Plant Growth Stimulation and Root Colonization Potential of In Vivo versus In Vitro Arbuscular Mycorrhizal Inocula

Inoculum of arbuscular mycorrhizal fungi, with growing use in horticulture, is produced mainly in two technically different cultivation systems: in vivo culture in symbiosis with living host plants or in vitro culture in which the fungus life cycle develops in association with transformed roots. To evaluate the effectiveness and the infectivity of a defined isolate obtained by both production methods, a replicated comparative evaluation experiment was designed using different propagules of Rhizophagus irregularis produced in vivo on leek plants or in vitro in monoxenic culture on transformed carrot roots. The size of the spores obtained under both cultivation methods was first assessed and bulk inoculum, spores, sievings, and mycorrhizal root fragments were used to inoculate leek plantlets. Spores produced in vitro were significantly smaller than those produced in vivo. Although all mycorrhizal propagules used as a source of inoculum were able to colonize plants, in all cases, leek plants inoculated with propagules obtained in vivo achieved significantly higher mycorrhizal colonization rates than plants inoculated with in vitro inocula. Inoculation with in vivo bulk inoculum and in vivo mycorrhizal root fragments were the only treatments increasing plant growth. These results indicate that the production system of arbuscular mycorrhizal fungi itself can have implications in the stimulation of plant growth and in experimental results. Arbuscular mycorrhizas (AM) are the most abundant microbial symbiosis for the majority of plant species, improving their nutrition and fitness. The presence of this mutualistic association in natural and agroecosystems, and especially the role of the fungal partner at increasing crop yields and plant health, has stimulated the search for mass production methods. Arbuscular mycorrhizal fungi (AMF) are unable to complete their life cycle and reproduce apart from the host plants, although their spores can germinate under different environmental conditions in the absence of a host (Giovannetti, 2001). Inoculum production has been improved in the last decades, but a recent article by Vosátka et al. (2012) confirms that commercial inocula sold on international markets are not always able to perform the arbuscular mycorrhizal symbiosis. There are basically three production systems for commercially available inoculants, including a variety of fungal isolates, formulations, and components: the substratebased production system, the substrate-free cultivation system (both under in vivo conditions), and the in vitro cultivation system (Ijdo et al., 2011). The in vivo substrate-based cultivation of AM in container-grown plants is the traditional and the most widely used technique for AM fungal inoculum production. It is useful for large-scale production that requires little technical support. The in vivo substrate-free system is based on hydroponics and aeroponics, but it has been limited to a few fungal species. The in vitro system allows obtaining sterile propagules with high potential for research and high-quality inoculum while lacking other microorganisms. This method has been widely tested and adopted by mycorrhizal research laboratories around the world (Declerck et al., 2005). Several reviews describe the methods for large-scale production of AMF and point out their potentials and limitations (Adholeya et al., 2005; Declerck et al., 2005; Diop, 2003; Elmes and Mosse, 1984; Ferguson and Woodhead, 1982; Fortin et al., 2002; Gianinazzi et al., 2002; Ijdo et al., 2011). Some studies on the intraradical AM development in monoxenic roots suggest characteristics of internal colonization with similar structures to those of in vivo soil-grown plants (Bago and Cano, 2005), but direct comparisons of host plant responses to these two types of inoculum are not found in the literature. In the present work, inocula of a defined isolate (BEG 72) of Rhizophagus irregularis simultaneously produced, for equal time, in vivo and in vitro, were evaluated in terms of host root colonization potential and plant growth stimulation. Materials and Methods The fungus used, obtained from Citrus aurantium L. (Camprubi and Calvet, 1996) in northeastern Spain, was isolate BEG 72 (International Bank of Glomeromycota) of Rhizophagus irregularis (Blaszk., Wubet, Renker & Buscot) C Walker & A. Sh€ussler comb nov according to 500 bp LSU sequence analysis (Camprubi et al., 2008) and formerly known as Glomus intraradices (Kr€uger et al., 2012; Stockinger et al., 2009). It was simultaneously cultivated during 6 months in association with leek (Allium porrum L., ‘‘carentan 2’’) plants and in monoxenic culture on Ri T-DNA (plasmid of Agrobacterium rhizogenes) transformed carrot (Daucus carota L.) roots (Mugnier and Mosse, 1987). For in vivo inoculum production, leek plants inoculated with R. irregularis were grown in 1-L volume containers filled with Terragreen (Oil-Dri Company, Wisbech, U.K.) for 6 months in a greenhouse under sodiumvapor lamps (temperature 20 to 28 C and 135 mmol·s·m 16 h·d). In vitro production consisted of monoxenic cultures of the same fungus established in bicompartmented petri plates supporting fungal development on modified Strullu-Romand medium (MSR) in the root compartment and on MSR medium without sucrose or vitamins in the root-free compartment (Declerck et al., 1996; St. Arnaud et al., 1996). Plates were incubated at 25 C for the same duration, i.e., during 6 months. To assure that similar material was used, we assessed mycorrhizal colonization in both the host plant root systems and the monoxenically grown roots after 6 months by the grid-line intersect method (Giovannetti and Mosse, 1980). Aiming to test a possible influence of the production method, we assessed spore diameter size, because previous observations on spore morphology indicated possible differences between spores produced in vitro and in vivo. Individual spores were extracted from the soil rhizosphere of a 6-month-old leek plant culture by wet sieving (250-, 125-, and 50-mm mesh) and decanting (Gerdemann and Nicolson, 1963) and also from the rootfree compartments of four petri plates with 6-month-old monoxenic cultures after liquefying the medium with sodium citrate buffer (Doner and Bécard, 1991). One hundred spores from each origin were included in an acid Received for publication 8 Apr. 2013. Accepted for publication 22 May 2013. We thank Dr. Christopher Walker for helpful advice on the taxonomy of the arbuscular mycorrhizas fungal isolate used in this work. We acknowledge financial support from the Spanish Ministry of Economy and Competitiveness MINECO grant AGL2010-15017. Paulo E. Lovato had a fellowship from CAPES Foundation, Ministry of Education of Brazil, Brası́lia/DF–Brazil. To whom reprint requests should be addressed; e-mail [email protected]. HORTSCIENCE VOL. 48(7) JULY 2013 897 glycerol solution (90 mL distilled water, 10 mL 1% HCl, 100 mL glycerin) individually placed on a glass slide and measured under a compound microscope. Two perpendicular transversal measures were recorded for each spore and their arithmetic mean value was the variable used in statistical analysis. A oneway analysis of variance (P # 0.05) indicated no significant differences among the size of spores produced in the four petri plates and because differences resulting from the monoxenic cultures themselves could be discarded, we compared a single sample of 100 spores produced in vitro with the sample of spores produced in association with a living plant. We carried out a comparison experiment to evaluate the influence of both inocula production methods on their effectiveness in root colonization and growth stimulation of leek plants. The experiment was replicated within a 4-month interval and plant growth was monitored at the end of the second experiment. Mycorrhizal leek plants and the four monoxenically cultured petri plates mentioned were used as inoculum sources for the experimental setup. Because preliminary analyses showed that the mycorrhizal colonization in the roots of both the leek plant culture and the Ri T-DNA transformed carrot roots in the in vitro cultures was similar and above 80%, we sought to apply equivalent amounts of potential mycorrhizal propagules originated in vivo and in vitro, and inoculation treatments were as follows. In vivo inoculation treatments consisted of: 1) bulk inoculum: 2 mL of substrate (with 50% pore space) from the host plant rhizosphere, which contained 50 ± 15 spores/g and 155 ± 54 mm of root fragments varying in shape and size (Newman, 1966) plus external mycelium fragments; 2) roots: 10 10-mm root fragments excised from the host plant heavily colonized root system; 3) spores: 50 individual spores recovered by wet sieving (pool of all spores retained on 250-, 125-, and 50-mm mesh sieves) of the bulk inoculum; and 4) sieving: 20-mL aliquot of a distilled water suspension of fungal material recovered after sieving a 2-mL sample of in vivo bulk inoculum through a 125-mm mesh sieve. In vitro inoculation treatments consisted of: 1) bulk inoculum: 1-mL portions excised from the MSR medium root compartment, which contained 70 ± 21 spores and 284 ± 83 mm of heavily colonized root fragments varying in shape and size (Newman, 1966) plus external fungal mycelium; 2) roots: 10 densely colonized mycorrhizal 10-mm root fragments excised from the same compartment after accurate observation under a stereomicroscope (·80 magnification) to confirm mycorrhizal colonization; 3) spores: 50 individual spores obtained after liquefying the medium in the root-free compartment with sodium citrate buffer as mentioned previously; and 4) sieving: 20-mL aliquot of a distilled water suspension including fungal material recovered after sieving through 125 mm a suspension obtained after dissolving the agar from the root-free compartment of a monoxenically cultured petri plate. A noninoculated control treatment was included in the experimental design. To verify possible nonmycorrhizal effects of other inoculum components on plant growth, after harvesting Expt. 2, four additional control treatments were compared with the former control treatment: 20-mL leachates free of mycorrhizal propagules from in vivo and in vitro bulk inoculum water suspensions, nonmycorrhizal leek roots, and nonmycorrhizal transformed carrot root fragments. To estimate the colonization density in roots, the number of propagules in 20 10-mm root fragments from a 6-month-old leek culture in Terragreen and from a 6-month-old in vitro culture of R. irregularis (BEG 72) were counted under the stereomicroscope (·80 magnification). Plant root samples were previously cleared and stained (Koske and Gemma, 1989), whereas structures inside in vitro roots, visible without any processing, were observed and counted with no clearing nor staining. Leek plantlets previously germinated in autoclaved sand trays were transplanted to 100-mL individual containers filled with pasteurized (100 C, 2 h, twice) sandy soil (950 g·kg sand, less than 2.5 g·kg organic matter, pH 8.2, 2.0 mg·kg Olsen-P), and the different inoculation treatments were applied under the plant roots at transplant. Containers, kept for 10 weeks in a greenhouse (20 to 28 C and 135 mmol·s·m 16 h·d), received weekly applications of Hoagland’s solution without phosphorus and irrigation was controlled daily. There were six replicates for each treatment and we carried out two consecutive experiments within a 4-month period. The percentage of root colonization was estimated (Giovannetti and Mosse, 1980) for both subsequent experiments and they were shown to be reproducible. When mycorrhizal root colonization percentage data obtained in each replicated experiment were analyzed as a factorial design, no significant differences (P # 0.05) between them could be detected. Therefore, 12 replications per inoculation treatment in a single experimental design were considered for statistical analysis in two one-way analyses of variance (ANOVAs) (P # 0.05), comparing first the origin of the inocula (either in vivo or in vitro) and second the type of propagules used as an inoculum source. At the end of the second experiment, plant shoot dry weight was determined. Results were statistically analyzed by ANOVA and means were compared with Tukey test (P # 0.05).