Biotechnology may offer unique opportunities for pest control in perennial tree and vine crops (Dandekar et al. 2002). Trap crops — plants that an insect pest prefers to the commercial crop — have been tested in a number of agricultural settings, but in most cases have not achieved control levels high enough to completely replace chemical pesticides. Insects are attracted to the trap plant, but they multiply there and can spread to the adjacent crop. A variant on this concept is to incorporate expression of the Bacillus thuringiensis (Bt) insecticidal protein into the trap plant. When the insect feeds on the transgenic trap plant, it dies and the insect population is reduced, thereby protecting the nearby commercial crop.

Dry Creek Laboratories of Hughson, Calif., demonstrated this concept with codling moth (Cydia pomonella), a major pest of apples, pears, walnuts and other fruits. The female moth lays eggs on the leaves or fruit, which then hatch into larvae that burrow into the fruits, making them unmarketable. Pesticide sprays and pheromone disruption are generally used to control this pest. However, the female moth prefers to lay its eggs on apple trees. Under license from Monsanto, Dry Creek Laboratories developed apple trees capable of expressing a Bt protein that was toxic to the codling moth larvae, with the intention of using these plants as trap crops in and adjacent to walnut orchards.

A 90-acre field trial was established in 1997, and in the 4 subsequent years worm damage to the walnuts was almost completely controlled without pesticide applications, equivalent to that in the plots sprayed three times per season with pesticides. While walnuts have also been transformed to express the Bt protein directly (Dandekar et al. 2002), an attractive feature of this scheme is that the walnuts themselves are not transgenic and the method could be used to protect existing orchards by interplanting the Bt-expressing apple or crabapple trees. Broader application of this approach could result in more effective trap crops for a number of annual and perennial crops. Unfortunately, Dry Creek Laboratories is unable to move forward at this time with commercialization of the Bt apple plants due to the costs associated with the regulatory process required for biotech crops ( see page 106 ).

Left, apple roots engineered to silence bacterial genes are resistant to crown gall formation. Right, control (nontransgenic) roots infected with the same bacterial strain show extensive gall proliferation.

Another opportunity for biotechnology in perennial crops that are normally grafted is to engineer only the rootstock for desirable traits. Commercial tree cultivars grafted onto transgenic rootstocks could benefit from increased rootstock productivity or disease resistance while producing nontransgenic pollen and fruit. For example, such applications in grapes could offer new solutions to Pierce's disease or Phylloxera by grafting traditional varieties onto resistant transgenic rootstocks. The feasibility of this approach was recently demonstrated for resistance to crown gall disease (Agrobacterium tumefaciens). Infections by the bacterium result in the formation of a gall, an unorganized mass of plant cells that results from overproduction of two plant hormones. The bacterium has the natural ability to transfer some of its genes into the host plant's genome following infection. The transferred genes code for three specific enzymes. When the plant expresses these genes, the enzymes synthesize the two hormones that induce the plant to form the tumor, or gall, on which the bacteria live. Eventually, the galls can girdle the stems and reduce the vigor of the tree or vine.

A biotechnology tool called “gene silencing” has been used to generate resistance to crown gall. This method involves transforming plants with DNA that, when expressed, produces signals that block the expression of any genes with the same sequence as the inserted DNA. Plants transformed with these interfering versions of the three enzyme genes would be primed to block the function of the corresponding bacterial genes in infected plants. This would prevent the formation of the damaging galls without even needing to kill the bacterium itself. The feasibility of this approach was demonstrated in tomato and Arabidopsis plants (Escobar et al. 2001). Furthermore, both walnut (see photo; Escobar et al. 2002) and apple (see photo; J. Driver et al., unpublished results) plants resistant to crown gall have been produced. As most crown gall infections occur in the rootstock, nontransgenic scions grafted on resistant transgenic rootstocks would be protected from the disease. Rootstock engineering holds great promise for the improvement of tree and vine crops by preserving the horticultural characteristics of existing varieties used as scions while incorporating beneficial traits into the rootstocks.

Crown gall formation was suppressed in walnut plants engineered to turn off specific bacterial genes. (A) The control shoot exhibits a large, undifferentiated tumor at 5 weeks after inoculation with a virulent A. tumefaciens strain, while (B) a shoot engineered for resistance exhibits no tumor. Source: Escobar et al. 2002.