et al. 2010).
One of the key questions of this study was to determine if, as cause for their establishment, exotic plant species would exploit soil nutrients better than natives. At establishment, there was no pattern in our study indicating that exotic plant species would differ consistently in nutrient acquisition from native congeners. However, previous studies have shown that exotic plant species experience a less negative effect from soil biota than native congeners (Van Grunsven et al. 2007; Engelkes et al. 2008). Therefore, direct interactions between exotic plant species and soil biota, such as release from soil-borne enemies (Reinhart et al. 2003), could play an important role in the establishment of exotic plant species. Increased shoot N concentrations of exotics in our study (data not shown) as well as in other studies (e.g. Agrawal et al. 2005) may suggest that exotic plant species are able to accelerate nutrient cycling via higher-quality litter inputs (Cornwell et al. 2008). Therefore, altered soil nutrient cycling may be a consequence of, rather than a cause for, plant invasions (Ehrenfeld 2010).
In conclusion, our results suggest that exotic plant species do not consistently differ from native congeners in their ability to exploit soil nutrients during their early stage of establishment.
We thank Staatsbosbeheer regio Oost for giving us permission to work in Millingerwaard. We thank Wiecher Smant, Maarten Nab, Henk Duyts, Roel Wagenaar, Marco Houtekamer and Ciska Raaijmakers for technical assistance. We thank Mirka Macel, Tim Engelkes, Elly Morriën, Wil Tamis and George Kowalchuk for discussions and advice on the experiment. We thank Roy van Grunsven for help with collecting Tragopogon dubius seeds. We thank Jeff Harvey, two anonymous referees and the Handling Editor for valuable suggestions that improved the manuscript. This study was funded by the Dutch Research Council NWO-ALW with a VICI-grant to W.H.v.d.P. This is Publication 5044 of the Netherlands Institute of Ecology (NIOO-KNAW).