Chronology and mechanisms of P uptake by mycorrhizal sweet potato plants

Transcription

1 New Phytol. (992), 22, Chronology and mechanisms of P uptake by mycorrhizal sweet potato plants BY D. M. O'KEEFE AND D. M. SYLVIA Soil Science Department, University of Florida, Gainesville, FL , USA (Received 22 July 99, accepted 28 July 992) SUMMARY Vesicular-arbuscular (VA) mycorrhizal and non-mycorrhizal sweet potato plants [Ipomoea batatas (L,S) Lam, cv. White Star] were grown in a glasshouse in a phosphorus (P)-fixing soil in order to (i) establish the chronology of mycorrhiza-mediated P uptake, (ii) document the distribution of total and metabolically active external hyphae in relation to roots, root hairs and P-depletion zones at the time when the P-uptake response first becomes apparent, (iii) evaluate the pore-size distribution of the soil relative to the radii of roots, root hairs and hyphae of mycorrhizal fungi, and (iv) determine if sweet potato mycorrhiza alter the organic acid composition of the rhizosphere. Phosphorus inflow, VA mycorrhizal root colonization, and active hyphae increased during the first half of an 8 wk experiment. During the latter half of the experiment, however, colonization and active hyphae increased while P inflow decreased. Roots colonized by mycorrhizal fungi supported a higher proportion of active hyphae in the soil than non-colonized roots. In a second experiment, hyphal distribution around individual roots colonized by mycorrhizal fungi was not consistently difl^erent from hyphal distribution around non-colonized roots. Hyphae of mycorrhizal fungi were of a diameter that would allow them to penetrate pores that hold water at water contents less than field capacity while roots, and to a large extent root hairs, would be excluded from these pores. We found no evidence that sweet potato roots, whether or not they were colonized by mycorrhizal fungi, altered the organic acid composition of the rhizosphere, W^e conclude that mycorrhiza mediated P uptake commenced vvith the onset of colonization by the fungi and was due to a change in the ability of the root to come into contact with the soil solution. Key words: Mycorrhiza, Ipomoea batatas (sweet potato), P inflow, organic acids, external hyphae, rhizal fungus and reported data that fit this chronology. Their finding, that proliferation of external Under P-deficient conditions, enhanced P uptake hyphae occurred d after planting, agrees with occurs when P inflow is greater for a tnycorrhizal the observations of Owusu-Bennoah & Wild (979), plant tban for a non-mycorrhizal plant, assuming vvhere difterences in P-depletion zones between that mycorrbizal status is tbe only difference between mycorrbizal and non-mycorrhizal onion roots octbe plants (Sanders & Tinker, 973), Proposed curred within 2 d. Sanders et al (977) reported mechanisms for mycorrhiza-mediated P uptake have that P inflow was higher in mycorrbizal plants been recently reviewed (O'Keefe & Sylvia, 99), It compared to controls after 7 d, wbile weigbt is generally accepted that tbe hyphae from mycor- differences became apparent 30 d after planting, rhizal roots are distributed in sucb a way as to In that same experiment, and a similar experiment improve tbe P-uptake cbaracteristics of tbe root witb onions (Sanders & Tinker, 973), P inflow into (Sanders & Tinker, 97 ; Mosse, 973). mycorrbizal roots exceeded inflow into non-mycor- Tbe mycorrbizal effect may be the result of eitber rhizal roots tbrougbout tbe experiment, suggesting a transitory event or a continuous phenomenon, tbat in tbese short-term experiments tbe mycorrhizal Knowledge of tbe cbronology of mycorrhiza-medi- response was a continuous phenomenon, ated P uptake may sbed ligbt on its mechanisms. To Soil pore sizes vary, and as a result, access to determine tbe chronology, multiple harvests or nutrients is partly dependent on the radii of the repeated, non-destructive measurements are necess- nutrient-absorbing structures. Estimates of tbe radii ary. A proposed cbronology is infection of tbe root, of roots, root hairs and hypbae of mycorrbizal fungi followed by proliferation of external bypbae, leading are needed to evaluate bow tbese structures explore to increased P inflow, and resulting in increased P soil pores for P. uptake. Gueye, Diem & Dommergues (987) studied Plants can increase P solubility and uptake by tbe response of cowpea to colonization by a mycor- exudation of organic acids into tbe rbizospbere 44 ANP 22

2 652 D. M. O'Keefe and D. M. Sylvia (Jungk, 987; Marschner, Romheld & Cakmak, 987; Nye & Kirk, 987). Many fungi possess the capacity to alter P availability in the soil by the same mechanisms as bigber plants (Beever & Burns, 980). However, there is no direct evidence of quantitative or qualitative changes in exudation of organic acids from plant roots due to infection by mycorrbizal fungi. The objectives of tbis researcb were to: (i) establish the chronology of mycorrhiza-mediated P uptake by sweet potato; (ii) determine the radial distribution of total and metabolically active external bypbae of mycorrbizal fungi in relation to tbe root, root bairs and tbe P-depletion zone at the time when enhanced P uptake first becomes apparent; (iii) evaluate pore-size distribution in tbe soil relative to the radii of roots, root hairs and hyphae of mycorrhizal fungi; and (iv) determine if sweet potato mycorrhiza alter the organic acid composition of tbe rhizosphere. M.^TERIALS AND METHODS Experiment : Chronology of the response Mycorrhizal and non-mycorrbizal sweet potato plants [Ipomoea batatas (L.) Lam. cv. Wbite Star] were grown from cuttings in a glasshouse in a phosphorus-fixing soil in order to establish a three (VA mycorrbizal inoculation treatments) by four (harvest date) factorial experiment. The experiment employed a completely randomized design with five replications per treatment. The VA mycorrbizal inoculation treatments consisted of two VA mycorrbizal isolates and a non-inoculated control. The mycorrhizal inocula consisted of soil, roots and spores from pot cultures of Glomus etunicatum Becker and Gerdemann [International Culture Collection of VA Mycorrhizal Fungi (INVAM) LETC26] and Acaulospora rugosa Morton (INVAM 98), maintained on Desmodium etunicatum D.C. growing in Arrendondo loamy sand (Grossarenic Paleudult). To inoculate plants, a 2 cm diameter by 4 cm deep core was removed from the centre of each pot and a 5 g pad of inoculum (from 2-wk-old pot cultures stored at 4 C) was placed in the bole and covered with the soil from the original core. In an attempt to standardize background microbial populations, a suspension - consisting of 20 g of pot-culture material from both isolates in 2 of tap water - was passed through a 0 mm mesh sieve and then five times through a 0-22 //m mesh sieve and applied, at a rate of 0 ml pot \ to all pots. prath, 970) it was determined that the soil had a high P-fixing capacity (784 /ig g"^) and a low native soil solution equilibrium P level (0-005 /ig ml"^). Phosphorus was added to the soil (37 mg kg"') resulting in soil solution-equilibrium P concentration after 6 d of 0-0 /ig m\~^. This P concentration had been shown previously to result in mycorrhizaenhanced P uptake by sweet potato (O'Keefe, 989). Pots (6 cm diameter) were filled with 3-5 kg of pasteurized soil, and 20 ml of a nutrient solution containing 2-9 g T^ NH.NOg, 4-3 gl'^ KCI and ml r ' of Hoagland's micronutrients (Hoagland & Arnon, 950) was added to each pot. Every 2 wk a further 0 ml of solution was added to each pot. Pots were watered to maintain approximate field capacity. Plants were grown during the summer in a glasshouse with average maximum PAR over the 8 wk growing period of 746/<E m'^ s"\ and average maximum and minimum daily temperatures of 32 and 24 C. Soil solution P concentration was sampled at 2 wk intervals in two pots from each inoculation treatment by lysimeters made from polyvinyl chloride (PVC) tubing (internal diameter 22 mm) cut to 6 cm lengths and closed at the bottom with sintered glass discs. The top end was closed with a rubber stopper fitted with a glass tube, vacuum hose and clamp. The solutions were extracted by a band-held vacuum pump, removed from the lysimeter with a pipette, and kept at 4 C with 3 drops of toluene until the completion of the experiment, when the solutions were analysed for P (Murphy & Riley, 962). Plants were harvested 2, 4, 6 and 8 wk after transplanting stem cuttings. At each harvest the following plant parameters were measured or calculated: shoot and root weight, shoot length, P concentration in shoots, roots and tbe first fully expanded leaf and petiole, P-infiow rates, percentage and total length of root with root hairs, root hair length, diameter and density, percentage and total length of root colonized by mycorrhizal fungi, and density and total number of entry points of mycorrhizal hyphae. Phosphorus content in shoot cuttings was measured at the beginning of the experiment. Phosphorus inflow was calculated for eacb plant by dividing total P uptake over a harvest interval by root length at the end of tbat interval. Shoot lengths were measured twice weekly. Root and root hair diameter were determined as the mean of ten random measurements on each of five cm long segments of roots with root hairs per plant. Total root fresh weight was determined and a 0-5 g subsample was collected for observation of root hairs and VAM colonization using tbe gridline-intersect method (Giovanetti & Mosse, 980). Screened (2-0 mm) soil from the Bt horizon of an Orangeburg series profile (Typic Paleudult) was used in the experiment. It bas tbe following At each harvest, the diameters of external hyphae characteristics after steam sterilization (75 C for 4 b): organic carbon, 0-04" ; sand, 7 0 % ; clay, were measured using an eyepiece micrometer at 7"o ; ph in H^O, 5-3; and Mehlich I-extractable P, 600 X magnification. Five cm long root segments 3-2 jug g \ From P adsorption curves (Fox & Kam- colonized by VAM fungi were examined per plant.

3 P uptake by sweet potato five measurements per segment were made, and values were averaged for each plant. The lengths of total and active hyphae in the soil were estimated on two samples taken from each pot: core A was taken directly over the plant stem and inoculum; core B consisted of three pooled cores taken 5 cm from the central core. Cores were taken to the bottom of the pot using a sharp, 20 mm diameter stainless steel corer. Each sample (whole core for A and an equivalent amount from B) was suspended in 5 of tap water, blended for 5 s, allowed to settle for 5 min, and than a 25 ml subsample was removed and placed on filter paper. Hyphae were treated with indonitrotetrazolium as the active stain and trypan blue as the counter stain and quantified using a gridline-intersect method (Sylvia, 988). Coarse, brown, septate hyphae were not included in hyphal counts. An additional study was done to test the effect of blending and settling time on the length of hyphae measured. Orangeburg soil was placed in 7 x 2 cm plastic trays to a depth of 3 cm and amended with 500 ml of a -0 gl"' malt extract solution to allow saprophytic fungal populations to increase. After incubation at 2 C for 4 wk, 25 g soil samples were processed as above except that the blending (5, 5 and 60 s) and settling (0, 0 5, 3, 5 and 0 min) times were varied. There were five replications per treatment combination. Experiment 2 : Hyphal distribution and organic acid production A second experiment was established using the same treatments and experimental design as the first experiment, except that there were only two harvests (2 and 4 wk) and P was applied 6 wk before transplanting. The plants were grown during the fall in the same glasshouse used for the first experiment. Mean daily maximum and minimum temperatures were 29 and 23 C and mean daily maximum natural PAR was 62//Em-s"' supplemented with metal halide lamps supplying 670//Em'^s~^ for 4 h d '. Shoot weight and P concentration, and P concentration of the first fully expanded leaf and petiole were determined at both harvests. Complete root systems were recovered at each harvest, except for portions contained in soil cores (coring procedure described below). Up to four 0-5 g samples were collected randomly from each root system, and these were assessed for total root length, root length and percentage colonized by mycorrhizal fungi, and percentage of root length with root hairs. Because the entire root system was used in this way, P inflow values were based on shoot P content alone. Soilsolution P concentrations were determined at 0, 2 and 4 wk. At the 4 wk harvest, four cores were removed from each pot, prior to disruption of the root system, in 653 Distance from the root (//m) Root = i z^ =d = =d =) =d Root hairs- A 30 Ei Figure. Calculated P-depletion zones at 7 days, root hairs, and zones used to determine hyphal distribution. The three broken lines represent the following: (A) extent of the root hair cylinder; (B) the extent of the wk Pdepletion zone excluding the effect of root hairs; and (C) the extent of the wk P-depletion zone assuming root hairs increase the effective radius of the root. The solid lines represent distances used to determine the distribution pattern of soil hyphae, the same pattern as in the first experiment. Cores were collected using 22 mm diameter, 8 cm long bevelled PVC tubes. Cores were impregnated with a 50 g r^ gelatin solution that contained 25 ml T^ of a trypan blue stain [trypan blue (0-5 g T'), H^O (500 ml r^), lactic acid (70 ml T^) and glycer'in (330 ml r^)]. Cores were packed at both ends with polyester fibre, rubber stoppers with glass tubes were placed in each end of the core, and the cores W'ere then placed upright in an incubator at 34 C. The gelatin solution was placed above the incubator on a hot plate and allowed to fiow through the tubes connected to the bottom of the cores. When the colour of the solution coming out of the top of the cores was the same as that going in, the cores were clamped at the bottom, removed from the incubator, and solidified at 4 C. Impregnated soil was pushed out of the PVC tubes in 5 mm increments using a tight-fitting wooden dowel and sectioned with a razor blade. Sections were inspected with a dissecting microscope and those with roots on the surface were examined at 60 x using a compound microscope. Hyphae adjacent to individual roots were quantified using an eyepiece whipple disc by the gridline-intersect method (Newman, 966). The grid was divided into five 222 /im wide zones (Fig. ), The root was aligned with the edge of the whipple disk and hyphae were quantified in each zone. Values from the three outer cores were averaged to give one value per zone per plant, and these values were analyzed separately from values obtained from the central core. 44-2

4 654 D. M. O'Keefe and D. M. Sylvia The soil that adhered to roots when they were removed from the pots after 4 wk was designated rhizosphere soil and the remainder of the soil was designated bulk soil. Rhizosphere soil was removed from the roots by dipping them into deionized water. The suspensions were then shaken gently in plastic hottles containing 00 ml of deionized water at room temperature for 2 h, filtered (0-2 /im), and frozen ( 20 C). The residual soil on the filter membrane was dried at 70 C for 24 h and weighed to adjust values to a dry wt basis. Filtrates were thawed and analysed for organic acids using high-pressure liquid chromatography (HPLC) (Lee & Lord, 987). Organic acids were separated at ambient temperature with H,^SO4 as the eluent at a flow rate of 2 ml min~^ and detected at 20 nm. Organic acids were quantified by comparing peak areas with external standards. Soil pore-size distribution was determined using water-release curves that relate the volumetric water content at varying pressure to soil pore size (Swanson & Peterson, 94). Bulk densities were determined by tbe method of Blake & Hartge (986). Soil water content was determined gravimetrically at each harvest, averaged and adjusted to volumetric water content {0). Volumetric water content was used to calculate P diffusion coefficient in soil D^, = D^0fb~^ () external byphae, and byphal density values, were correlated positively witb tbeir variance and were tberefore power transformed according to the methods of Glenn, Brown & Takeda (987). Exponents used in tbese transformations were 0-45 for hypbal density, 0-30 for total hypbal length and 0-0 for active hyphal length. E 03 o CO Time (wk) Figure 2. Shoot length of sweet potato grown in a glasshouse for 8 wk inoculated with Acaulospora rugosa (#), Glomus etunicatum (V) and uninoculated (O)- Differences {P ^ 0-05) at a given date between the inoculated treatments and the control are indicated by an asterisk. Symbols represent means of five replicates. where Z)^^ is the P diffusion coefficient in water (Schenk & Barber, 979), buffer power (b) was calculated from P adsorption curves (Barber, 984), and tbe tortuosity factor (/) was taken from Warncke 6 Barber (972) as 0-5, Tbe widtb of P depletion zones around the root and root-hair cylinder after 7 d were calculated to be 338 and 468 //m, respectively, from x2dj (2) D) O o CD o E E_ -^^ (D) V (Baldwin, Tinker, Nye, 972). Data analysis Data were subject to analysis of variance using tbe General Linear Models Procedure (SAS Institute, Inc., 985). Data baving a significant barvest or harvest x inoculation effect were subjected to regression analysis. Significant inoculation effects were analysed by single degree of freedom contrast procedures tbat compared the effect of inoculation with each mycorrhizal fungus to tbe control. Data for root lengtb, lengtb of root witb root bairs, root bair lengtb, root hair density and total length of root colonized by mycorrbizal fungi were log-transformed, wbile tbe percentage of the root with root bairs and percentage of tbe root colonized by mycorrbizal fungi were square-root-transformed prior to analysis according to Gomez & Gomez (976). Means for tbe length of active and total o o JZ o cn Q- E Ib , - D- 2 - O -- S "ra - ^ ' (E) 0 / \ \ E " \ - -,y^ o \ 06 - V Time (wk) / \ - \ ' Figure 3. Effect of Acaulospora rugosa (#) and Glomus etunicatum (V) on (A) root and (B) shoot P content, (C) shoot P concentration, (D) first fully expanded leaf and petiole P concentration, and (E) P inflow to sweet potato grown in a glasshouse for 8 wk. Symbols represent means of five replicates except for the first harvest, where all inoculated plants had three replicates and the second harvest, where A. rw^osa-inoculated plants had four replicates. Regression equations are given in Table, (O) uninoculated controls.

5 P uptake by sweet potato 655 Table. Regression equations and coefficients of determination (r^) showing the relationship of harvest date (wk) to root and shoot P content (mg), shoot P concentration (mg g~^ J, first fully expanded leaf and petiole P concentration (mgg~^) and P inflow (ng cm~^ d'^) of sweet potato grown in a glasshouse for 8 wk with two mycorrhizal fungi (^Acaulospora rugosa and Glomus etunicatumj Inoculation treatment Regression equations n r^ A. rugosa G. etunicatum A. rugosa G. etunicatum A. rugosa G. etunicatum A. rugosa G. etunicatum A. rugosa G. etunicatum Root P content y=0-26 J^H-0-26 y=0-65 X-0-7 y=0-42 X-0-04 Shoot P content y = 0-5 A'---52 y=0-66 X-l-0-39 y = -38 X-2-48 Shoot P concentration y = -0-3 X+\-6\ y = J^---46 y=0-09 X + Q-b\ First fully expanded leaf and y=-0-2.y-i-2-93 y=-0-30x y=-0-5 X-l-3-3 P inflow y=-0-3 X -F ^--9-6 y = 0-6 X ^-0-8 y = 0-5 X X'-Q-l petiole P concentration RESULTS Soil solution P levels Mean soil solution P levels in tbe first experiment were initially 0-05//g mr\ dropped to 0-Q2 figm\'^ at tbe 2 wk barvest, and tben stabilized at 0-0 fig ml~\ Soil solution P levels in the second experiment were stable throughout the experiment at fig m\~^. The lower values in the second experiment were probably due to the 6 wk equilibration period. Experiment : Chronology of the response Shoot (y= 0-7 X-0-46, P = 0-000, 7?^ = 0-7) and root (y=0-67 X-0-52, P = 0-000, i?^ ^ 0-69) weights increased linearly with harvest, but inoculation had no affect on the rate of increase. Differences in shoot length between controls and plants inoculated with A. rugosa and G. etunicatum appeared at 3-5 and 6 wk respectively (Fig. 2). Root (Fig. 3 a) and shoot (Fig. 3 b) P increased more rapidly in mycorrhizal plants than in control plants (Table ). Shoot P concentration increased in plants inoculated with G. etunicatum, but decreased in plants inoculated with A. rugosa and control plants (Fig. 3 c). Phosphorus concentration in the first fully expanded leaf and petiole decreased with harvest in all inoculation treatments (Fig. 3d). Inoculation with mycorrhizal fungi gave higher concentrations of P in roots (P ^ 0-0) than in control plants (means of 684yWg P g^^ in control, 852 /ig P g"^ in A. ru^osa-inoculated, and 736//g P g"^ in G. etunicatum-inoculated plants). Phosphorus infiow increased for 6 wk to plants inoculated with fungi, while P inflow into control plants decreased with time (Fig. 3e). Mean (±SD) root hair density ( mm '^), diameter ( /tm), length (3±8//m), and root diameter ( //^m) were not affected by inoculation treatment or harvest. The percentage of the root length with root hairs decreased linearly with time (Y = 0-4 X+6-47, P = 0-000, R^ = 0-42), but was not affected by inoculation. Colonized root length (total and percentage) increased with time, but there was no difference between the two mycorrhizal fungi; mean colonization at the first harvest was 4 % and this increased linearly {Y = 0-55 ^+4-25, P^O-02, i?' = 0-7) to a maximum of 8 "o at 8 wk. There was no significant difference between the two inoculated treatments in density of entry points cm^^ of colonized root (3 + 0). There was a linear increase (y= X+90000, P = 0-0, R^ = 0-20) in the total number of entry points with time. Hyphal diameter of both mycorrhizal fungi was jurn. The densities of total hyphae in core A

6 656 D. M. O'Keefe and D. M. Sylvia hyphae recoverable at 3 min were observed. With settling times less than 3 min, suspended clay collected on the filter and interfered with observation of the hyphae. 4 5 Time (wk) 7 8 Figure 4. Effect of time on (A) active hyphal density in core A, and (B) percentage of hyphae that were active in cores A (O) and B (#) in pots containing sweet potato growing in a glasshouse. Symbols represent means of replicates for the first harvest, 4 for the second harvest, and 5 for the last two harvests. Regression equations and R^ values are: (A) y = ;*(' + O-337, R- = 03; (B) core A, Y= 7-67 Js:+9-55, R' = 0-9, core B, Y = 245 A' , J^- = 0-4), ( cm cm"'^) and B ( cm cm"'') were not affected by inoculation or harvest. However, the density of active hyphae increased linearly with time (Fig, 4a). In core A, the mean proportion of active hyphae was 0-24 for the control, 0-73 for A. rugosainoculated, and 0-52 for G. etunicatum-inocu\ated plants. The proportion of active hyphae also increased linearly with time in both Core A and B (Fig, 4b). The estimates of length of external hyphae were affected by settling time, but not by blending time. After a settling time of 5 min, only 56% of the Experiment 2 : Hyphal distribution and organic aeid production At tbe 2 wk harvest, mean ( + SD) weight ( mg), total shoot P (322 ± 20//g), P concentration of the shoot ( //g P g^'), first leaf and petiole P concentration ( //g P g~'), total ( ) and colonized root length (5 + 6 cm), percentage colonization ( + '5) and P inflow ( ng cm~^) were not different between treatments. Phosphorus infiow was negative because only shoot P was used in the calculation. At the 4 wk harvest, there was no difference between treatments for weight, first leaf and petiole P concentration, and total and colonized root length with mean values of mg, 2-74±2-7 mg g"^ cm and cm, respectively. However, shoot P content and concentration, and P inflow rate were all higher in inoculated plants (Table 2), Colonization by G, etunieatum (29,o) was greater (P ^ 0-05) than that by A. rugosa (3%). There was no effect of distance from the root on the density of hyphae in core A ( cm cm"^). There w as a slight decline in hyphal density (Y = X + 0-4, P ^ 0-0, R^ -= 0-06) away from the root in core B, with mean values ranging from 0-34 cm cm " in the //m zone to 0-23 cm cm~^ in the 888-//rn zone. There were increases in hyphal density due to mycorrhizal inoculation in the outermost zone of core A (G. etunicatum 0-75 cm cm~^ compared to 0-06 cm cm"'^ for control) and in the middle zone in core B {A. rugosa 036 cm cm ' compared to 00 cm cm"^ for control). Oxalic acid {20 + \ 5 fig g~^) was the only organic acid detected in the soil extracts. There was no difference between rhizosphere and non-rhizosphere soil, nor between inoculated and non-inoculated plants. The pore-size distribution in the soil in relation to the average radii of hyphae, root hairs and roots Table 2. Effect of inoculation with mycorrhizal fungi on shoot P content and concentration, and P inflow to sweet potato after 4 wk in a glasshouse Inoculation treatment Shoot P Content (/yj;;) Concentration i/'g g"') P inflow (ng cm' d') Acaulospora rugosa Glomus etunicatum * ** 520** ** 2-2* Values are the mean of 5 replicates, ** and * indicate that the mean is significantly different from that of the uninoculated treatment at P ^ 00 and 0-05, respectively.

7 P uptake by sweet potato 657 -Radius of root hairs -Radius of VAM hyphae Pore radius (//m) Figure 5. The relationship between the size of the nutrient-absorbing structures and pore size distribution in Orangeburg soil. Bars represent the mean radius of roots, root hairs and mycorrhizal hyphae, 0, Water-filled pores at field capacity. is shown in Figure 5. The average volumetric water content during the experiment was cm^ cm"^. According to the relationship between volumetric water content and pore size outhned by Swanson & Peterson (94), pores greater than 20 /im radii were, on average, drained for the duration of the experiment. This means that when the soils were at field capacity roots were excluded from saturated pores while mycorrhizal hyphae and root hairs were not. DISCUSSION Root colonization, the density of active hyphae and P uptake increased during the first half of the 8 wk experiment. However, while colonization and density of active hyphae continued to increase, P inflow decreased in the latter half of the experiment. Furthermore, the rate of increase of P inflow early in the experiment was greater than the rate of increase of either colonization or density of active hyphae. This suggests that the relationship between P uptake, colonization and density of external hyphae is neither simple nor proportional. Plant demand for phosphorus may aflfect this relationship; however, this should not have been an important factor in our experiments as rate of P uptake did not decrease over the course of the experiment. A more likely explanation is that the presence of mycorrhizal hyphae altered the physical contact between the root and soil solution, thereby affording a qualitative advantage to the mycorrhizal plant. Mycorrhizal fungi could improve P uptake by roots by improving physical exploration of the soil pore space in two ways. Firstly, hyphae adhere to soil particles, which improves contact with the soil solution. While measuring hyphal density in gelimpregnated soil cores, we observed that fungal hyphae follow the contours of, and were in direct contact with, the soil particles. This can also be observed in Foster, Rovira & Cook (983). In contrast, roots and root hairs do not readily conform to soil particle surfaces. Secondly, hyphae can enter smaller soil pores than plant roots and root hairs. Pores with radii greater than 5 //m will generally be drained at fleld capacity (Marshall, 979). In the soil used in our experiments, roots would be excluded from all undrained pores at field capacity, while both root hairs and hyphae of mycorrhizal fungi would have direct access to undrained pores. As soil water content decreases, the relative advantage of hyphae over root hairs increases. Within these undrained pores, P availability should be higher than in drained pores due to the positive relationship between volumetric water content and soil P diffusion rates (Schenk & Barber, 979). Contact and access alone may not fully explain why P inflow was higher to roots with root hairs and mycorrhizal hyphae than to roots with root hairs alone. Hyphae were approximately one-third the diameter of root hairs, but P inflow was increased more than tenfold. It is likely that the increase is due to a combination of: (i) the radial extension of hyphae of mycorrhizal fungi; (ii) the nature of the contact between hyphae and the soil surface; and (iii) entry of hyphae into smaller pore spaces than roots and root hairs. Proliferation of mycorrhizal hyphae may not increase P uptake in a proportional fashion. The physical contact between soil and roots appears fundamentally to change following initial infection by the fungi. It follows that early, but low, infection levels with correspondingly low densities of external hyphae may have a proportionally greater impact on P uptake than later, higher infection levels with higher densities of external hyphae. It is possible that our methods were not sensitive enough to distinguish small but significant levels of hyphae in the soil. The amount of hyphae recovered was less than reported in other studies (Sylvia, 990). An unknown amount of hyphae was lost during settling in the first experiment; however, the direct observations in the second experiment also gave low values. This study supports the contention of Sylvia (988) that active hyphal counts are more useful than total ones. Other researchers have observed that mycorrhizal development and increased P inflow precede extensive hyphal proliferation. Sylvia (988) observed that proliferation of active external hyphae followed increased root colonization by several weeks. Sanders et al. (977) observed that P inflow increased while colonization by mycorrhizal fungi and proliferation of external hyphae were still relatively low. This is not to suggest that external hyphae are not required for enhanced P uptake, but that small quantities of hyphae can significantly improve P uptake. The mycorrhizal response was continuous, not

8 658 D. M. O'Keefe and D. M. Sylvia transitory, for the duration of this experiment. After it first became apparent, P inflow to plants colonized with mycorrhizal fungi was consistently greater than to non-inoculated control plants. This is not to suggest that the response would be continuous if the experiment continued beyond the 8 wk period. In fact, P inflow was decreasing in the inoculated treatments when the experiment ended. Our attempt to determine the radial distribution of external hyphae relative to the calculated depletion zones surrounding roots revealed no significant pattern. Other researchers (Hattingh, Gray & Gerdemann, 973; Rhodes & Gerdemann, 975; Alexander, Alexander & Hadley, 984) worked on a scale of centimetres when addressing the question of how far the hyphae of mycorrhizal fungi extend from the root. At that scale, inter-root distances would result in overlap among hyphae radiating from roots of most crop plants. Phosphorus-depletion zones around roots occur on the scale of micrometres to millimetres, and this may be the most appropriate scale for studying hyphal distribution. Nonetheless, only in the outer cores, which should have had a higher proportion of young roots than the central cores, was there a detectable decline in hyphal density away from individual roots. It may not be possible to determine hyphal distribution around single roots since, even at the micrometre scale, hyphae from different roots will overlap, obscuring distribution patterns. ACKNOWLEDGEMENTS The authors thank N. B. Comerford for performing HPLC analysis. This work was partly sponsored by the Gas Research Institute as part of a cooperative programme of research with the Institute of Food and Agricultural Sciences, University of Florida, Gainesville. Published as Florida Agricultural Experimental Station journal series no. R REFERENCES R, C, ALEXANDER, I. J, & HADLEY, G. (984). Phosphate uptake by Goodyera repens in relation to mycorrhizal infection. Nezv Phytologist 97, -4. BALDWIN, J. P., TINKER, P. B. H. & NYE, P. H. (972). Uptake of solutes by multiple root systems from soil. II. The theoretical effects of rooting density and pattern of uptake of nutrients from soil. Plant and Soil 36, BARBER, S. A. (984). Soil Nutrient Bioaraitabitity. John Wiley and Sons, Inc., New York. BEEVER, R. E. & BtRNS, D. J. W. (980). Phosphorus uptake, storage and utilization by fungi. Advances in Botanical Research 8, BLAKE, G. R. & HARTGE, K. H. (986). Bulk density. In: Methods of Soil Analysis: Part, Physical and Mineralogical Methods, 2nd Edn (Ed. by A. Klute), pp Xmerican Society of Agronomy, Inc., Madison, Wisconsin. FOSTER, R. C, ROV;RA, A. D. & COOK, T. W. (983). Ultra- structure of the Root-soil Interface. The.American Phytopathological Society, St Paul, Minnesota. Fox, R. L. & KAMPRATH, E. J. (970). Phosphate sorption isotherms for evaluating the phosphate requirements of soils. Soil Science Society of America Proceedings 34, GiovANNETTi, M. & MossE, B. (980). An evaluation of techniques for measuring VA mycorrhizal infections in roots. New Phytologist 84, GLENN, D. M., BROWN, M. W. & TAKEDA, F. (987). Statistical analysis of root count data from minirhizotrons. In: Minirhizotron Observation Tubes-. Methods and Applications for Measuring Rhizosphere Dynamics. ASA Special Publication 50 (Ed. by H. M. Taylor), pp American Society of Agronomy Inc., Madison, Wisconsin. GOMEZ, K. A. & GOMEZ, A. A. (976). Statistical Procedures for Agricultural Research with Emphasis on Rice. The International Rice Research Institute, Los Banos, Laguna, Philippines. GuEYE, M., DIEM, H. G. & DOMMERGUES, Y. R. (987). Variation in N.^ fixation, N and P content of mycorrhizal Vigna unguiculata in relation to the progressive development of extraradical hyphae of Glomus mosseae. MIRCEN Journal 3, HATTINGH, M. J., GRAY, L. E. & GERDEMANN, J. W. (973). Uptake and translocation of ^'^P-labeled phosphate to onion roots by endomycorrhizal fungi. Soil Science 6, HOAGLAND, D. R. & ARNON, D. I. (950). The Water-culture Method for Growing Plants without Soil. California Agricultural Experiment Station, Berkeley, California. JtNGK, A. (987). Soil-root interaction in the rhizosphere affecting plant available phosphorus. Journal of Plant Nutrition 0, 97. LEE, D. P. & LORD, A. D. (987). A high performance phase for the organic acids. Liquid Chromatograph-Gas Chromatographv 5, MARSCHNER, H., ROMHELD, V. & CAKMAK, I. (987). Root- induced changes of nutrient availability in the rhizosphere. Journal of Plant Nutrition 0, MARSHALL, T. J. (979). Soil Physics. Cambridge University Press, Cambridge. MOSSE, B. (973)..Advances in the study of vesicular-arbuscular mycorrhizae. Annual Review of Phytopathology, MURPHY, J. & RILEY, J. P. (962). A modified single-solution method for the determination of phosphate in natural waters. Analytica Chimica Acta 27, NEWMAN, E. I. (966). A method for estimating the total length of root in a sample. Journal of Applied Ecology 3, NYE, P. H. & KIRK, G. J. D. (987). The mechanism of rock phosphate solubilization in the rhizosphere. Plant and Soil 00, O'KEEFE, D. M. (989). Mechanisms of enhanced phosphorus uptake by VA mycorrhizal sweet potato. Ph.D. Dissertation, Llniversity of Florida, Gainesville, Florida, U.S.A. O'KEEFE, D. M. & SYLVIA, D. M. (99). Mechanisms of the vesicular-arbuscular mycorrhizal plant-growth response. In: Handbook of Applied Mycology, vol. ; Soil and Plants (Ed. by D. K. Arora et al.), pp Marcel Dekker, Inc., New York. OWUSU-BENNOAH, E. & WILD, A. (979). Autoradiography ofthe depletion zone of phosphate around onion roots in the presence of vesicular-arbuscular mycorrhiza. New Phvtologist 82, 33. RHODES, L. H. & GERDEMANN, J. W. (975). Phosphate uptake zones of mycorrhizal and non-mycorrhizal onions. New Phytologist 75, SANDERS, F. E. & TINKER, P. B. (97). Mechanism of absorption of phosphate from soil by endogone mycorrhizas. Nature 233, SANDERS, F. E. & TINKER, P. B. (973). Phosphate fiow into mycorrhizal roots. Pesticide Science 4, SANDERS, F. E., TINKER, P. B., BLACK, R. L. B. & P.^LMERLEY, S. M. (977). The development of endomycorrhizal root systems. I. Spread of infection and growth-promoting effects with four species of vesicular-arbuscular endophyte. New Phytologist 78, SAS Institute Inc. (985). SAS User's Guide: Statistics. Version 5. SAS Institute, Cary, N.C., USA. ScHENK, M. K. & BARBER, S. A. (979). Phosphate uptake by corn as affected by soil characteristics and root morphology. Soil Science Society of America Journal 43, SwANSON, C. L. W. & PETERSON, J. B. (94). The use of the

9 P uptake by sweet potato 659 macrometric and other methods for the evaluation of soil external hyphae of vesicular-arbuscular mycorrhizal fungi. In: structure. Soil Science 53, 73-85, Rhizosphere Dynamics (Ed, by J, E, Box and L, C, Hammond), SYLVIA, D, M, (988), Activity of external hyphae of vesicular- pp, 44-67, Westview Press, Boulder, CO, arbuscular mycorrhizal fungi. Soil Biology and Biochemistry 20, WARNCKE, D, D, & BARBER, S, A, (972), Diffusion of zinc in soil, T'^^T'- I, The influence of soil moisture. Soil Science Society of America SYLVIA, D, M, (990), Distribution, structure, and function of Proceedings 26, 39-42,

10