Context de...rrhizal species richness on

OIKOS 93: 353–364. Copenhagen 2001

Context dependent effects of ectomycorrhizal species richness on tree seedling productivity Lena M. Jonsson, Marie-Charlotte Nilsson, David A. Wardle and Olle Zackrisson

Jonsson, L. M., Nilsson, M.-C., Wardle, D. A. and Zackrisson, O. 2001. Context dependent effects of ectomycorrhizal species richness on tree seedling productivity. – Oikos 93: 353 – 364. While there has been much recent interest about the relationships between plant diversity and plant productivity, much remains unknown about how the diversity of mycorrhizal fungi affects plant productivity. We investigated the effects of ectomycorrhizal fungal community composition and diversity on the productivity and growth characteristics of seedlings of two tree species (Pinus syl6etris and Betula pendula) as well as their interactions with each other. This involved setting up a mycorrhizal fungal diversity gradient from one to eight species using a design previously demonstrated to be able to separate diversity effects from compositional effects. We found that the eight mycorrhizal fungal species differed in their effects on seedling productivity and that the nature of effects was determined by the fertility of the substrate. Fungal species richness effects were also important in affecting seedling productivity over and above what could be explained by ‘‘sampling effect’’ but only in some situations. For B. pendula in a low fertility substrate there were clear positive causative effects between fungal species richness and productivity with the eight species treatment having over double the productivity of any of the eight monoculture treatments; no diversity effects were, however, detected in a high fertility substrate. For P. syl6estris in a high fertility substrate there were significant negative effects of fungal diversity on productivity while in a low fertility substrate no effects were apparent. The possible mechanistic bases for these results are discussed. The growth of P. syl6estris relative to that of B. pendula when grown in combination was unaffected by mycorrhizal treatments. Our results provide clear evidence that effects of mycorrhizal fungal diversity on productivity are context dependent and may be positive, negative or neutral depending on the situation considered. L. M. Jonsson, M.-C. Nilsson (correspondence) and O. Zackrisson, Dept of Forest Vegetation Ecology, Swedish Uni6. of Agricultural Sciences, SE-901 83 Umea˚, Sweden ([email protected]) (present address of LMJ: Rutgers Pinelands Field Station, 501 Four Mile Road, P.O. Box 206, New Lisbon, NJ 08064, USA). – D. A. Wardle, Dept of Animal and Plant Sciences, Uni6. of Sheffield, UK S10 2TN.

There has been much recent interest in determining the nature of relationships which may exist between components of biodiversity (most notably species richness) and plant productivity. The majority of experimental studies which have been conducted to assess productivity responses to biodiversity have involved manipulation of plant diversity (e.g. Naeem et al. 1994, 1996, Tilman et al. 1996, Hooper and Vitousek 1997, 1998, Hector et al. 1999, Wardle et al. 2000) and the interpretations of many of these studies continue to be debated

(e.g. Aarssen 1997, Grime 1997, Huston 1997, Tilman et al. 1997, Wardle 1999). However, other biotic components of ecosystems are also important determinants of net primary productivity, including herbivores, micro-organisms which form symbiotic relationships with plants, pathogens and decomposer biota. Few studies have attempted to determine the effects of this consumer diversity on plant productivity. However, Laakso and Seta¨la¨ (1999) found that stimulation of tree seedling growth by soil fauna was unaffected by their

Accepted 6 February 2001 Copyright © OIKOS 2001 ISSN 0030-1299 Printed in Ireland – all rights reserved OIKOS 93:3 (2001)

353

diversity. Further, Van der Heijden et al. (1998) found greater productivity in experimental units which had a greater species richness of arbuscular mycorrhizae. However, the interpretation of the results of that study have been debated; Wardle (1999) suggested that their results can be explained entirely in terms of ‘‘sampling effect’’ [proposed by some ecologists as being an artefact of certain experimental designs (Aarssen 1997, Huston 1997)] while Van der Heijden et al. (1999) proposed that these results represent a real diversity effect on the grounds that mycorrhizal species that have disproportionate effect on plant productivity do not exist. In many temperate and boreal forest ecosystems the dominant tree species present form ectomycorrhizal associations. These associations are a major contributor to both soil biodiversity and ecosystem functioning, and are powerful regulators of ecosystem productivity and forest composition (Smith and Read 1997, Jonsson et al. 1999). However, although different ectomycorrhizal fungal species differ in their effects on tree growth (Smith and Read 1997, Herrmann et al. 1998), the effects of fungal species richness on host plants remain entirely unknown. In the present study we sought to experimentally determine the effects of ectomycorrhizal fungal species richness on tree seedling productivity using species that are characteristic of the boreal forests of northern Sweden. Because we were interested in determining whether diversity-productivity relationships were dependent upon context we performed the same experiment for each of two different tree species with different growth characteristics [i.e. Pinus syl6estris L. (coniferous) and Betula pendula Roth. (deciduous hardwood)], and for each of two soil types of differing fertility. The ultimate goal of our study was to contribute to a better understanding of how diversity and composition of heterotrophs may influence ecosystem productivity. Table 1. Fungal species combinations used in the mycorrhizal diversity experiment. Treatment

Species richness

Ectomycorrhizal species

A B

1 1

C

1

D E F G H I J K L M N O

1 1 1 1 1 2 2 2 2 4 4 8

Amanita muscaria (i) Cenococcum geophilum (ii) Hebeloma crustuliniforme (iii) Laccaria bicolor (iv) Lactarius rufus (v) Paxillus in6olutus (vi) Xerocomus badius (vii) X. subtomentosus (viii) (vi)+(viii) (i)+(iv) (iii)+(v) (ii)+(vii) (i)+(iv)+(vi)+(viii) (ii)+(iii)+(v)+(vii) all species

354

Materials and methods Selection of organisms and humus The growth substrate used for this experiment consisted of humus collected from the field at each of two sites located at Varjisa˚ n in the northern boreal zone of Sweden (66°01%N; 19°15%E) which differed in terms of fertility. One of these sites (hereafter referred to as the ‘‘low fertility’’ site) was dominated by P. syl6estris with a ground layer vegetative cover of the ericaceous dwarf shrubs Vaccinium myrtillus L. (bilberry), V. 6itis-idaea L. (cowberry), and Empetrum hermaphroditum Hagerup (crowberry) and the feathermoss Pleurozium schreberi (Bird) Mitt. At the site humus pH was 3.5 and the concentrations of total N was 0.82% and of total C was 43.2%. The other site (hereafter referred to as the ‘‘high fertility’’ site) was dominated by Picea abies (L.) Karst. and the ground layer consisted of herbaceous species [e.g. Melica nutans L., Geranium syl6aticum L. and Gymnocarpium drypoteris (L.) Newm.] characteristic of highly productive, fertile areas (Ha¨ gglund and Lundmark 1977). The humus pH was 4.7 and the concentrations of total N was 1.46% and of total C was 33.3%. Humus was collected from each site and after removal of large roots and coarse woody materials it was airdried and homogenised using a ‘‘garden mill’’ (BioMaster, Stiga, Sweden). PVC tubes (diameter= 50 mm; height =120 mm), each with a plastic perforated lid at the bottom, were filled with 10 mm of washed quartz sand (Silversand 90 mm) and 23 g (100 mm) of airdried humus placed on top of the sand. The tubes were then g-sterilised (25.0 kGy, Stril AB, Kopparberg, Sweden) to kill the resident mycoflora. Previous studies have demonstrated that g-sterilisation does not change soil nitrogen concentrations (Biro et al. 2000, Olff et al. 2000). We selected two tree species, i.e. B. pendula and P. syl6estris, for use in this experiment. Seeds of each species were surface sterilised by soaking in H2O2 for 10 min (B. pendula) or 20 min (P. syl6estris) to remove possible adhering fungi, rinsed in sterilised water, and transferred to autoclaved glass jars containing sterilised quartz sand. The seedlings were left to grow for six weeks in a climate chamber [L:D 16:8, temperature range= 20°C (day), 13°C (night)]. The birch seedlings were amended with a weak full nutrient solution (containing 7.1 mg N l − 1, 1.5 mg P l − 1, and 5.9 mg K l − 1) on one occasion. We selected eight ectomycorrhizal fungal species (Table 1) known to be compatible with both B. pendula and P. syl6etris and which occur commonly in Swedish boreal forests. Colonies of each species were isolated from sporocarps or sclerotia collected in boreal forests throughout Sweden, and maintained on MMN media (but with sucrose concentration reduced by 50%) (Marx 1969). These were maintained at room temperature and in darkness before the experiment. OIKOS 93:3 (2001)

Experimental set-up

Harvesting and measurements

We set up a full factorial experiment consisting of 15 fungal species combination treatments (Table 1), two humus treatments (high fertility and low fertility as described above) and three tree seedling treatments (P. syl6estris monoculture, B. pendula monoculture, and P. syl6estris +B. pendula mixture), resulting in 90 combinations in total. This experiment was set up with six replicate experimental units (tubes) for each possible treatment combination; all data analyses were done on the basis of six independent replicate blocks. Each tube was planted with two tree seedlings and was inoculated by the prescribed fungal species by placing eight 3-mm diameter agar plugs containing actively growing hyphae of the desired fungal species within the humus at time of planting. This was performed under sterile conditions. A total of eight plugs was used for each tube so that each tube had the same amount of total fungal inoculum at the start of the experiment; the number of total plugs used was therefore independent of mycorrhizal fungal species richness. These plugs were placed aseptically on the root system at the time of planting. After planting and inoculation, 3 mm of sterilised coarse sand was placed on the top of the humus in each tube in order to reduce evaporation and algal growth. The mycorrhizal species combination treatments (Table 1) represented a diversity gradient from one to eight species, with all species being represented in monoculture. This design means that performance of the multiple fungal species treatments can be directly compared with that of all component fungal species in monoculture. This design is conceptually the same as that used by Hooper and Vitousek (1997), Wardle et al. (1997a) and Nilsson et al. (1999), and enables detection of effects of species richness over and above that which can be explained by ‘‘sampling effect’’ (Huston 1997, Wardle 1999), in which the probability of including those fungal species which have the greatest effect on plant growth becomes greater when more fungal species are included (see Huston et al. 2000). Effects of fungal diversity on plant productivity over and above sampling effect occur without ambiguity whenever the productivity of the mixture is greater than that of any of the monocultures (Garnier et al. 1997, Huston et al. 2000). After set-up the tubes were maintained in a climate chamber with a light regime of L:D 16:8 and with the temperature ranging from 20°C (day) to 15°C (night); the light intensity used was 580 mmol m − 2 s − 1. The climate chamber was treated with UV-light and hot water to reduce potential contaminants before the experiment. Each tube was irrigated with 10 ml of distilled water daily for the first week and every two to three days thereafter.

The seedlings were harvested between 183 and 219 d after set up depending upon replicate block; replicate blocks were harvested sequentially. Upon harvest all plants were removed from the humus and shoot and root material rinsed and separated. Care was taken to ensure that root systems of different individuals in the same tube were separated from each other and kept as intact as possible. Root systems were maintained in a moist condition at 4°C for mycorrhizal assessment. For all seedlings in the seedling monoculture tubes, the root system was cut up in 2-cm pieces and 10 fragments of B. pendula and 20 fragments of P. syl6estris were randomly chosen. For each fragment, the total number of root tips and total number of mycorrhizal inoculated root tips (identified to fungal species based on morphology) were quantified. The mass of root subsamples used for mycorrhizal assessments, and the total root mass and shoot mass of each seedling were determined after oven drying at 70°C for 48 h. For the seedlings in seedling mixture tubes total root and shoot mass were determined after oven drying at 70°C for 48 h.

OIKOS 93:3 (2001)

Statistical analysis For each response variable, ANOVA was performed to test for effects of mycorrhizal species treatment, humus treatment, the interaction of these treatments, and blocking; tubes were assigned to one of six blocks at the time of harvest. Significance of mycorrhizal species effects between treatments was assessed by using the Least Significant Difference test. Data were transformed by log transformation or rank transformation as necessary to satisfy assumptions of ANOVA such as normality and homogeneity of variances.

Results Mycorrhizal fungal colonisation In the fungal monoculture treatments, all species successfully established except for C. geophilum on P. syl6estris grown in the high fertility substrate (Table 2: left hand data column). All other fungal species colonised at least 10% of the seedling root tips in the monoculture treatments and the majority of species colonised over 80% of the root tips. In the multiple fungal species treatments, significant competitive reduction (and possibly competitive exclusion) of fungal species frequently occurred (Table 2), although in some cases the percentage colonisation of some fungal species was actually higher (though not significantly so) in the fungal mixtures than in the monocultures. Generally, the total proportion of root tips that were colonised by 355

Table 2. Percentage of root tips colonised at the end of the experiment by each mycorrhizal species in monoculture, and when each species was grown together with one, three or seven other fungal species1. Species richness2 Tree species

Site

Fungal species

1

2

4

8

P. syl6estris

Low fertility

P. syl6estris

High fertility

B. pendula

Low fertility

B. pendula

High fertility

A. muscaria C. geophilum H. crustiliniforme L. bicolor L. rufus P. in6olutus X. badius X. subtomentosus A. muscaria C. geophilum H. crustiliniforme L. bicolor L. rufus P. in6olutus X. badius X. subtomentosus A. muscaria C. geophilum H. crustiliniforme L. bicolor L. rufus P. in6olutus X. badius X. subtomentosus A. muscaria C. geophilum H. crustiliniforme L. bicolor L. rufus P. in6olutus X. badius X. subtomentosus

36.8a 2.8a 55.2a 94.5a 76.5a 95.5a 88.3a 99.3a 87.3a 0.0a 78.1a 98.9a 78.9a 98.3a 85.9a 90.8a 94.3a 51.9a 59.1a 99.8a 99.8a 97.8a 99.8a 94.9a 55.0a 17.8a 93.5a 65.5a 10.6a 20.5ab 17.2a 34.3a

0.6b 6.6a 0.1b 95.4a 32.5b 95.8a 92.3a 0.5b 0.1b 0.0a 83.5a 86.4a 12.3b 86.7a 78.7a 0.0b 68.0ab 48.9a 4.1b 20.0b 95.9a 94.5a 51.0b 0.0b 0.0b 0.0a 78.8ab 63.2a 9.7a 37.4a 6.3b 0.0b

0.0b 2.5a 0.0b 85.8a 21.3b 13.6b 74.8a 0.0b 0.0b 0.5a 73.0a 50.5b 4.5b 48.7b 23.0b 0.0b 75.5ab 25.0ab 3.3b 17.3b 16.7b 12.8b 49.4b 0.0b 0.8b 0.0a 66.3b 57.6a 2.2a 12.6b 8.3ab 0.0b

0.8b 0.1a 0.0b 77.3a 0.0b 21.8b 0.0b 0.0b 0.9b 0.0a 0.0b 41.5b 0.0b 53.5b 0.0b 0.0b 48.2b 0.0b 0.0b 37.1b 8.3b 6.5b 0.0c 0.0b 0.0b 0.0a 0.0c 58.6a 0.0a 0.8b 0.0b 0.0b

1 2

Within each row numbers followed by different letters are significantly different at P =0.05 (Least Significant Difference test on ranked transformed data). Refer to Table 1 for species combinations.

Table 3. Total percentage colonization of root tips by mycorrhizas in the multiple fungal species treatments1. Treatment2

Range of monoculture treatments 2-species treatments:

4-species treatments: 8-species treatments: 1 2

Pinus syl6estris

I J K L M N O

Betula pendula

Low fertility

High fertility

Low fertility

High fertility

2.8–99.3 96.3a 96.0a 32.6b 98.9a 99.4a 98.6a 100.0a

0.0–98.9 86.7ab 86.5ab 95.9a 78.7b 99.2a 100.0a 95.9a

51.9–99.8 94.5a 88.0a 100.0a 99.0a 100.0a 94.4a 100.0a

17.2–93.5 37.4ab 63.2a 88.5a 6.3b 71.0a 76.8a 66.4a

Within each column numbers followed by different letters are significantly different at P =0.05 (Least Significant Difference test on ranked transformed data). Treatment codes as for Table 1.

mycorrhizal fungal in the multiple fungal species treatments was comparable to that of some of the monoculture treatments, though was much greater than that of other monoculture treatments (Table 3). However, for the high fertility B. pendula total root colonisation for the eight-species mixture was actually less than that for some other less diverse treatments (Table 3). 356

In the eight-species treatment for P. syl6estris, over 90% of root tips were colonised by just two species, i.e. L. bicolor and P. in6olutus. For B. pendula in the low fertility substrate, over 85% of root tips in the eightspecies treatment were colonised by L. bicolor and A. muscaria, while in the high fertility substrate nearly all colonisation was performed by L. bicolor and A. musOIKOS 93:3 (2001)

caria was entirely absent (Table 2). On average, no treatment had more than three mycorrhizal species present on the root tips at time of harvest and many inoculated species did not appear in any of the multiple species treatments at harvest (Table 2). Mycorrhizal contamination of root tips by non-inoculated species was extremely low and contamination was detected only in 12% of tubes; however 81% of tubes showed contamination in the C. geophilum treatment, in which colonisation of root tips by the inoculated species was lowest. To the best of our knowledge, none of the eight fungal species that were used for our treatments contaminated any of the experimental units to which they were not added. The density of root tips per unit mass of P. syl6estris fine roots was independent of mycorrhizal treatment (Table 4), but was significantly greater in the low fertility substrate (overall mean=12.6 tips/mg) than in the high fertility substrate (5.5 tips/mg). Betula pendula root tip density was significantly related to both mycorrhizal and fertility treatments. In the fungal monoculture treatments in the low fertility soil, the highest root tip density was in the X. badius treatment (28.3 tips/ mg); all other treatments resulted in root tip densities of less than 20.0 tips/mg with the least being in the A. muscaria treatment (8.0 tips/mg). For the high fertility substrate root tip density was greatest in the X. badius treatment (22.5 tips/mg) and least in the P. in6olutus treatment (11.0 tips/mg). In all cases root tip density in the multiple fungal species treatments did not differ significantly from what we would expect based on the data for the component fungal species in monoculture and root tip density was not significantly correlated with fungal diversity (Table 4). The total proportion of root tips colonised by mycorrhizal fungi was significantly influenced by both mycorrhizal species treatment and substrate (Tables 2–4) but the total colonisation in the multiple fungal species mixtures generally did not differ much from that of some of the monoculture treatments (Tables 2, 3). Further, total colonisation at harvest was not related to fungal diversity regardless whether it was measured at the beginning or the end of the experiment (Table 7). For P. syl6estris, a significantly smaller proportion of root tips was colonised by mycorrhizae in the low

fertility substrate (overall mean= 78.2%) than in the high fertility substrate (84.3%). For B. pendula the reverse pattern was observed, i.e. 92.8% colonisation in the low fertility substrate and 47.9% in the high fertility substrate.

Seedling response Mycorrhizal fungal treatments had significant effect on P. syl6estris shoot weight, root weight and shoot:root ratio, and on B. pendula seedling height and shoot weight (Table 5). Seedling productivity was frequently influenced by fungal species across the fungal species monoculture treatments (Table 6). For P. syl6estris in the low fertility substrate seedlings grew largest when inoculated with X. badius and C. geophilum and least when inoculated with H. crustiliniforme and L. bicolor (Table 6). A similar pattern was observed for P. syl6estris in the high fertility substrate although L. rufus also had strong effects on seedling growth. With regard to B. pendula in the low fertility substrate, seedling growth was best when inoculated with A. muscaria or X. subtomentosus and least with L. rufus and P. in6olutus (Table 6). In the high fertility substrate there was a partial reversal of these effects with growth being greatest in the L. rufus treatment and least in the A. muscaria treatment. In the low fertility substrate different patterns of shoot to root allocation existed between P. syl6estris and B. pendula. The growth of both seedling species was significantly greater in the high fertility than the low fertility substrate, and P. syl6estris shoot:root ratio was greater for the high fertility than the low fertility substrate (Table 6). There was no effect of increasing mycorrhizal species richness on P. syl6estris productivity in the low fertility substrate (Table 7) and performance of P. syl6estris in the multiple fungal species treatments was consistent with what we would expect based on the performance of seedlings in the fungal monoculture treatments (Fig. 1). However, in the high fertility substrate fungal diversity had significant negative effects on shoot weight and shoot:root ratio (Table 7). Shoot mass, root mass and shoot:root ratio of P. syl6estris seedlings grown on the substrate was less in the eight-fungal species treatment

Table 4. Effects of mycorrhizal species combinations (M) and soil type (S) on root tip density and mycorrhizal colonisation as shown by F-values derived from analysis of variance1. Response variable P. B. P. B.

2

syl6estris root tip density pendula root tip density syl6estris colonisation3 pendula colonisation

M

S

M×S interaction

Blocking

0.72 6.49*** 19.53*** 2.57**

226.29*** 186.80*** 4.15* 5.46*

0.98 6.94*** 3.83*** 1.00

3.88*** 0.88 2.94* 0.25

1

Degrees of freedom for F are: 14,145 for M; 1,145 for S; 14,145 for M×S; and 5,145 for blocking. *, **, ***= F-value significantly different from zero at P= 0.05, 0.01, 0.001, respectively. Number of root tips per unit dry weight of root. 3 Proportion of total root tips colonised by mycorrhizae. 2

OIKOS 93:3 (2001)

357

Table 5. Effects of mycorrhizal species combinations (M) and soil type (S) on plant response variables as shown by F-values derived from analysis of variance1. Response variable Seedling monocultures P. syl6estris seedling height B. pendula seedling height P. syl6estris shoot weight B. pendula shoot weight P. syl6estris root weight B. pendula root weight P. syl6estris s:r ratio2 B. pendula s:r ratio Seedling mixtures P. syl6estris shoot weight B. pendula shoot weight P. syl6estris root weight B. pendula root weight Total plant weight per tube P. syl6estris mix:mono3 B. pendula mix:mono Mass mix:mono4

M

S

M×S interaction

Blocking

1.30 2.42** 3.71*** 1.80* 2.05* 1.06 1.80* 1.16

29.64*** 970.60*** 858.20*** 978.03*** 810.13*** 1155.59*** 156.30*** 19.05

1.32 2.89*** 1.80* 2.66** 1.25 1.55 1.38 1.41

1.08 1.09 1.51 2.67* 1.06 1.39 1.83 0.42

1.34 0.62 1.21 0.61 1.30 1.46 0.34 1.44

1.47 2.81 0.77 1.58 1.17 0.47 1.38 0.74

1.48 0.73 0.98 0.72 1.41 1.40 0.93 1.36

160.97*** 62.17*** 149.27*** 68.45*** 766.10*** 9.71*** 19.37*** 22.54***

1

Degrees of freedom for F are: 14,145 for M; 1,145 for S; 14,145 for M×S; and 5,145 for blocking. *, **, ***= F-value significantly different from zero at P= 0.05, 0.01, 0.001, respectively. Shoot:root mass ratio. 3 Mean seedling mass in two seedling species mixture as a proportion of that in seedling monoculture. 4 Total plant mass in two seedling species pots as a proportion of the average of that in the P. syl6estris and B. pendula monoculture pots. 2

than in any of the fungal monoculture treatments (Fig. 1). For the low fertility substrate there were very strong effects of species diversity on productivity of B. pendula both above and below ground (Table 7), and this was particularly apparent in the treatments with four and eight fungal species (Fig. 2). Root mass in the eight fungal species treatment was more than double that in any of the eight fungal monoculture treatment (Fig. 2). In the high fertility substrate there was no general effect of diversity on B. pendula productivity (Table 7), although there was a strong positive effect of mixing C. geophilum and X. badius together relative to the monoculture treatments of these two fungal species (Fig. 2), indicative of a strong synergistic effect. When P. syl6estris and B. pendula seedlings were grown together, total plant mass (or that of either of the species considered alone) was not significantly affected by mycorrhizal treatment (Table 5). However, in the low fertility treatment there was a general trend of lower total productivity of both shoots and roots when more mycorrhizal species were present; this was especially apparent in the eight species treatments (Fig. 3). There were no clear trends with regard to total shoot or root biomass in the mixed seedling tubes for the higher fertility soil (Fig. 3). Individual P. syl6estris seedlings grew larger when grown with a B. pendula seedling than when grown with another P. syl6estris seedling. The magnitude of this effect was not influenced by mycorrhizal fungal species treatment (Table 5) but was affected by substrate fertility. In the low fertility substrate the mass of individual 358

P. syl6estris seedlings when grown with a B. pendula seedling was 2.67 times greater than when grown with another seedling of its own species; in the high fertility substrate this ratio was only 1.83. Meanwhile individual B. pendula seedlings grew smaller when grown with a P. syl6estris seedling than with another B. pendula seedling. Again the magnitude of this effect was not influenced by fungal species treatment but was affected by substrate fertility (Table 5, Fig. 3). The ratio of mass of individual B. pendula seedlings when grown with P. syl6estris relative to mass of B. pendula grown with another B. pendula was 0.23 in the low fertility substrate and 0.57 in the high fertility substrate. The total productivity of tubes containing both tree seedling species was greater than the mean productivity of tubes containing P. syl6estris singly and B. pendula singly. The magnitude of this stimulation was unaffected by mycorrhizal treatment but did vary according to substrate fertility (Table 5); the ratio of productivity in the two seedling species tubes to the mean productivity in the P. syl6estris and B. pendula monospecific tubes was 2.09 in the low fertility substrate and 1.20 in the high fertility substrate.

Discussion It is becoming increasingly recognised that individual plant species effects can be very important in determining ecosystem properties (Vitousek and Walker 1989, Hobbie 1992, Wardle et al. 1997b). Our study provides evidence that effects of individual mycorrhizal fungal OIKOS 93:3 (2001)

4.9 119 111 0.22 3.5 40 20 1.56 8.1 186 111 0.32 11.6 231 214 0.92 32.3 101 139 0.73 11.5 54 31 1.72 32.8 677 139 1.08 46.7 796 786 1.04 28.0 272 279 0.83 12.1 48 25 2.2 37.1 839 279 1.30 60.2 918 658 1.41 31.7 140 193 0.73 10.0 31 15 3.89 36.5 723 193 1.19 51.4 870 714 1.13 29.7 146 196 0.80 11.4 21 10 2.28 37.3 924 196 1.53 47.6 824 806 1.08 31.3 101 131 0.78 10.9 50 15 1.61 37.6 753 131 1.32 37.2 603 731 0.82 31.9 88 142 0.68 11.0 36 21 1.81 35.0 694 142 1.19 45.3 607 689 0.92 32.8 170 256 0.69 12.3 47 24 2.48 29.5 867 256 1.34 49.0 836 693 1.35 1

B. pendula

P. syl6estris High fertility

B. pendula

Treatment codes correspond to Table 1.

31.0 109 173 0.65 11.8 46 38 1.19 37.3 776 173 1.32 33.0 576 679 2.06 P. syl6estris Low fertility

Seedling height (mm) Shoot weight (mg/pot) Root weight (mg/pot) Shoot:root ratio Seedling height (mm) Shoot weight (mg/pot) Root weight (mg/pot) Shoot:root ratio Seedling height (mm) Shoot weight (mg/pot) Root weight (mg/pot) Shoot:root ratio Seedling height (mm) Shoot weight (mg/pot) Root weight (mg/pot) Shoot:root ratio

LSD0.05 H G F E A

B

C

D

Mycorrhizal treatment Response variable Tree species Site

Table 6. Seedling growth characteristics for mycorrhizal monoculture treatments1.

OIKOS 93:3 (2001)

species can influence ecosystem properties through affecting net primary productivity. This is consistent with earlier studies suggesting that mycorrhizal fungal species differ in their effects on plant growth (Le Tacon 1992, Jones et al. 1998, Van der Heijden et al. 1998, Kiens et al. 2000). Because the two plant species used in our study also differed with regard to the mycorrhizal fungal species that they preferentially supported, our study supports the view that a feedback may exist between plant community structure and mycorrhizal community structure (Allen et al. 1995, Herrmann et al. 1998, Kiens et al. 2000). Our results also showed that different mycorrhizal fungal species differ in their effects on plant physiological attributes. For example, the shoot:root ratio of P. syl6estris was influenced by mycorrhizal fungal species indicating that fungal species differed with regard to their effects on nutrient acquisition by these seedlings (Chapin 1980, Smith and Read 1997). Similarly, different mycorrhizal fungal species varied in their effects on root tip density of B. pendula seedlings, suggesting that fungal species effects are important determinants of seedling root morphology and therefore nutrient foraging ability. Soil fertility was important in influencing plant-fungal associations and in particular determined which mycorrhizal fungal species dominated on the roots of B. pendula. The fact that different fungal species were favoured by different host plant species and different soil fertilities points to a considerable potential for resource partitioning between fungal species, although in the multiple fungal species treatments there was also evidence of considerable niche overlap given the frequency with which substantial competitive reduction (and possible competitive exclusion) of subordinate fungal species occurred. The degree of resource partitioning (i.e. resource use complementarity) relative to niche overlap (i.e. compensatory effects) is likely to be an important determinant of whether diversity-function relationships occur (Lawton and Brown 1993, Hooper and Vitousek 1997). In our study we found mycorrhizal fungal species diversity to be important in determining plant productivity for two out of four cases. For one of these cases, i.e. B. pendula in the low fertility substrate, there were clear positive effects of fungal diversity on plant growth. This is likely to be because a wider range of fungal species would result in a greater range of resources in the soil being accessed and therefore a greater availability of nutrients for plant growth. This effect was not observed in the high fertility treatment, reflecting that seedlings are more dependent upon mycorrhizae in low fertility than in high fertility situations, and that resource use complementarity by different fungal species is greater when nutrient conditions are poorer. For the second of these cases, i.e. P. syl6estris in the high fertility situation, the reverse pattern was detected and increasing fungal diversity reduced seedling growth. This suggests that in contrast to the pattern observed for B. pendula 359

Table 7. Pearson’s correlation coefficient (r) between plant response variables and two measures of mycorrhizal diversity [i.e. mycorrhizal species richness (log2-transformed) at the start of the experiment (S1) and Shannon-Weiner diversity index for mycorrhizal morphotypes at time of harvest (S2)]1. Response variable

2

Root tip density Mycorrhizal colonisation3 Plant height Shoot mass Root mass Shoot:root ratio

Low fertility site

High fertility site

P. syl6estris

B. pendula

S1

S1

S2

S1

S2

S1

S2

−0.148 0.356

−0.433 0.347

−0.312 0.357

−0.097 0.356

−0.101 0.365

−0.205 0.479

0.827*** 0.729** 0.807*** −0.458

0.620* 0.463 0.545* −0.365

0.048 −0.621* −0.207 −0.659**

0.068 −0.631* −0.241 −0.646**

−0.044 −0.093 −0.262 −0.184

−0.141 −0.083 −0.393 −0.039

S2 0.052 0.390

0.343 0.412

−0.335 −0.045 −0.021 −0.120

−0.156 0.106 0.167 −0.001

P. syl6estris

B. pendula

*, **, *** = F-value significantly different from zero at P=0.05, 0.01, 0.001, respectively. All correlations are performed using 15 independent data points, i.e. the means of each of the 15 treatments A–O shown in Table 1. Diversity indices for mycorrhizal morphotypes at time of harvest are determined based on the proportions of root tips colonised by each of the morphotypes (i.e. fungal species). 2 Number of root tips per unit dry weight of root. 3 Proportion of root tips colonised. 1

Fig. 1. Shoot and root weight and shoot:root ratio for seedlings of P. syl6estris in all multiple fungal species treatments. For each bar in each panel, associated crosses indicate mean values for monoculture (single mycorrhizal species) treatments of each of the mycorrhizal species present in the mixture. Crosses outside dotted bars differ significantly from the treatment mean at P= 0.05 (Least Significant Difference test). Treatment codes correspond to Table 1.

360

OIKOS 93:3 (2001)

Fig. 2. Shoot and root weight and shoot:root ratio for seedlings of B. pendula in all multiple fungal species treatments. Legend as for Fig. 1.

FIg. 3. Total shoot and root weights in tubes containing both tree seedling species in all multiple fungal species treatments. Legends as for Fig. 1; for each bar the mass below each solid dividing line represents that of B. pendula while the mass above each diving line represents that of P. syl6estris. The dotted lines for each bar apply only to total mass values.

(low fertility), greater resource use complementarity by fungi in the more diverse treatment probably resulted in the fungi being able to compete more effectively against the host plant thus reducing host plant biomass. This is consistent with some studies suggesting that mycorOIKOS 93:3 (2001)

rhizal fungi can suppress plant growth through direct competition or ‘‘controlled parasitism’’ (Dighton and Mason 1985, Colpaert et al. 1992). In the case of B. pendula, the lower colonisation by mycorrhizal fungi in the higher fertility soil is consistent 361

with what we would expect based on previous investigations (reviewed by Smith and Read 1997). The finding that there was greater colonisation of mycorrhizal fungal under higher fertility for P. syl6estris runs counter to most previous work, but we identify three possible mechanisms as to why higher fertility could have promoted greater colonisation on roots in our study. Firstly, plants that grew better in the high fertility soil could have provided more carbon exudates from their roots, thus promoting greater initial fungal colonisation (cf. Salonen et al. 2000). Secondly, fungi in high fertility situations may be better able to utilise carbon for sources other than those directly supplied by living plants (e.g. root litter, soil organic matter), thus enabling improved colonisation of live roots (see Smith and Read 1997). Thirdly, the low fertility soil was collected from under ericaceous vegetation which is well known to produce high amounts of phenolics; phenolics derived from ericaceous vegetation in the Swedish boreal forest have previously been shown to significantly reduce mycorrhizal colonisation of tree seedling roots (Nilsson et al. 1993). In the case of B. pendula (low fertility) for which productivity was greatest in the eight-species treatment, there were only two to three fungal species present on the root tips at the time of harvest. This level of fungal species richness did not differ much from species richness on root tips at time of harvest for the two- and four-fungal species treatments. This result is explicable in terms of those fungal species which were at extremely low densities on the root tips (and may have been competitively excluded) surviving in the soil as saprophytes and enabling the plant to indirectly access a greater amount of mineralised nutrients. In this context all eight fungal species we used are capable of survival in soil as saprophytes independent of the host plant. Most ectomycorrhizal fungal species demonstrate some degree of cellulose-degrading and proteolytic ability (Smith and Read 1997), and all eight species can be grown on agar without a host plant present which is indicative of at least some degree of independent saprophytic ability. Previous experimental studies have claimed to present evidence that plant productivity can be greatly enhanced by species richness of plants (Naeem et al. 1994, Tilman et al. 1996, Hector et al. 1999) and mycorrhizal fungi (Van der Heijden et al. 1998). However, it is unclear as to the extent to which these results can be explained by sampling effect or by resource use complementarity, and the interpretation of these studies continues to be debated (Huston 1997, Van der Heijden et al. 1999, Wardle 1999, Hector et al. 2000, Huston et al. 2000). To unambiguously detect effects of diversity over and above that which can be explained by sampling effect alone requires replicated monoculture treatments of all component species, and demonstration that the mixtures perform better than all of the monocul362

tures (Garnier et al. 1997, Huston et al. 2000). In the present study, we can demonstrate that in two out of four cases fungal species diversity clearly affected plant productivity in excess of what can be explained by sampling effect because in both of these cases the performance of the eight-species treatment differed significantly from that of all of the monocultures. However, although we have demonstrated that effects of species richness on productivity in our study must be due at least in part to mechanisms which do not include sampling effect, there are still difficulties with this type of design in quantifying the relative contribution of sampling effect and true diversity mechanisms (such as resource use complementarity) to the measured results (Hector et al. 2000, Huston et al. 2000). Previous studies investigating the effects of species richness of organisms on ecosystem properties have usually expressed their results in terms of species richness at the start of the experiment rather than at the time of harvest, and have often not measured the species richness at the end of the experiment (or even whether all of the species used actually survived through the course of the experiment) (see Wardle 1999). In contrast, in this experiment we demonstrated that all eight fungal species survived in monoculture throughout the experiment, although not all species were detected in the eight-species mixture. It is unclear as to whether those species which were not detected in the eight-species mixture were entirely absent from those mixtures, or whether they actually occurred but at such low levels that they failed to be detected simply because only a subsample of the total root system of each seedling was examined. Regardless, the relationships between species richness and plant growth were similar regardless of whether we considered fungal species diversity at the start of the experiment or at the time of harvest (Table 7). For organism groups regulated by resource availability or amount, the effects of their species richness on plant productivity are likely to be greater when there is greater resource use complementarity (i.e. resource partitioning) and when competition intensity among species is therefore less (Hooper 1998). In our study, the effects of fungal diversity on productivity were influenced by soil fertility for both tree species, suggesting that fertility influences the balance between resource use complementarity and competition among fungal species. This is consistent with studies which suggest that fertility can influence competition intensity among saprophytic fungal species (Carreiro and Koske 1992, Stahl and Christensen 1992), and that effects of organism species richness on productivity can depend upon nutrient availability (Austin and Austin 1980). This points clearly to the effects of species richness on productivity being context dependent. Further, even when effects of species richness occurred, the direction of these effects was not predictable. However, our OIKOS 93:3 (2001)

results suggest that both fungal species composition and fungal species diversity can operate as determinants of ecosystem productivity. This may also have longerterm consequences for plant community structure. Although we failed to detect effects of mycorrhizal fungal treatments on the performance of each tree seedling species in mixture relative to in monoculture, we did identify specific effects of individual mycorrhizal fungal species on seedling growth. Our results in tandem ultimately provide evidence that mycorrhizal community structure and diversity may both have important consequences for the ways that boreal forest ecosystems function. Acknowledgements – We thank Ulla Ahonen-Jonart and Andy Taylor for providing fungal cultures, Morgan Karlsson, Linda Berglund, Anna Shevtsova, Erik Hellberg and Bjo¨ rn Eriksson for help at various stages with running the experiment. Financial support came from the Swedish Council for Forestry and Agricultural Research.

References Aarssen, L. W. 1997. High productivity in grassland ecosystems: effected by species diversity or productive species? – Oikos 80: 183–184. Allen, E. B., Allen, M. F., Helm, D. J. et al. 1995. Patterns and regulation of mycorrhizal plant and fungal diversity. – Plant Soil 170: 47–62. Austin, M. P. and Austin, B. O. 1980. Behaviour of experimental plant communities along a nutrient gradient. – J. Ecol. 68: 191–198. Biro, B., Ko¨ ves-Pe´ chy, K., Vo¨ ro¨ s, I. et al. 2000. Interrelations between Azospirillum and Rhizobium nitrogen-fixers and arbuscular mycorrhizal fungi in the rhizosphere of alfalfa in sterile, AMF-free or normal soil condition. – Appl. Soil Ecol. 15: 159–168. Carreiro, M. M. and Koske, R. E. 1992. The effect of temperature and substratum on competition among three species of forest litter microfungi. – Mycol. Res. 96: 19– 22. Chapin, F. S. III. 1980. The mineral nutrition of wild plants. – Annu. Rev. Ecol. Syst. 11: 223–260. Colpaert, J. V., Vanassche, J. A. and Luijtens, K. 1992. The growth of the extramatrical mycelium of ectomycorrhizal fungi and the growth-response of Pinus syl6estris L. – New Phytol. 120: 127–135. Dighton, J. and Mason, P. A. 1985. Mycorrhizal dynamics during forest tree development. – In: Moore, D., Casselton, L. A., Wood, D. A. and Frankland, J. C. (eds), Developmental biology of higher fungi. Cambridge Univ. Press, pp. 117–139. Garnier, E., Navas, M. L., Austin, M. P. et al. 1997. A problem for biodiversity-productivity studies: how to compare the productivity of multispecific plant mixtures to that of monocultures. – Acta Oecol. 18: 657– 670. Grime, J. P. 1997. Biodiversity and ecosystem function: the debate deepens. – Science 277: 1260–1261. Ha¨ gglund, B. and Lundmark, J. E. 1977. Site index estimation by means of site properties of Scots pine and Norway spruce in Sweden. – Stud. For. Suec. 138: 1– 38. Hector, A., Schmid, B., Beierkuhnlein, C. et al. 1999. Plant diversity and productivity in European grasslands. – Science 286: 1123–1127. Hector, A., Schmid, B., Beierkuhnlein, C. et al. 2000. No consistent effects of plant diversity on productivity (response). – Science 289: 1255a. Herrmann, S., Munch, J.-C. and Buscot, F. 1998. A OIKOS 93:3 (2001)

gnotobiotic culture system with oak microcuttings to study specific effects of mycobionts on plant morphology before, and in the early phase of, ectomycorrhiza formation by Paxillus in6olutus and Piloderma croceum. – New Phytol. 138: 203– 212. Hobbie, S. E. 1992. Effects of plant species on nutrient cycling. – Trends Ecol. Evol. 7: 336– 339. Hooper, D. U. 1998. The role of complementarity and competition in ecosystem responses to variation in plant diversity. – Ecology 79: 704– 719. Hooper, D. U. and Vitousek, P. M. 1997. The effects of plant composition and diversity on ecosystem processes. – Science 277: 1302– 1305. Hooper, D. U. and Vitousek, P. M. 1998. Effects of plant composition and diversity on nutrient cycling. – Ecol. Monogr. 68: 121– 149. Huston, M. A. 1997. Hidden treatments in ecological experiments: re-evaluation the ecosystem function of biodiversity. – Oecologia 110: 449– 460. Huston, M. A., Aarssen, L. W., Austin, M. P. et al. 2000. No consistent effect of plant diversity on productivity. – Science 289: 1255a. Jones, M. D., Durall, D. M. and Tinker, P. B. 1998. Comparison of arbuscular and ectomycorrhizal Eucalyptus coccifera: growth response, phosphorus uptake efficiency and external hyphal production. – New Phytol. 140: 125 – 134. Jonsson, L., Dahlberg, A., Nilsson, M.-C. et al. 1999. Continuity of ectomycorrhizal fungi in self-regenerating boreal Pinus syl6estris forests studied by comparing mycobiont diversity on seedlings and mature trees. – New Phytol. 142: 151 – 162. Kiens, E. T., Lovelock, C. E., Krueger, E. L. and Herre, E. A. 2000. Differential effects of tropical arbuscular mycorrhizal fungal inocula on root colonization and tree seedling growth: implications for tropical forest diversity. – Ecol. Lett. 3: 106– 113. Laakso, J. and Seta¨ la¨ , H. 1999. Sensitivity of primary production to changes in the architecture of belowground food webs. – Oikos 87: 57 – 64. Lawton, J. H. and Brown, V. K. 1993. Redundancy in ecosystems. – In: Schulze, E.-D. and Mooney, H. A. (eds), Biodiversity and ecosystem function. Springer-Verlag, pp. 225 – 270. Le Tacon, F. 1992. Variations in field response of forest trees to nursery ectomycorrhizal inoculation in Europe. – In: Read, D. J., Lewis, D. H., Fitter, A. H. and Alexander, I. (eds), Mycorrhizas in ecosystems. CAB International, pp. 119 – 134. Marx, D. H. 1969. The influence of ectotrophic mycorrhizal fungi on the resistance of pine roots to pathogenic infections. I. Antagonism of mycorrhizal fungi to root pathogenic fungi and soil bacteria. – Phytopathology 59: 153 – 163. Naeem, S., Thompson, L. J., Lawler, S. P. et al. 1994. Declining biodiversity can alter the performance of ecosystems. – Nature 368: 734– 737. Naeem, S., Ha˚ kansson, K., Lawton, J. H. et al. 1996. Biodiversity and plant productivity in a model assemblage of plant species. – Oikos 76: 259 – 264. Nilsson, M.-C., Ho¨ gberg, P., Zackrisson, O. and Fengyou, W. 1993. Allelopathic effects by Empetrum hermaphroditum on development and nitrogen uptake by roots and mycorrhizae of Pinus syl6estris. – Can. J. Bot. 71: 620– 628. Nilsson, M.-C., Wardle, D. A. and Dahlberg, A. 1999. Effects of plant litter species composition and diversity on the boreal forest plant-soil system. – Oikos 86: 16 – 26. Olff, H., Hoorens, B., Goede, R. G. M. et al. 2000. Small scale shifting mosaics of two dominant grassland species: the possible role of soil-borne pathogens. – Oecologia 125: 45 – 54. Salonen, V., Seta¨ la¨ , H. and Puustinen, S. 2000. The interplay between Pinus syl6estris, its root hemiparasite, Melampyrum pratense, and ectomycorrhizal fungi: influ-

363

´ coscience 7: ences on plant growth and reproduction. – E 195 – 200. Smith, S. E. and Read, D. J. 1997. Mycorrhizal symbiosis. – Academic Press. Stahl, P. D. and Christensen, M. 1992. In 6itro mycelia competition among members of a soil microfungal community. – Soil Biol. Biochem. 24: 309–316. Tilman, D., Wedin, D. and Knops, J. 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. – Nature 379: 718–720. Tilman, D., Naeem, S., Knops, J. et al. 1997. Biodiversity and ecosystem properties. – Science 278: 1866–1867. Van der Heijden, M. G. A., Klironomos, J. N., Ursic, M. et al. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. – Nature 396: 69– 72. Van der Heijden, M. G. A., Klironomos, J. N., Ursic, M. et al. 1999. ‘‘Sampling effect’’, a problem in biodiversity manipulation? A reply to David A. Wardle. – Oikos 87: 408– 410.

Vitousek, P. M. and Walker, L. R. 1989. Biological invasion by Myrica faya in Hawaii: plant demography, nitrogen fixation, ecosystem effects. – Ecol. Monogr. 59: 247 – 265. Wardle, D. A. 1999. Is ‘‘sampling effect’’ a problem for experiments investigating biodiversity-ecosystem function relationships? – Oikos 87: 403 – 407. Wardle, D. A., Bonner, K. I. and Nicholson, K. S. 1997a. Biodiversity and plant litter: experimental evidence which does not support the view that enhanced species richness improves ecosystem function. – Oikos 79: 247 – 258. Wardle, D. A., Zackrisson, O., Ho¨ rnberg, G. and Gallet, C. 1997b. The influence of island area on ecosystem properties. – Science 277: 1296– 1299. Wardle, D. A., Bonner, K. I. and Barker, G. M. 2000. Stability of ecosystem properties in response to above ground functional group richness and composition. – Oikos 89: 11– 23.

.

364

OIKOS 93:3 (2001)