American Journal of Botany 88(10):1786-1793. (2001)
Effects of mycorrhizal infection and soil phosphorus availability on in vitro and in vivo pollen performance in Lycopersicon esculentum (Solanaceae).
Jennifer L. Poulton, Roger T. Koide and Andrew G. Stephenson

The effects of mycorrhizal infection and soil P availability on in vitro and in vivo pollen performance were studied in two cultivars of tomato (Lycopersicon esculentum). In the first study, plants were grown in a greenhouse under three treatment combinations: nonmycorrhizal, low P (NMPO): nonmycorrhizal, high P (NMP3) and mycorrhizal, low P (MPO). Mycorrhizal infection and high soil P conditions significantly increased in vitro pollen tube growth rates but not percentage of germination. In addition, pollen from NMP3 and MPO plants sired significantly more seeds than pollen from NMPO plants in pollen mixture studies. In the second study, plants were grown initially in a greenhouse under two treatment combinations: NMPO and MPO. After all plants began to flower, they were placed in experimental arrays in the field. Under open pollination, pollen from MPO plants sired significantly more seeds than pollen from NMPO plants. This result was primarily attributed to increased flower production (and thus pollen production) in MPO plants. Thus, mycorrhizal infection and high soil P conditions can increase pollen quality (in vitro and in vivo pollen performance) as well as pollen quantity, thereby enhancing fitness through the male function. Anthocyanin production (used to determine paternity) also affected pollen performance.

Low soil P conditions often limit plant growth and reproduction, impacting agricultural and natural plant communities worldwide (e.g., Greenwood, 1981; Mengel and Kirkby, 1982; Coltman et al., 1987; Vose, 1987). Under these conditions, infection of plant roots by vesicular-arbuscular mycorrhizal (VAM) fungi is especially beneficial. Mycorrhizal fungi form mutualistic relationships with -85% of terrestrial angiosperm species (Law, 1988). Mycorrhizal infection enhances P uptake from the soil, primarily by increasing the absorptive surface area in contact with the soil solution (Hayman, 1983). Thus, mycorrhizal plants generally have a higher P status than non-mycorrhizal plants (Koide, 1998). When soil P is limiting, mycorrhizal infection often increases vegetative growth in the host plant (Smith and Read, 1997). Despite the large body of research on mycorrhizal effects on vegetative growth, relatively little attention has been given to its effects on plant reproduction, especially the male function. However, it is reasonable to expect that mycorrhizal infection and high soil P conditions will have beneficial effects on pollen performance because of increased resource availability. It is also reasonable to predict that mycorrhizal infection will improve other factors (e.g., flower production, pollen production, and pollinator visitation) that influence fitness through the male function under natural field conditions.

During pollen development, the tapetal layer of the anther wall endows pollen grains with storage products (or their precursors) that are metabolized upon germination and initial tube growth (e.g., Stanley and Linskens, 1974; Baker and Baker, 1979; Jackson, Jones, and Linskens, 1982; Wetzel and Jensen, 1992; Clément, Burrus, and Audran, 1996). For example, stored phytate is hydrolyzed into phosphate and myoinositol, which are used by the pollen tube for cell wall and membrane synthesis (Jackson and Linskens, 1982; Dickenson and Lin, 1986). The quantity and quality of these storage products can affect pollen performance, as measured by percentage of germination, pollen tube growth rates, and the ability to sire seeds in competition with pollen from other plants (reviewed in Stephenson et al., 1994; Delph, Jóhannsson, and Stephenson, 1997). Thus, any environmental conditions that affect resource availability to the sporophyte, and therefore provisioning of storage products during pollen development, can potentially influence pollen performance.

Recent studies have shown that pollen performance is affected by environmental conditions during pollen development, such as leaf herbivory, temperature, and soil nutrient availability. Simulated leaf herbivory resulted in reduced in vitro pollen tube growth rates in Silene vulgaris (Delph, Jóhannsson, and Stephenson, 1997), reduced in vivo pollen tube growth rates in Lobelia siphilitica (Mutikainen and Delph, 1996), and reduced ability to achieve fertilization under competitive conditions in Cucurbita texana and S. vulgaris (Quesada, Bollman, and Stephenson, 1995; Delph, Jóhannsson, and Stephenson, 1997). In Trifolium repens, pollen developed under cool temperatures had higher percentage of germination and grew longer pollen tubes in vitro than pollen developed under warm temperatures (Jakobsen and Martens, 1994). Similarly, pollen developed under cool temperatures from wild and cultivated Cucurbita pepo plants grew longer pollen tubes in vitro and sired more seeds in pollen mixtures than pollen developed under warm temperatures (Jóhannsson and Stephenson, 1998). In Raphanus raphanistrum, pollen from plants grown under low nutrient conditions sired fewer seeds in competition than pollen from plants grown under better nutrient conditions (Young and Stanton, 1990). Similarly, when soil nutrient levels were varied independently in C. pepo, pollen from plants grown under low nitrogen and low phosphorus conditions sired fewer seeds in competition than pollen from plants grown under high nitrogen and high phosphorus conditions (Lau and Stephenson, 1993, 1994).

In a preliminary study, mycorrhizal infection increased in vitro pollen tube growth rates of C. pepo grown in the field (Stephenson et al., 1998). If there is a correlation between in vitro and in vivo measures of pollen performance, then mycorrhizal infection should also improve in vivo competitive ability. In the study reported here, the effects of mycorrhizal infection and soil P availability on pollen performance (i.e., pollen germination, pollen tube growth, and the ability to achieve fertilization in pollen mixtures) were examined in a cultivated variety of tomato (Lycopersicon esculentum Mill.). In a related study, the effects of mycorrhizal infection on fitness through the male function (i.e., siring success in open-pollinated experimental arrays) were examined under field conditions. Only a few other studies have measured the effects of soil P availability on in vitro and in vivo pollen performance. Furthermore, this research is the first to consider mycorrhizal effects on pollen performance.


DISCUSSION

Although the beneficial effects of mycorrhizal infection on vegetative growth have been widely studied, relatively little attention has been given to its effects on plant reproduction, especially the male function. This study clearly demonstrates that mycorrhizal infection and high soil P conditions improve both in vitro and in vivo pollen performance. In other words, pollen from NMP3 and MPO plants outperforms (both in vitro and in vivo) pollen produced by NMPO plants. Moreover, there were no significant differences in pollen performance between NMP3 and MPO plants, indicating that, in terms of maintaining pollen performance, MPO plants are far more efficient at utilizing soil P. Other studies have shown that environmental conditions that affect resource availability to the sporophyte during pollen development (e.g., leaf herbivory and soil fertility) can influence in vitro and in vivo pollen performance (Young and Stanton, 1990; Lau and Stephenson, 1993, 1994; Jakobsen and Martens, 1994; Quesada, Bollman, and Stephenson, 1995; Mutikainen and Delph, 1996; Delph, Jóhannsson, and Stephenson, 1997; Jóhannsson and Stephenson, 1998; Stephenson et al., 1998). For example, high soil P conditions in the field improved siring success in pollen mixture studies with C. pepo (Lau and Stephenson, 1994). In another study with C. pepo grown in the field, pollen from my- corrhizal plants showed faster in vitro pollen tube growth rates than pollen from nonmycorrhizal plants (Stephenson et al., 1998). Thus, it is not surprising that mycorrhizal infection and high soil P conditions have beneficial effects on pollen performance in tomato.

This study also revealed that, under field conditions in low P soils, mycorrhizal plants produced significantly more flowers (and thus more pollen) and sired significantly more seeds than nonmycorrhizal plants. Consequently, mycorrhizal infection translates into increased reproductive output through the male function (male fitness) under field conditions. This large increase in the number of seeds sired appears to be due primarily to the increase in flower production (pollen production), but there was also an increase in the number of seeds sired (although not significant) that was independent of flower number (i.e., more seeds were sired due to the greater in vivo performance of pollen from mycorrhizal plants).

In this study, there is a strong correlation between in vitro and in vivo pollen performance for each treatment in the VFNT Cherry cultivar. Pollen from NMP3 and MPO plants grew longer pollen tubes in vitro and sired more seeds in vivo than NMPO plants. (Differences among the treatments in siring ability were not due to differences in percentage germination, at least under in vitro conditions.) Similarly, other studies have shown a strong correlation between in vitro and in vivo measures of pollen performance (e.g., Delph, Jóhannsson, and Stephenson, 1997; Jóhannsson and Stephenson, 1998). Moreover, pollen tube length in vitro did not differ significantly between NMP3 and MPO plants, nor did they differ in their in vivo ability to sire seeds when their pollen was deposited simultaneously onto stigmas. Thus, mycorrhizal effects on pollen performance are presumably related to improved P acquisition.

Tomato Bibliography

Physiological Predetermination

Mycorrhizal Fungi