• Paris -type arbuscular mycorrhizas (AM) are reportedly the most common morphological type of AM; however, most research has focused on the Arum -type. Asphodelus fistulosus , a common weed in southern Australia, forms Paris -type AM when colonised by Glomus coronatum . It is often found in sites with low nutrient levels, and may therefore be dependent on its AM associations for growth and phosphorus (P) nutrition.
• A. fistulosus was inoculated with G. coronatum and grown in pots containing a soil/sand mixture with P added to give five soil P concentrations. The plants were grown in a glasshouse and harvested 6 and 9 wk after planting, at which times growth, P nutrition and colonisation were measured.
• At low soil P, A. fistulosus showed very marked positive responses to colonisation both in P uptake and growth; both responses decreased with increasing P supply. Colonisation was not greatly reduced by increasing P supply.
• This study appears to be one of the first detailed investigations of P responses in a Paris -type AM, providing insight into what is reportedly the more common but less well studied morphological type of AM.
Arbuscular mycorrhizas (AM) are formed when the roots of plants are colonised by Glomalean fungi, and occur in most terrestrial plant communities. Benefits to the plant as a result of the formation of AM are well described, based on work carried out almost exclusively using only one of the two main morphological types of AM: the Arum-type (Abbott & Robson, 1977; Bruce et al., 1994; Dickson et al., 1999, and many others). The most widely studied benefit is that of improved acquisition of mineral nutrients, particularly phosphorus (P), and subsequent transfer to the plant. The fungus in turn benefits from a supply of organic carbon (C) from the plant.
An extensive survey of the literature has established that the other main morphological type of AM (the Paris-type) occurs in more plant families than the Arum-type (Smith & Smith, 1997). Despite this, there is a paucity of developmental or physiological research on plants forming Paris-type AM (Smith & Smith, 1997), and where this has been carried out (e.g. on Acer saccharum and Panax quinquefolius) the potential functional significance of the structural differences between the Arum- and Paris-types, particularly with respect to occurrence of both simple and arbusculate coils within root cortical cells of Paris-type AM, has not been appreciated (Gallaud, 1905; Smith & Smith, 1997; Dickson & Kolesik, 1999; Cavagnaro et al., 2001a,b,c). Burleigh et al. (2002) showed that Paris-type AM formed in tomato by Scutellospora calospora and Gigaspora rosea were relatively inefficient in promoting shoot growth and P uptake, compared with Arum-type AM formed by G. mosseae and G. versiforme. The experiment had one harvest and one P level and was not designed to investigate details of the development and physiology of the interaction. We have investigated the effects of mycorrhizal colonisation on growth and P nutrition of Asphodelus fistulosus L., a plant of Mediterranean origin and an important weed in southern Australia that forms Paris-type AM with Glomus coronatum (Cavagnaro et al., 2001b, 2001c). This is the first detailed investigation of the physiology and P nutrition of a plant that forms Paris-type AM, under controlled experimental conditions.
Materials and Methods
Plant growth and phosphorus treatments
Asphodelus fistulosus L. seeds (collected from Ardrossan, South Australia) were surface sterilised in a 3% sodium hypochlorite solution for 10 min, imbibed in aerated reverse osmosis (RO) water for 20 min and placed on filter paper in sealed Petri dishes in the dark at 25°C for 3 d to germinate. The germinated seeds were planted into individual 70 mm diameter, plastic, nondraining pots (Polar Cup, Australia) containing 400 g of a 1 : 9 (w/w) mixture of inoculum and soil/sand mix. The inoculum was a mixture of soil and roots derived from a pot culture of Glomus coronatum Giovannetti (WUM 16, formerly known as Glomus sp. ‘City Beach’) raised on Allium porrum L. in the same soil/sand mix as that used for the experiment. The soil/sand mix was composed of a 1 : 9 (w/w) mixture of soil collected from Mallala, South Australia, with 16 mg bicarbonate extractable P kg −1 soil (see below) and washed river sand. The soil/sand mix had a final bicarbonate extractable P concentration of 6.6 mg kg −1 .
Five P treatments were included. Prior to planting the germinated A. fistulosus seeds, the P content of the inoculum and soil/sand mix was altered by mixing NaH2PO4 solutions thoroughly into the soil. The final soil P concentrations were 6.6 (no added P), 14.7, 30.7, 43.0 and 61.5 mg P kg−1 soil, referred to as P1, P2, P3, P4 and P5, respectively. Soil P was determined using a modification of the method of Colwell (1963), sulfuric acid was replaced with hydrochloric acid. The plants were grown in a glasshouse with mean annual diurnal temperatures of 22°C day and 14°C night and watered three times each week to 10% of soil d. wt. Once a week, starting 2 wk after planting, N and other nutrients were added in 5 ml of modified Long Ashton solution minus P (Cavagnaro et al., 2001c). For all treatments, nonmycorrhizal control pots were prepared in the same manner, the only difference being that the inoculum was replaced with soil and roots derived from non-mycorrhizal pot cultures. There were four replicate pots for each treatment.
Plants were harvested 6 and 9 wk after planting. The roots were carefully washed free from the soil with RO water. The roots and shoots were separated and fresh weights determined. A weighed subsample of root material was kept for assessment of colonisation. The remainder of the roots and the whole shoots were dried in an oven at 80°C for 48 h, weighed, and P concentrations were determined using the phosphovanado-molybdate method (Hanson, 1950). Seed d. wt and P contents were also determined to give a starting point for analysis of growth responses.
Assessment of colonisation
The roots were cleared and stained using a modification of the method of Phillips & Hayman (1970) omitting phenol from reagents. The percentage of the root length colonised and percentage of the colonised root length with external hyphae (EH), intracellular hyphal coils (HC) and arbusculate coils (AC) were determined using the magnified intersects technique (McGonigle et al., 1990).
Calculations and data analysis
Mycorrhizal growth responses (MGR) were calculated using the individual total plant d. wt of colonised (M) plants and mean weights of uncolonised (NM) plants at each harvest: Eqn 1.
Mycorrhizal P responses (MPR) were calculated from Eqn 2
This was done first using the individual P content of colonised plants and mean P content of uncolonised plants and second using equivalent shoot P contents. The rationale here was that because an unknown quantity of root P in the colonised plants is in the fungus, use of whole plant P may provide misleading results in terms of benefit to the plant. This complication is avoided if MPR is based on shoot P.
Uptake of P per unit root d. wt, that is specific P uptake (SPU) was calculated using Eqn 3:
Where t = harvest time.
Specific P uptake between 0 and 6 wk used seed P content and total plant P at 6 wk. There were no roots at time zero.
Specific P uptake (SPU) between 6 and 9 wk was calculated using first the total plant P contents, and second shoot P contents, and individual and mean root d. wt. The rationale was again that root P would include P in intraradical AMF and inclusion might be misleading with respect to benefits to the plant.
ANOVA was performed on all data using Genstat 5 Release 4.1 (1998). Where significant differences were found, differences between treatment means were calculated using the least significant differences (LSD) method.
The percentage of the A. fistulosus root length colonised decreased with increasing soil P, from 59.1% (P1) to 40.6% (P5) and from 56.5% (P1) to 36.7% (P5) at the 6 and 9 wk harvests, respectively (Table 1). Addition of P to the soil did not have a significant effect on the percentage of the colonised root with EH, HC or AC at either harvest. However, the ratio of the percentage of the colonised root containing HC to the percentage containing AC (HC : AC ratio) generally increased significantly with increasing P.
Table 1. Percentage of root length colonised by Glomus coronatum , and percentage of mycorrhizal root length containing EH, HC, AC and HC : AC ratio of Asphodelus fistulosus at different soil P concentrations at the 6 and 9 wk harvests. Means followed by the same letter are not significantly different at the P < 0.05 level. Valid statistical comparisons cannot be made between harvest times or between different colonisation features. ( n = 4)
6 wk colonisation
9 wk colonisation
HC : AC
HC : AC
At 6 wk, the SDW of the colonised plants grown at P1 was significantly greater than that of the uncolonised plants (Fig. 1a). The same trend was apparent for RDW, but not all differences were significant. At P2 to P5 the SDW and RDW of the colonised plants were smaller than the uncolonised plants, but the difference in RDW at P5 was not significant. The maximum SDW and RDW of the uncolonised plants was achieved at P2. Between P2 and P5 the difference between the SDW (and also RDW) of the uncolonised and colonised plants decreased.
Between the 6 and 9 wk harvests the SDW and RDW of the uncolonised plants grown at P1 did not change, but the colonised plants increased in weight dramatically (compare Fig. 1a,b). At 9 wk, both SDW and RDW of the colonised plants grown at P1 were much larger than the uncolonised plants. At P2 to P5, the SDW of colonised and uncolonised plants were not significantly different from each other or from the colonised plants at P1. The RDW of uncolonised plants grown at P2 were higher than those grown at P1. There was no increase with any further P addition. There was no consistent trend in the RDW of the colonised plants. As at the 6 wk harvest, the increasing addition of P to the soil resulted in a decrease in the difference between the RDW of uncolonised and colonised plants at 9 wk. The data in Fig. 1 were used to calculate the MGR, using Eqn 1. At both the 6 and 9 wk harvests the MGR was positive at P1 and negative (although not always significantly less than zero based on LSD values) at P2 to P5 (Fig. 2).
At the beginning of the experiment the mean (± SE) seed P content of the plants was 23.0 (± 5.5) µg P per seed (6.4 mg P g−1 seed). Shoot and root P concentrations generally increased with increasing P supply, although differences between P treatments were not always significant (Fig. 3). At 6 wk colonised and uncolonised plants had similar shoot P concentrations, except P5, where the concentration in colonised plants was higher (Fig. 3a). Root P concentrations in colonised and uncolonised plants were the same at P1; from P2 to P5 those of colonised plants were higher, although the difference decreased with increasing P addition.
At the 9 wk harvest, the P concentrations in the shoots of colonised plants at P1 were lower than those of uncolonised plants (Fig. 3b). At P2 and P3 the shoot P concentrations in the colonised plants were higher than in the uncolonised plants, but with further addition of P to the soil there were no significant differences. The root P data at this harvest were very similar to those observed at six weeks in both concentrations and trends.
P contents of the tissues (i.e. the products of d. wt and P concentrations) were used to calculate the MPR for total plant P content (Fig. 4a) and shoot P content (Fig. 4b), using Eqn 2. Both methods showed that MPRs were significantly higher at lower soil P concentrations compared with the higher soil P concentrations. The MPR was greatest at the 9 wk harvest at P1 and in most cases was higher at the 9 wk harvest compared with the 6 wk harvest.
Between 0 and 6 wk the SPUs of colonised and uncolonised plants were not significantly different at P1, but between P2 and P5 colonisation significantly increased SPU. SPU's of uncolonised plants increased with increased soil P until a plateau was reached at P4, whereas SPUs of colonised plants increased steadily with increased soil P (Fig. 5a). Between 6 and 9 wk, SPU's of uncolonised and colonised plants at P1 and P2 were not significantly different regardless of the method of calculation (i.e. using shoot P or total plant P), although there was a clear trend towards greater SPU in colonised plants (Fig. 5b,c). At P3 to P5 the SPUs of the colonised plants were significantly greater than those of the uncolonised plants when total plant P was used but not when shoot P only was used.
Overall the results show that the development and function of Paris-type AM formed between A. fistulosus and G. coronatum has many similarities to Arum-type symbioses, for which much information is available. There was a positive growth response at low soil P and mycorrhizal colonisation had clear effects on plant growth and P nutrition at higher P levels. This contrasts with findings in tomato with two other AM fungi (see Introduction) and shows that some Paris-type AM can be highly effective. We did not include an Arum-type symbiosis in this experiment, but closely similar work (using the same soil, fungal isolate, etc.) has been carried out with Allium porrum (Pearson et al., 1991; Smith et al., 1994; Dickson et al., 1999), a plant with generally similar life-form to A. fistulosus.
In A. fistulosus, colonisation was moderately sensitive to P supply (reduced from 60% to just over 30% over the range of P used). This contrasts with Al. porrum colonised with the same fungus in which a reduction of 10% in percentage of the root length colonised was achieved with P addition equivalent to P1 to P2 concentrations used here (Dickson et al., 1999). The reductions in both A. fistulosus and Al. porrum were much less than in plants in which colonisation is highly P-sensitive, such as cereals (e.g. Baon et al., 1992). P supply had little effect on the proportion of the colonised root length with external hyphae, hyphal coils or arbusculate coils when these features were considered separately. However, the increase in the ratio of hyphal coils to arbusculate coils (Table 1), suggests a possible influence of P on the formation of the arbuscules (relative to hyphal coils), as has been observed for typical Arum-type arbuscules in many (but not all) plant/fungus combinations (Braunberger et al., 1991; Pearson et al., 1991; Bruce et al., 1994; Dickson et al., 1999). The function of hyphal coils and arbusculate coils is not known. However, as they provide interfacial areas as large as those of arbuscules there is no a priori reason why they should not play a role in nutrient transfer (Dickson & Kolesik, 1999).
A. fistulosus was highly responsive to formation of Paris -type mycorrhizas when soil P supply was low (P1). At this P level, uncolonised plants remained small throughout the experiment, whereas colonised plants grew well and showed positive MGRs and MPRs at both harvests. At 6 wk the MGR at P1 was less than that reported for Al. porrum colonised by the same fungus under the same experimental conditions (MGR = 112%) ( Dickson et al., 1999 ). The difference may be due to the slower growth of A. fistulosus or slower rate of colonisation compared to Al. porrum , or a combination of both. P concentrations in both shoots and roots were similar in the two groups of plants at P1 ( Fig. 3 ). Those in uncolonised plants remained unchanged between 6 and 9 wk and growth ceased ( Fig. 1 ). Mycorrhizal roots at P1 did not accumulate high concentrations of P compared with uncolonised roots, which probably reflects rapid transfer from the fungal compartment to the plant, supporting increased shoot growth. The results suggest that the positive responses in A. fistulosus at P1 in terms of growth and P uptake were the result of increased efficiency of the colonised root systems to absorb P. Although there were no significant differences in SPU at P1, there was a trend towards higher values in mycorrhizal plants; SPU's for the uncolonised plants were negative between 6 and 9 wk, regardless of method of calculation. It can be concluded that non-mycorrhizal A. fistulosus roots are very inefficient at acquiring P from low soil concentrations and in consequence are highly dependent on their Paris -type mycorrhizal symbiosis.
At higher levels of P supply (P2–P5) the results are more complex. At 6 wk, d. wt of both roots and shoots of mycorrhizal plants were lower than equivalent uncolonised plants and there were no marked changes in shoot growth associated with mycorrhizal colonisation; maximum shoot growth of colonised plants was achieved at P4 and for uncolonised plants at P2. At 9 wk SDW of colonised and uncolonised plants were very similar, but RDW were higher in uncolonised plants. These differences highlight the contrasting patterns of C allocation to nutrient absorbing tissue in the two groups of plants (roots in uncolonised and fungus in mycorrhizal plants). However, the colonised roots were considerably more efficient in P absorption, with SPU's generally greater than those of uncolonised roots, regardless of method of calculation (Fig. 5 and below). P accumulation in roots and shoots also showed differences between uncolonised and colonised plants (Fig. 3). Again over the range P2–P5, the shoot P concentrations increased, with those of colonised plants either equal to or greater than those of the uncolonised plants. At these higher soil P concentrations, A. fistulosus appears to be able to obtain sufficient P to achieve maximum growth regardless of mycorrhizal colonisation. The progressive increases in P concentration suggest that a factor or factors other than P availability limited shoot growth for both groups of plants, particularly for the mycorrhizal ones. Reduction in root growth may reduce acquisition of nutrients that do not enter by the mycorrhizal pathway, and lead to growth limitations and accumulation of P, as observed here. However, a similar, although smaller effect was observed in uncolonised plants and further investigation is required to determine the underlying cause. In any event, the fungus may have become superfluous with respect to P uptake in the short term, and potentially a cost to the plant. This is consistent with the suggestion of Johnson et al. (1997), that increasing availability of a limiting soil resource can convert balanced mutualistic relationships into less balanced ones, of which some can be parasitic.
The negative MGR (at P2 to P5) may in part be due to a C drain to the fungus and can be considered a cost of establishment of the symbiosis. Again, similar effects have been found with Arum-type mycorrhizas (e.g. Dickson et al., 1999). The percentage of the root length colonised was relatively low compared with other studies where C drains have been observed (Graham & Eissenstat, 1998). However in Paris-type AM the fungal biomass per unit root length of colonised root is likely to be much higher than in Arum-types because of the formation of extensive coils in consecutive cells. In Arum-type AM, by contrast, arbuscules are frequently more scattered, occurring in every second or third cell. Furthermore Dickson & Kolesik (1999) showed that for representative arbuscules and coils of the same surface area, the volume of the coils was much greater. In consequence, fungal biomass (and hence C drain) per unit root length is likely to be greater in Paris-type than in Arum-type AM and requires further investigation. Calculation of MPR and SPU were both useful in interpreting the results in terms of effects of mycorrhizal formation and P supply on A. fistulosus. Similar trends were observed regardless of whether the calculation was based on whole plant P or shoot P (Figs 4 and 5). This confirms that the fungi confer a benefit in terms of more efficient uptake from the soil and transfer to the shoots. The roles of the extensive hyphal coils in Paris-type AM are not known. They may be sites of nutrient transfer (see above); however, given their large volume and their long life span, they may also provide important organs for the storage of P and C. The data presented here suggest that P is sequestered (stored) in the coils, which constitute a major proportion of the fungal biomass. In a perennial plant storage over periods when the plant is not actively growing, and subsequent release of the P to the plant when it starts to grow, might give the plant a ‘head start’ over the other plants in the community. Increased success of the plant may ultimately benefit the fungus. In southern Australia, sites invaded by A. fistulosus are often unfavorable in terms of plant growth, thus successful formation of symbiotic AM may provide a competitive advantage over invasive nonmycorrhizal species colonising these sites (see O’Connor et al., 2002).
The results presented here suggest that in soils with low P, A. fistulosus will benefit from forming AM both in early and later stages of growth, and colonisation may in fact provide an essential advantage in terms of competition and success over several seasons. Conversely, in higher P soils, A. fistulosus may benefit only in later stages of growth and possibly only if a further stress is imposed upon the plant. The delay in apparent benefit may be of particular importance given that A. fistulosus is a perennial herb. As discussed, the longer term effects of colonisation need to be considered, in particular the roles of coils in AM. A. fistulosus is a useful plant to study the physiology and ecology of AM (particularly Paris-type AM) given its large change in growth with and without colonisation by AMF over a narrow range of soil P concentrations (P1 and P2).
Our work is funded by the Australian Research Council. TRC held an Australian Postgraduate Award. These data were presented in a preliminary form at the Third International Conference on Mycorrhizas (ICOM3), July 2001, Adelaide, Australia. We wish to thank participants for valuable discussions on this work. We also wish to thank Mr S. Anstis for performing seed P determinations and Ms V. Cavagnaro for valuable discussions. Finally, we wish to thank the two anonymous referees of the paper for their valuable comments.