Effect of enhanced atmospheric CO2 on mycorrhizal ... - BES journal

0 downloads 0 Views 446KB Size Report
(Table 1). The species (all perennial) were selected to represent both different plant types (i.e. grass, legume and non-legume dicot) and different growth types.
ORIGINAL ARTICLE OA Functional Ecology 1999 13, 190–199

000

EN

Effect of enhanced atmospheric CO2 on mycorrhizal colonization and phosphorus inflow in 10 herbaceous species of contrasting growth strategies P. L. STADDON, J. D. GRAVES and A. H. FITTER Department of Biology, The University of York, PO Box 373, York YO1 5YW, UK

Summary

1. Ten herbaceous species were grown over a 4-month period under ambient (360 µmol mol–1) and elevated (610 µmol mol–1) atmospheric CO2 conditions. Plants were inoculated with the arbuscular mycorrhizal (AM) fungus Glomus mosseae and given a phosphorus (P) supply which was not immediately available to the plants. 2. Multiple harvests were taken in order to determine whether the effect of elevated CO2 on mycorrhizal colonization and phosphorus inflow was independent of its effect on plant growth. 3. All species grew faster under elevated CO2 and carbon partitioning was altered, generally in favour of the shoots. All species responded similarly to elevated CO2. 4. Elevated CO2 did not affect the percentage of root length colonized by AM fungi, but the total amount of colonized root length was increased, because the plants were bigger. 5. Elevated CO2 increased total P content, but had little or no effect on P concentration. At a given age, P inflow was stimulated by elevated CO2, but when root length was taken into account the CO2 effect disappeared. 6. In these host species there is no evidence for a direct effect of elevated CO 2 on mycorrhizal functioning, because both internal mycorrhizal colonization and P inflow are unaffected. 7. Future research should concentrate on the potential for carbon flow to the soil via the external mycelial network. Key-words: Arbuscular mycorrhizas, carbon partitioning, elevated CO2, Glomus mosseae, soil carbon storage Functional Ecology (1999) 13, 190–199

Introduction

© 1999 British Ecological Society

We are witnessing a steady increase in atmospheric CO2 concentration resulting from human activities (Houghton, Jenkins & Ephraums 1990; Wyman 1991). However, not all the CO2 released by human activity stays in the atmosphere and it is estimated that between 1 and 2 Gt C year–1 is being sequestered into an unknown (or ‘missing’) sink (Watson et al. 1992). Many plants show increased photosynthesis and growth under elevated CO2 conditions (e.g. twice current ambient level) (Poorter 1993; Rogers, Runion & Krupa 1994). Many plants also show changes in their C allocation pattern, generally in favour of the roots (Stulen & den Hertog 1993; Rogers et al. 1994, 1996). It has been hypothesized that these plant responses to elevated CO2 could lead to increased C storage in soils, which would explain the missing sink (Gifford 1994). Carbon flow to the soil can be directly via the plant roots or indirectly via other soil biota including mycorrhizas. Carbon allocation to

mycorrhizal fungi can be relatively high, in the order of 10% of photosynthetically fixed C (Harris & Paul 1987; Tinker, Durall & Jones 1994). More C being fixed by plants means that more C should be available for mycorrhizal symbiosis. If mycorrhizal fungi are stimulated by this extra C (see Tester et al. 1986), then they could be a major route by which increased C is entering the soil. The effect of elevated CO2 has been studied in both ectomycorrhizal (ECM) and arbuscular mycorrhizal (AM) systems. ECM colonization is often stimulated by elevated CO2 (Norby, O’Neill & Luxmoore 1986; Norby et al. 1987; O’Neill, Luxmoore & Norby 1987). The evidence for AMs is however, less clear. Monz et al. (1994) and Morgan et al. (1994) found increases in root length colonized (RLC) in Bouteloua gracilis, and Klironomos, Rillig & Allen (1996) found the same in Artemisia tridentata, Jongen, Fay & Jones (1996) found no difference in RLC for Trifolium repens. A major problem with the above studies is that they did not carry out their investigations over a time 190

191 CO2, mycorrhizal colonization and P inflow

series and that they were therefore comparing plants of different developmental stages and sizes. A previous seven-harvest experiment investigating Glomus mosseae colonization of Plantago lanceolata and T. repens at ambient and elevated CO2 concentrations, showed that there was no direct effect of elevated CO2 on mycorrhizal colonization and functioning in relation to uptake of phosphorus (Staddon et al. 1998). The experiment reported in this paper was designed to test that the above conclusion was not restricted to P. lanceolata and T. repens, but generally applicable to a much wider range of species of different growth types, both dicots and monocots and legumes and non-legumes.

Materials and methods PLANT MATERIAL

Most seeds were obtained from commercial seed suppliers with the exception of Chamerion angustifolium L. seeds which had to be collected from the wild (Table 1). The species (all perennial) were selected to represent both different plant types (i.e. grass, legume and non-legume dicot) and different growth types [based on Grime, Hodgson & Hunt (1988) and the simplified groupings of Hunt et al. (1993)] (Table 1). All of the species can be colonized by arbuscular mycorrhizas (Harley & Harley 1987) and most are known to be responsive to elevated CO2 (Hunt et al. 1991, 1993; Poorter 1993) with the exception of Lotus corniculatus. EXPERIMENTAL SET-UP

After germination on moist filter paper (Whatman No. 1) seeds were planted out into pots (11·5 cm deep, 12·5 cm diameter) containing a 1:1 volumetric mix of coarse builder’s sand and Terragreen (an attapulgite clay soil conditioner, Turfpro Ltd, Staines, UK). This mix was amended with a supply of phosphorus in the form of bonemeal at 0·20 g litre–1. The mycorrhizal

Table 1. Plant material description. Species are listed along with their seed source, their plant type and their group type. The commercial seed suppliers are Johnsons Seeds (Boston, Lincolnshire, UK) and Emorsgate Seeds (Norfolk, UK). Pretty Wood is a site in North Yorkshire, UK. NLD, non-legume dicot; SNC, slow growing and non-competitive; FC, fast growing and competitive; IC, intermediate type Species

Seed source

Plant type

Growth type

Festuca ovina L. Holcus lanatus L. Plantago lanceolata L. Prunella vulgaris L. Digitalis purpurea L. Taraxacum officinale L. Chamerion angustifolium L. Trifolium repens L. Lotus corniculatus L. Medicago sativa L.

Johnsons Emorsgate Emorsgate Johnsons Emorsgate Johnsons Pretty Wood Emorsgate Emorsgate Johnsons

Grass Grass NLD NLD NLD NLD NLD Legume Legume Legume

SNC FC IC IC IC FC FC IC SNC IC

fungal inoculum (provided by J. Merryweather, University of York, York, UK) consisted of maize-root medium infected with G. mosseae (isolate UY21, from original Rothamsted isolate), including the sandTerragreen growth medium. The inoculum was premixed into the experimental growth medium at 1:10 and delivered into the pots in a horizontal band 1 cm thick, at 2–3 cm depth. To guarantee having nodulated T. repens and L. corniculatus, Rhizobium trifolii inoculum (provided by D. Davies, University of York, York, UK) grown on agar slopes was delivered in deionized water. Nodulation of Medicago sativa was unsuccessful. The plants were watered daily with deionized water, and after 2 weeks they were fed every 2 days with half strength Long Ashton nitratebased nutrient solution amended to contain no phosphorus (Hewitt & Smith 1975). Plants were grown in two glasshouse chambers with different atmospheric CO2 concentrations, ambient (360 ± 40 µmol mol–1) and elevated (620 ± 40 µmol mol–1). The CO2 enrichment system consisted of a canister of compressed CO2 (BOC Ltd, UK) linked to an infra-red gas analyser (WA524, The Analytical Development Co. Ltd, UK) which constantly monitored the CO2 concentration and adjusted the CO2 released via a control valve. The ambient CO2 concentration was measured with a portable gas analyser (ADC Ltd, UK). So as to eliminate any potential chamber effect, the chambers were switched every week (i.e. the CO2 treatments and the plants were switched between the two chambers). Strictly speaking, there is still the problem of pseudoreplication, however, we are satisfied that any chamber effect was rendered negligible by the chamber switching and thus the only difference in environmental conditions experienced by the two sets of plants was the CO2 concentration (see also Norby 1987). The plants were grown under natural light supplemented with light from a 400 W halogen bulb (giving a minimum PAR of 100 µmol m–2 s–1 at plant level, sensor SKP 210, Skye Instruments Ltd, UK) with a 16 h light period. Plants were randomized weekly to avoid the problem of light patchiness. Temperature control was set at 20/15 °C day/night, with maximum day temperature ranging beteen 17 °C and 26 °C and minimum night temperature between 10 °C and 17 °C. Plants were grown over a 4 month period with three sequential harvests each consisting of three replicates for both CO2 treatments. Holcus lanatus harvests occurred at 49, 70, 105 days after planting (dap), Taraxacum officinale and M. sativa at 54, 75, 111 dap, P. lanceolata and Festuca ovina at 59, 80, 115 dap, Prunella vulgaris and T. repens at 63, 84, 119 dap, Digitalis purpurea at 65, 86, 121 dap, L. corniculatus at 66, 87, 122 dap and C. angustifolium at 59, 95, 119 dap. For all species the time between the first and last harvest was 8 weeks except for C. angustifolium where it was 8·5 weeks. At harvest, shoots were cut off at ground level and roots were carefully washed in water to free them of sand and Terragreen particles. Root

192 P. L. Staddon et al.

lengths were measured using the grid line intercept method (Tennant 1975). Subsamples from the top 5 cm of the root system (defined by distance from the stem base) were removed for mycorrhizal colonization assessment. Shoot, root and crown dry masses (DM) were obtained after oven drying at 80 °C for 48 h. Leaf area was measured with a leaf area meter (Lambda Instuments Corporation, USA). The random root samples (within the top part of the root) for mycorrhizal assessment consisted of segments about 1 cm long. The roots were cleared in KOH (80 °C, 2 min), acidified in HCl (room temperature, 1 min) and stained with acid fuchsin (80 °C, 20 min) (cf. Kormanik & McGraw 1982) but omitting phenol. Mycorrhizal colonization was assessed under a compound microscope at × 250 magnification using epifluorescence (Merryweather & Fitter 1991). Scoring was carried out with the line intercept method (McGonigle et al. 1990) looking at a minimum of 60 intersections (although for the first harvest this was not always possible owing to small sample size). The percentage of intercepts where hyphae are present is termed root length colonized (RLC). Phosphorus assay of plant seed and shoot and root tissue (for harvests 2 and 3) was carried out after triple acid digestion using the molybdenum blue method (Allen 1974). Phosphorus (P) tissue concentrations, total seed, plant, shoot and root P contents were obtained. P inflow was then calculated for two periods: period 1 from the start of the experiment to harvest 2, and period 2 from harvest 2 to harvest 3. P inflow for period 1 was based on the difference in P content between individual plants at harvest 2 and the mean seed value and likewise P inflow for period 2 was based on the difference in P content between individual plants at harvest 3 and the mean value at harvest 2.

ANCOVAs. P inflows for the two periods were analysed separately and for both periods together using two-way and three-way ANOVAs and ANCOVAs, respectively.

Results EFFECT OF ELEVATED ATMOSPHERIC CO 2 ON PLANT GROWTH

Over the three harvests, all species showed increased growth (i.e. they grew faster, resulting in bigger plants at a given time) under elevated CO2 conditions although not all by the same extent (Fig. 1). There was, however, no significant interaction between species and treatment (for any of the plant parameters analysed by two-way ANCOVAs with time as the covariate) which means that all 10 species were affected in a similar way by the elevated CO2 treatment. Both shoot and root growth were stimulated by elevated CO2 (Table 2). Analysis of covariance, using age as covariate, showed that, at a given time, plants under elevated CO2 had significantly greater biomass, root length and leaf area (Table 2). There was no overall significant change (P = 0·088) in the allometry of growth: root dry mass was unaffected by CO2 treatment when variation in shoot dry mass was accounted for (Table 2), although at elevated CO2 there was nonetheless a trend of less carbon being allocated to the root. Specific root length (SRL) analysed as a function of total plant biomass showed a decrease in value (trend only) at elevated CO2. The analysis of root length per unit plant dry mass showed a significant (P = 0·033) decrease at elevated CO2 (Table 2), and when this parameter was analysed for the last two harvests only, the decrease became even more significant (three-way ANCOVA, P = 0·001) (Fig. 2).

STATISTICAL METHODS

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

Statistical analysis was performed on SPSS v6 (Norusis 1994). Where necessary, data were natural log (ln), square root or arcsine square root transformed to conform to normality. This was tested for by the non-parametric 1-sample Kolmogorov-Smirnov test. Plant growth (various plant parameters) was analysed using two-way ANCOVAs with species and treatment (CO2) as factors and days after planting (dap) as the covariate. Carbon partitioning between plant parts (root and shoot) was analysed using allometric relationhips. Specific root length (SRL) was analysed by two-way ANCOVA (species and treatment as factors) with plant DM as the covariate. Because plant growth was stimulated by elevated CO2 treatment (see Results), covariates had to be included in many of the analyses in order to allow comparisons between plants at similar stages of growth. Root length per unit plant dry mass (DM) was analysed by three-way ANCOVA (species, treatment and harvest as factors). Phosphorus (P) concentration [P] and tissue P content were analysed using three-way (as above) ANOVAs and

EFFECT OF ELEVATED ATMOSPHERIC CO 2 ON MYCORRHIZAL ROOT LENGTH

Mycorrhizal colonization generally increased with time over the successive harvests (e.g. under ambient CO2, the mean percentage RLC for all species ranged from 26% for F. ovina to 70% for P. vulgaris at harvest 1, and from 42% for H. lanatus to 84% for P. vulgaris at harvest 3, and a similar pattern could be seen under elevated CO2). The only significant (P = 0·001) effect of CO2 on mycorrhizal colonization in the range of species was a stimulation of total amount of root length colonized at a given time (Table 3). This difference disappeared when root length was taken into account. There was no CO2 effect on percentage RLC (Table 3). The analysis showed a highly significant species effect (P < 0·001), but no species by CO2 interaction (see caption to Table 3). The analysis was also performed with the various species grouped by plant type (grass, non-legume dicot, legume; M. sativa was treated separately as it was not nodulated) or by group type (fast competitive, intermediate competitive and slow non-

193 CO2, mycorrhizal colonization and P inflow

Fig. 1. Effect of CO2 treatment on plant growth in the 10 species studied. The lines shown are the fitted linear regression lines of total plant dry mass (ln transformed) regressed against days after planting (dap). Full lines and squares, elevated CO2; dashed lines and diamonds, ambient CO2.

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

competitive). These groupings did not significantly change the above results on the effect of elevated CO2 on RLC; in both cases, the groups differed significantly and neither showed any interaction with the CO2 effect (Table 3 caption). These results showed that there was no difference in mycorrhizal response to elevated CO2 in the various host plant species, and that the CO2 effect on mycorrhizal colonization seen on an

age basis was simply owing to plants growing faster (see Fig. 1) under elevated CO2 conditions. EFFECT OF ELEVATED ATMOSPHERIC CO 2 ON PLANT TISSUE PHOSPHORUS

For all species, plant shoot P concentration [P] was not affected by CO2 treatment (Fig. 3a, Table 4). Root

Table 2 . Effect of elevated atmospheric CO2 on various plant growth parameters. The significance values given for the CO2 treatment effect were obtained from twoway ANCOVAs. In all cases species (10 levels) was the second factor and was highly significant (P < 0·001), as was the covariate (P < 0·001). There was never a significant interaction effect between the two factors. The direction of the elevated CO2 effect was obtained visually from scatter plots fitted with regression lines. NS, not significant; * P < 0·05, *** P < 0·001 Elevated CO2 treatment effect

Total plant DM Shoot DM Root DM Root length Leaf area (LA) Root DM Root DM Specific root length (SRL) Root length per unit plant DM

Covariate

P-value

Direction

age age age age age shoot DM total plant DM total plant DM age

*** *** *** *** *** NS (0·088) NS (0·135) NS (0·088) * (0·033)

increase increase increase increase increase decrease decrease decrease decrease

[P] was weakly (P = 0·040) affected and there was the complication of a significant three-way interaction. When plant size (total plant DM) was taken into account there was no CO2 effect on [P] (Table 4). Total plant, shoot and root P content was signifi-

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

cantly increased by elevated CO 2 at the two harvests analysed (harvests 2 and 3) (Table 4, Fig. 3b,c). This was not a direct CO2 effect as the significant effect disappeared when plant size was taken into account. In all cases there was a highly significant species effect and in most cases a significant harvest effect and species by harvest interaction (Table 4). Interactions involving treatment were (with one exception) all weak (P < 0·05), but are nonetheless problematic and will be discussed later. EFFECT OF ELEVATED ATMOSPHERIC CO 2 ON PHOSPHORUS INFLOW

There were marked differences between species in P inflow, although in all species, P inflow (i.e. uptake rate per unit root length) was several times higher during period 1 than during period 2 (Fig. 4). For period 1, P inflow was increased by elevated CO2 in nine of the 10 species (Fig. 4a), although this was not strictly significant (ANOVA P = 0·065). When P inflow was analysed as a function of percentage RLC, the CO2 effect was significant (ANCOVA P = 0·011), but when analysed as a function of total plant DM, root length or total colonized root length the CO2 effect disappeared. For period 2, P inflow was also increased by elevated CO2

Fig. 2. Effect of CO2 treatment on the ratio of root length (ln transformed) to plant dry mass (ln transformed) in the 10 species studied for (a) harvest 2 and (b) harvest 3. Hl, Holcus lanatus; To, Taraxacum officinale; Ms, Medicago sativa; Pl, Plantago lanceolata; Fo, Festuca ovina; Pv, Prunella vulgaris; Tr, Trifolium repens; Dp, Digitalis purpurea; Lc, Lotus corniculatus; Ca, Chamerion angustifolium. ■ ambient CO2, ■ elevated CO2. Error bars represent SE.

Table 3. Significance levels of CO2 treatment effect on Glomus mosseae colonization. Significance levels were obtained from two-way ANCOVAs. In all cases the second factor [species, plant type (grass, legume, non-legume dicot) and growth type (fast competitive, intermediate competitive, slow non-competitive)] was highly significant (P < 0·001). The interaction term between the two factors was not significant in any case. dap, days after planting. Root length colonized (RLC) is expressed as a percentage (%) and as the total amount of root length colonized per plant (total). NS, not significant; * P < 0·05, ** P < 0·01 Variate:

% RLC

Total RLC

% RLC

Total RLC

Covariate:

vs dap

species

NS (0·52)

** (0·001)

NS (0·11)

NS (0·06)

plant type

NS (0·57)

** (0·020)

NS (0·21)

NS (0·14)

growth type

NS (0·65)

** (0·025)

NS (0·51)

NS (0·40)

vs root length

Second Factor

(ANOVA P = 0·022) in eight out of nine species (T. officinale could not included owing to missing values) (Fig. 4b). RLC parameters could not be used as covariates in this case as they were not significant. When P

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

inflow was analysed as a function of total plant DM, the CO2 effect was significant (ANCOVA P = 0·023), but when analysed as a function of root length the CO2 effect disappeared. For all the above analyses there was no species by treatment interaction. Phosphorus inflow was also analysed for both periods simultaneously (Table 5). In all cases there were significant differences between species (P < 0·001) and between periods (P < 0·001) and significant species by harvest interactions. Elevated CO2 significantly increased P inflow at a given time. Percentage RLC (which is not affected by elevated CO2, Table 3) actually increased the significance of the CO2 effect on P inflow (Table 5). When root length or total colonized root length were taken into account the CO2 effect on P inflow disappeared, but this was not the case when plant DM was used as the covariate (Table 5). Also, when plant DM was used as the covariate, there were significant interactions between (1) species and treatment and (2) between harvest and treatment. Interaction (1) is due to the different species having

Fig. 3. Effect of CO2 treatment on (a) shoot phosphorus concentration (mg g–1) at harvest 3 (b) total plant phosphorous content (µg P, ln transformed) at harvest 2, and (c) total plant phosphorus content (µg P, ln transformed) at harvest 3. In all three cases data shown for all 10 species: Hl, Holcus lanatus; To, Taraxacum officinale; Ms, Medicago sativa; Pl, Plantago lanceolata; Fo, Festuca ovina; Pv, Prunella vulgaris; Tr, Trifolium repens; Dp, Digitalis purpurea; Lc, Lotus corniculatus; Ca, Chamerion angustifolium. ■ ambient CO2, ■ elevated CO2. Error bars represent SE (absent error bars due to missing data points). NA, not available.

different slopes for the relationship between P inflow and plant DM. Interaction (2) is due to different slopes for the above relationship in the two periods.

196 P. L. Staddon et al.

Discussion Plants under elevated CO2 grew significantly faster than those under ambient CO2 (Fig. 1), which is consisTable 4. Effect of elevated atmospheric CO2 on root and shoot phosphorus concentration [P] and on root, shoot and plant total phosphorus content. Results (P-values) from three-way ANOVAs and ANCOVAs are given. In all cases, the covariate (plant DM) is significant. Significant interactions (P < 0·05) are noted and are as follows: s.h, species by harvest; s.t, species by treatment; h.t, harvest by treatment; s.h.t, species by harvest by treatment. All data ln transformed to satisfy conditions of normality. NS, not significant; * P < 0·05, *** P < 0·001 Factors

Covariate

Species

Harvest

CO2 treatment

Significant interactions

Root [P]

none plant DM

*** ***

* NS

* NS

s.h, s.h.t s.h, s.h.t

Shoot [P]

none plant DM

*** ***

*** ***

NS NS

s.h s.h, s.t

Total root P

none plant DM

*** ***

*** *

*** NS

s.h, h.t, s.h.t s.h, h.t, s.h.t

Total shoot P

none plant DM

*** ***

*** ***

*** NS

none s.h, s.t

Total plant P

none plant DM

*** ***

*** *

*** NS

s.h s.h

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

tent with previous work on many of these species (Hunt et al. 1991, 1993; Poorter 1993). Under elevated CO2 there was a general increase in partitioning towards the shoot (Table 2), and although there was no species by CO2 interaction, some species (e.g. C. angustifolium, D. purpurea, F. ovina and L. corniculatus) showed this more than others (e.g. H. lanatus and P. lanceolata). Although elevated CO2 generally has the opposite effect over a very wide range of species a proportion of species do show the effect we observed (Rogers et al. 1996). Of course various factors could account for this, not just the species in question but also the growth conditions (e.g. growth medium, water and nutrient availability, temperature and lighting). Mycorrhizal colonization was assessed in the equivalent top section of the root system in all species. This was done because (1) the mycorrhizal inoculum was delivered near the top of the pots (a commonly used method: e.g. Jongen et al. 1996) and (2) roots were reaching the bottom of the pots at different times for the various species. There were significant differences between species, between plant types and between growth types in relation to amount of mycorrhizal colonization. For all species, mycorrhizal colonization responded in a similar way to elevated CO2 (Table 3). The analysis of mycorrhizal colonization showed that elevated CO2 only stimulated total mycorrhizal colonized root length for plants of a given age. When root length was taken into account, the CO2 treatment effect on mycorrhizal colonization disappeared (Table 3). This means that for all these plant

Fig. 4. Effect of CO2 treatment on P inflow (pmol m–1 s–1) in the 10 species studied during (a) period 1 and (b) period 2 (owing to missing data, Taraxacum officinale was not included in this period). The 10 species were Hl, Holcus lanatus; To, Taraxacum officinale; Ms, Medicago sativa; Pl, Plantago lanceolata; Fo, Festuca ovina; Pv, Prunella vulgaris; Tr, Trifolium repens; Dp, Digitalis purpurea; Lc, Lotus corniculatus; Ca, Chamerion angustifolium. ■ ambient CO2, ■ elevated CO2. Error bars represent SE. Note the change in scale.

197 CO2, mycorrhizal colonization and P inflow

species there was no direct stimulation of mycorrhizal colonization by elevated CO2, and that any increase on an age basis is simply owing to faster growth of the host plants. This observation agrees with the conclusion of our previous experiment (Staddon et al. 1998) and is consistent with the host plant being able to regulate internal mycorrhizal colonization quite tightly and not allowing the fungal partner to utilize all the available soluble carbon in the root. Lewis, Thomas & Strain (1994) reached a similar conclusion on ectomycorrhizal colonization at elevated CO2. Much of the previous work on the effects of elevated CO2 on mycorrhizal colonization has not taken plant developmental stage into account (e.g. Monz et al. 1994; Morgan et al. 1994; Jongen et al. 1996), a crucial factor especially as plant development can be altered by both elevated CO2 (Poorter 1993) and mycorrhizal colonization (Hetrick 1991). Although in this experiment we only had three harvests, this was still sufficient with the degree of replication (both on an individual and a species basis) to take altered plant development into account. Allometric analysis to test for changes in allocation pattern between plant parts is a powerful and necessary tool in such experiments (Farrar & Williams 1991). Internal mycorrhizal colonization, as a function of root length available, followed the same pattern in roots of plants grown in ambient and elevated CO2 (Table 3). Similarly, at a given age, elevated CO2 significantly increased total tissue P content in both root and shoot, but had little effect on root P concentration ([P]) and no effect on shoot [P] (Table 4). The interactions involving the CO2 treatment (see Table 4) must be briefly explained here. The species by treatment interactions (P < 0·05) seen for shoot [P] and total shoot P as a function of plant DM were because of the different species having different slopes (in relation to the CO2 treatment) for the relationship with plant DM. The harvest by treatment interaction (P < 0·05) for total root P without the covariate was because of the CO2 effect being non-significant at harvest 2. The same interaction for total root P with plant DM as the

Table 5. Effect of elevated atmospheric CO2 on phosphorus inflow. Results are from three-way ANOVA and ANCOVAs. In all cases the two other factors were species and harvest, which were always highly significant (P < 0·001). In all cases covariates were significant. Significant interactions are noted and are as follows: s.h, species by harvest; s.t, species by treatment; h.t, harvest by treatment. RLC, root length colonized expressed as a percentage (%) and as the total amount per plant. P inflow square root transformed,% RLC arcsine square root transformed, others variables ln transformed. NS, not significant; * P < 0·05, ** P < 0·01

With no covariate With percentage RLC as covariate With plant DM as covariate With root length as covariate With total RLC as covariate

Significance of CO2 treatment

Significant interactions

** (0·010) ** (0·002) ** (0·018) NS (0·521) NS (0·328)

s.h s.h s.h, s.t, h.t s.h s.h

covariate was because of the two harvests showing different slopes (in relation to CO2 treatment) for the relationship with plant DM. The significant three-way interactions for root [P] and total root P with and without the covariate were because of several of the species (especially L. corniculatus and M. sativa) showing a different trend in the CO2 effect at the two harvests. It can nonetheless be concluded that elevated CO2 had little or no effect on tissue [P] but significantly increased total plant P (at a given age). Although they only had a single harvest, it appears that Jongen et al. (1996) found a similar effect in their mycorrhizal T. repens. To conclude, increased P content was simply a result of bigger plants under elevated CO2 conditions, which is what is generally reported (Conroy 1992; Owensby, Coyne & Auen 1993) with a few exceptions (e.g. Morgan et al. 1994). Phosphorus inflows were calculated over two periods (period 1: start of experiment to harvest 2; period 2: harvest 2 to harvest 3). The much lower P inflows for period 2 were probably due to a combination of factors including greater total root length per plant and less P being available in the growth medium (all P was delivered to the medium at the start of the experiment). When analysed separately, P inflow over the two interharvest periods showed similar CO2 treatment effects as when both periods were analysed simultaneously (Table 5). In all cases there were highly significant species and harvest effects and species by harvest interaction. Therefore, the different species showed different P inflows and also a difference in the decrease in P inflow between the two periods. At a given age, P inflow was stimulated (P = 0·010) by elevated CO2, an effect also observed in 104 day-old P. lanceolata by Rouhier & Read (1998), who did not however, determine whether the effect was a result of the host plant showing altered growth and partitioning under elevated CO2. Percentage RLC (as a covariate) did not change this, as it was not itself affected by CO2 treatment. At a given plant DM, P inflow was still stimulated by elevated CO2 (P = 0·018), and there were significant interactions between CO2 and species (different species having different slopes in relation to CO2 treatment as a function of the covariate) and CO2 and harvest (different slopes for the relationship for the two periods). When root length or total root length colonized were taken into account, the CO2 effect on P inflow disappeared. The fact that using plant DM as the covariate did not eliminate the CO2 effect was owing to the difference in plant carbon allocation seen between the two CO2 treatments (Fig. 2 and Table 2), and can be explained as follows: (1) there was a general negative relationship between P inflow and plant size, (2) plants of a given plant DM at ambient and elevated CO2 had different root lengths (less under elevated) and (3) therefore, for a given plant DM, greater P inflow at elevated CO2 resulted. From our analyses, it can be concluded that elevated CO2

198 P. L. Staddon et al.

had no direct effect on P inflow and that any effect seen on an age basis was simply because of faster growing plants with altered C partitioning. Therefore, in these host plant species, there is no evidence for a direct effect of elevated CO2 on mycorrhizal functioning as internal mycorrhizal colonization and the mycorrhizal P uptake mechanism are unaffected. Following on from our previous experiment (looking at the effects of elevated CO2 on mycorrhizal colonization by G. mosseae in P. lanceolata and T. repens over seven harvests) (Staddon et al. 1998), this experiment clearly demonstrates that internal mycorrhizal colonization (in the case of G. mosseae) is unresponsive to elevated CO2 treatment in a wide range of species of different growth types, both dicots and monocots and legumes and non-legumes. An outstanding question, however, is whether the external mycorrhizal network responds to an elevated CO 2 environment, as a consequence of increased plant growth. Also, we acknowledge that the G. mosseae isolate used in these experiments is not necessarily representative of the numerous other mycorrhizal fungus species present in a natural ecosystem and that further research focused on the effects of elevated CO 2 on mycorrhizal colonization and external hyphal production by the different members of the mycorrhizal community is urgently needed. Only then can we begin to understand how mycorrhizas integrate with increased carbon flow to the soil under elevated atmospheric CO2 conditions (see Fitter et al. 1996) and how this fits in with the various possible feedback mechanisms (Dixon & Turner 1991) related to the global C cycle.

Acknowledgements We thank the Department of Biology at the University of York for awarding P.L.S. a studentship. We also wish to thank the horticultural staff and in particular Colin Abbott for the care given in looking after the plants.

References

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

Allen, S.E. (1974) Chemical Analysis of Ecological Materials. Blackwell Scientific Publications, Oxford. Conroy, J.P. (1992) Influence of elevated atmospheric CO2 concentrations on plant nutrition. Australian Journal of Botany 40, 445–456. Dixon, R.K. & Turner, D.P. (1991) The global carbon cycle and climate change: responses and feedbacks from belowground systems. Environmental Pollution 73, 245–262. Farrar, J.F. & Williams, M.L. (1991) The effects of increased atmospheric carbon dioxide and temperature on carbon partitioning, source-sink relations and respiration. Plant, Cell and Environment 14, 819–830. Fitter, A.H., Self, G.K., Wolfenden, J., van Vuuren, M.M.I., Brown, T.K., Williamson, L., Graves, J.D. & Robinson, D. (1996) Root production and mortality under elevated atmospheric carbon dioxide. Plant and Soil 187, 299–306. Gifford, R.M. (1994) The global carbon cycle: a viewpoint on the missing sink. Australian Journal of Plant Physiology 21, 1–15.

Grime, J.P., Hodgson, J.G. & Hunt, R. (1988) Comparative Plant Ecology: a Functional Approach to Common British Species. Unwin Hyman, London. Harley, J.L. & Harley, E.L. (1987) A check-list of mycorrhiza in the British flora. New Phytologist Suppl. 105, 1–102. Harris, D. & Paul, E.A. (1987) Carbon requirements of vesicular–arbuscular mycorrhizae. Ecophysiology of VA Mycorrhizal Plants (Ed. G. R. Safir), pp. 93–105. CRC Press, Boca Raton, FL. Hetrick, B.A.D. (1991) Mycorrhizas and root architecture. Experientia 47, 355–362. Hewitt, E.J. & Smith, T.A. (1975) Plant Mineral Nutrition. The English Universities Press, London. Houghton, J.T., Jenkins, G.J. & Ephraums, J.J. (1990) Climate Change: the IPCC Scientific Assessment. Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Hunt, R., Hand, D.W., Hannah, M.A. & Neal, A.M. (1991) Response to CO2 enrichment in 27 herbaceous species. Functional Ecology 5, 410–421. Hunt, R., Hand, D.W., Hannah, M.A. & Neal, A.M. (1993) Further responses to CO2 enrichment in British herbaceous species. Functional Ecology 7, 661–668. Jongen, M., Fay, P. & Jones, M.B. (1996) Effects of elevated carbon dioxide and arbuscular mycorrhizal infection on Trifolium repens. New Phytologist 132, 413–423. Klironomos, J.N., Rillig, M.C. & Allen, M.F. (1996) Belowground microbial and microfaunal responses to Artemisia tridentata grown under elevated atmospheric CO2. Functional Ecology 10, 527–534. Kormanik, P.P. & McGraw, A.-C. (1982) Quantification of vesicular-arbuscular mycorrhizae in plant roots. Methods and Principles of Mycorrhizal Research (Ed. N. C. Schenck), pp. 37–46. American Phytopathological Society, St Paul, Minnesota. Lewis, J.D., Thomas, R.B. & Strain, B.R. (1994) Effect of elevated CO2 on mycorrhizal colonisation of loblolly pine (Pinus taeda L.) seedlings. Plant and Soil 165, 81–88. McGonigle, T.P., Miller, M.H., Evans, D.G., Rairchild, G.L. & Swan, J.A. (1990) A new method which gives an objective measure of colonization of roots by vesicular– arbuscular mycorrhizal fungi. New Phytologist 15, 495–501. Merryweather, J.W. & Fitter, A.H. (1991) A modified method for elucidating the structure of the fungal partner in vesicular–arbuscular mycorrhiza. Mycological Research 95, 1435–1437. Monz, C.A., Hunt, H.W., Reeves, F.B. & Elliott, E.T. (1994) The response of mycorrhizal colonization to elevated CO2 and climate change in Pascopyron smithii and Bouteloua gracilis. Plant and Soil 165, 75–80. Morgan, J.A., Knight, W.G., Dudley, L.M. & Hunt, H.W. (1994) Enhanced root system C-sink activity, water relations and aspects of nutrient acquisition in mycotrophic Bouteloua gracilis subjected to CO2 enrichment. Plant and Soil 165, 139–146. Norby, R.J. (1987) Nodulation and nitrogenase activity in nitrogen-fixing woody plants stimulated by CO2 enrichment of the atmosphere. Physiologia Plantarum 71, 77–82. Norby, R.J., O’Neill, E.G., Hood, W.G. & Luxmoore, R.J. (1987) Carbon allocation, root exudation and mycorrhizal colonization of Pinus echinata seedlings grown under CO2 enrichment. Tree Physiology 3, 203–210. Norby, R.J., O’Neill, E.G. & Luxmoore, R.J. (1986) Effects of atmospheric CO2 enrichment on the growth and mineral nutrition of Quercus alba seedlings in nutrient-poor soil. Plant Physiology 82, 83–89. Norusis, M.J. (1994) SPSS Advanced Statistics 6.1. SPSS Inc, Chicago.

199 CO2, mycorrhizal colonization and P inflow

© 1999 British Ecological Society, Functional Ecology, 13, 190–199

O’Neill, E.G., Luxmoore, R.J. & Norby, R.J. (1987) Increases in mycorrhizal colonisation and seedling growth in Pinus echinata and Quercus alba in an enriched CO2 atmosphere. Canadian Journal of Forest Research 17, 878–883. Owensby, C.E., Coyne, P.I. & Auen, L.M. (1993) Nitrogen and phosphorus dynamics of a tallgrass prairie ecosystem exposed to elevated carbon dioxide. Plant, Cell and Environment 16, 843–850. Poorter, H. (1993) Interspecific variation in the growth response of plants to an elevated ambient CO2 concentration. Vegetatio 104/105, 77–97. Rogers, H.H., Runion, G.B. & Krupa, S.V. (1994) Plant responses to atmospheric CO2 enrichment with emphasis on roots and the rhizosphere. Environmental Pollution 83, 155–189. Rogers, H.H., Prior, S.A., Runion, G.B. & Mitchell, R.J. (1996) Root to shoot ratio of crops as influenced by CO2. Plant and Soil 187, 229–248. Rouhier, H. & Read, D.J. (1998) The role of mycorrhiza in determining the response of Plantago lanceolata to CO2 enrichment. New Phytologist 139, 367–373. Staddon, P.L., Graves, J.D. & Fitter, A.H. (1998) Effect of enhanced atmospheric CO2 on mycorrhizal colonisation by Glomus mosseae in Plantago lanceolata and Trifolium repens. New Phytologist 139, 571–580.

Stulen, I. & den Hertog, J. (1993) Root growth and functioning under atmospheric CO2 enrichment. Vegetatio 104/105, 99–115. Tennant, D. (1975) A test of a modified line intersect method of estimating root length. Journal of Ecology 63, 995–1001. Tester, M., Smith, S.E., Smith, F.A. & Walker, N.A. (1986) Effects of photon irradiance on the growth of shoots and roots, on the rate of initiation of mycorrhizal infection units in Trifolium subterraneum L. New Phytologist 103, 375–390. Tinker, P.B., Durall, D.M. & Jones, M.D. (1994) Carbon use efficiency in mycorrhizas: theory and sample calculations. New Phytologist 128, 115–122. Watson, R.T., Meira Filho, L.G., Sanhueza, E. & Janetos, A. (1992) Sources and sinks. Climate change 1992. The Supplementary Report to the IPCC Scientific Assessment (Eds J. T. Houghton, B. A. Callander & S. K. Varney), pp. 25–46. Cambridge University Press, Cambridge. Wyman, R.L. (1991) Global Climate Change and Life on Earth. Chapman & Hall, New York. Received 10 February 1998; revised 23 June 1998; accepted 7 July 1998