SCOPE 49 - Methods to Assess Adverse Effects of Pesticides on Non-target Organisms

16.1 INTRODUCTION

Integrated management of pests (IPM) or pathogens is an important part of crop production strategies. Integrated management strategies have as their goal providing maximum economic yields for the farmer, while improving and/or maintaining the production site and protecting adjacent habitats. IPM has as its goal, therefore, the protection of crops from pests and pathogens without producing negative environmental impacts. IPM differs from conventional control in that former deals with the long-range goal of maintaining or increasing productivity; whereas the latter protects the cropped environment and does not include site preservation or non-target effects on adjacent habitats and organisms. Consequently, as a component of integrated crop production strategies, IPM requires long-term investment at the production site and adjacent sites; thus, it mandates an investment in the future that can reduce short-term productivity and profits. However, production practices afforded by integrated crop management strategies can lead ultimately to enhanced productivity and minimal environmental impact. Certainly, producers will not apply IPM strategies if the short-term benefits do not result in profitable returns, since losing the farm because of financial losses is a powerful argument for short-term solutions. IPM must be used to develop more cost-effective methods for farming while protecting both production and adjacent environments for the future. Biotechnology has a significant role to play in contemporary IPM strategies and applications (Napoli and Staskawicz, 1985).

IPM requires coordination of biology, ecology, chemistry, and engineering for crop production; thus, biologists, ecologists, chemists, and agricultural engineers must work together to solve anticipated and unexpected problems. Biotechnology will accelerate the production of more vigorous, productive, and resistant crop plants less susceptible to environmental stresses. This advance includes microbes that contribute to plant productivity by causing more rapid plant development or by protecting plants from pathogens or pests. Engineered plants and beneficial microbes also have the potential to cause problems, since careless development or release may inadvertently lead to the introduction of environmental problems capable of self-replication (Tauber et al., 1985). This chapter deals with the methodologies for assessing the impact of organisms engineered for IPM applications, by examining possible strategies to evaluate and develop biotechnology methodologies related to IPM, describing organisms currently being developed that may impact IPM, and identifying research needs. 

16.2 BIOLOGICAL CONTROL CONCEPTS

Historically, biological control has been an important part of IPM. Biological control is defined as the management of a pest or pathogen using a second organism, or biological control agent, and cultural methods directed towards reducing the impact of a specific pest. Biotechnology will have an impact on IPM in the near future (Mets et al., 1986). The first biotechnological successes related to IPM have already been achieved, with engineered microorganisms or genes of microbes engineered into plants (Lincoln et al., 1985). A stated goal of IPM projects is to reduce the levels and frequency of application of agricultural chemicals. This goal creates an expanded role for the use of biological control methods, including genetically engineered microorganisms (GEMs) designed as antagonists towards specific plant pathogens. The following sections will address recombinant DNA technology, including the introduction of microbial genes in plants, applied to microbes for biological control. Genetic engineering of animals and products designed for the improvement of animal health will not be covered.

16.3 BIOTECHNOLOGY CONCEPTS

Biotechnology has been difficult to define for both laymen and scientists. Simplistically stated, biotechnology is the application of biological principles to the technologies for production of food, fibre, fuel, or pharmaceuticals. Realistically, biotechnology has been utilized since the first human intentionally planted crop plants or husbanded animals. The addition of recombinant DNA techniques to biotechnology may provide a revolution as profound as the prehistoric discoveries that using seed from more productive plants or breeding animals with desirable characteristics increased the number or quality of agricultural products (Ouellette and Cheremisinoff, 1985).

In the United States, the consistent development of agricultural biotechnology began with the establishment of the first agricultural experiment station in 1887 in New Haven, Connecticut. Until that time, biotechnology had developed in a non-directed manner similar to initial developments in some other field of science or engineering and similar to the application of developments in agriculture. Within the current system of experiment stations, land-grant universities, federally supported research, and the Extension Service, the process of biotechnology development has accelerated at an extraordinary pace, with new scientific breakthroughs reported almost weekly. Biotechnology has evolved to include the application of the newest scientific methodologies to biological productivity. This situation means that, for agricultural biotechnology, those sciences related to genetic engineering will actually encompass recombinant DNA technologies along with the support disciplines for the regeneration of whole organisms, including plant regeneration, in vitro fertilization, and embryo implantation of animals. 

16.4 SHORTCOMINGS OF BIOCONTROL AGENTS

16.4.1 DETECTION, RECOVERY, AND ENUMERATION

To monitor the fate of recombinant DNA in the environment, methods utilized for detection, recovery, and enumeration must be precise, accurate, and sensitive. An important concept is that the products of recombinant DNA may reside in the engineered organism, in other organisms, or in the inorganic or organic components of the habitat to which the organism was applied. Therefore, monitoring techniques should have enough flexibility to cover all these possibilities. Obviously, a combination of techniques is usually more effective than a single procedure (Milewski, 1985; Rissler, 1984).

16.4.1.1 Culture

Two limitations of culture techniques exist: the organism(s) must be able to grow on some selective medium in order to be detected at low population levels. If gene exchange has occurred, the recipient organism may not be detectable by these means. A final limitation is that viable counts, by definition, will only record actively growing cells. Stressed and hypobiotic cells or free DNA that are non-detectable may be important in the dissemination of recombinant DNA (Atlas, 1982; Roszak and Colwell, 1987).

16.4.1.2 Detection limits

Methodologies for assessing safety of engineered biological control agents are only useful if the limits of detection are clearly understood. By currently available methods, detection of viable or non-viable cells at population levels below 10³ colony-forming units per gram of soil is usually inaccurate.

16.4.1.3 Direct counts

Direct counting procedures (e.g., acridine orange staining, immunofluorescence assays ELISA, serological staining, and monoclonal antibody procedures) usually reveal higher numbers in a population, and may be more sensitive than viable cell counts; but they are limited, since they are usually less accurate (i.e., acridine orange stains nucleic acids in all bacteria and fungi) and often count dead as well as living cells. Polyclonal serological systems recognize a battery of antigens that may occur in several organisms. Monoclonal serological systems may be specific enough to follow the presence of a gene product related to recombinant DNA. An attractive system would be to use a battery of monoclonals to recognize recombinant DNA gene products and the possible hosts of that DNA.

16.4.1.4 Genetic markers

Genetic markers, such as spontaneous indicators to antibiotic resistance, are useful in selecting strains that have been introduced into a habitat; and they often increase the levels of detection, especially if two or more markers are used in combination (Hagedorn, 1986). The limitations of this procedure are that only introduced strainswhich are weak because they are mutantscan be followed, and that there are no markers to follow indigenous strains, such as the wild-type parent (Hagedorn, 1986). The use of genetic markers also hinges on the stability of the markers. For instance, ampicillin mutants occur at a frequency approximately 100 times higher than tetracycline-resistant mutants. At high population densities, spontaneous ampicillin mutants among indigenous bacteria will interfere with the interpretation of results before spontaneous tetracycline mutants would do so. Recently, there has been considerable discussion of the potential mobility of the mutant gene within the general microbial population. The issue is the transfer of the resistant element to other bacterial strains, thus dispersing into the environment resistance to clinically important antibiotics. There is also the possibility that such strains or genetic recipients might enter food chains and contribute to antibiotic resistance in microbes inhabiting the gastrointestinal tract of humans or animals. However, new technologies afforded by transposable elements have provided solutions to many of the problems, such as marker stability, that are inherent in selecting for spontaneous mutants. As an example, genes coded for lactose utilization (Monsanto LacZ) have bene introduced into a strain of Pseudomonas aureofaciens that has been introduced in a field test. This marker is selective and non-antibiotic and provides a specific identifying characteristic for the introduced strain.

16.4.1.5 Detection of DNA from soil

Perhaps the most sensitive method for detecting genetically engineered organisms in the soil is to detect the recombinant DNA itself in the environment. This method may be quite important, since it will detect cell-free DNA as well as DNA in non-viable and viable cells. Cell-free DNA and DNA in non-viable cells may be important in the dissemination of recombinant DNA through the process of transformation of competent recipient cells with naked DNA. DNA may persist for long periods adsorbed on mineral particles or in or on organic debris. Some of these sites may be relatively protected from soil nucleases (Kelman et al., 1987). The analytical technique is based upon amplification of the DNA extracted from the environment, followed by detection using specific hybridization probes (Somerville et al. 1988). Simply, total DNA is isolated from soil using a variety of techniques, usually featuring polyvinyl polypyrrolidone to remove humic compounds, sodium dodecyl sulphate as a lysis detergent, and purification by buoyant density centrifugation in caesium chloride gradients. The DNA amplification cycle consist of rounds of denaturing the DNA to single strands, annealing a specific primer, and DNA polymerase-mediated manufacture of a complete second strand (Ou et al., 1988; Saiki et al., 1988). By repeating these three steps several times, DNA from as little as one bacterial cell per gram of soil may be detected. This is a significant increase in sensitivity and level of detection compared to viable-cell-count and immunofluorescence procedures described above.

16.4.1.6 Movement of genetic material

Genetic transfer by microorganisms has been recorded in aquatic and marine habitats (Morrison, et al., 1978; Baross et al., 1967), in soil (Graham and Istock, 1978; Weinberg and Stotzky, 1972), and in plants (Kerr et al., 1977; Lacy, 1978; Lacy and Leary, 1975; Talbot et al., 1980; Lacy et al., 1984). Philosophically, some level of genetic exchange among microbesengineered or notmust be expected, since they have so many useful exchange mechanisms, including conjugation, transformation, transduction, and cell fusion. The frequency of transfer may be expected to be lower than in the laboratory in soil and water, since the conditions may not be as optimal as on laboratory media. However, in plants, transfer frequencies have been recorded that are higher than laboratory frequencies, possibly due to the absence of microbial competitors, the presence of abundant nutrients, and support for mating pairs in plant tissue (Kerr et al., 1977; Lacy, 1978; Lacy and Leary, 1975; Lacy et al., 1984). Some tentative evidence has been collected indicating that engineered microbes may rearrange recombinant DNA in vivo and transfer DNA to indigenous organisms. Confirmation will be required for these results (Jansson and Tiedje, 1988; Orvos et al., 1988).

Detection of transfer has been performed by conventional genetic procedures which require expression of a genetic trait and detection by viable counts of the bacterium acquiring the genetic trait. Dissemination of genetic elements below the limits of detection cannot be measured; likewise, unexpressed DNA, which may later be disseminated to a second organism in which expression may occur, cannot be detected. Enrichment culturing may increase the sensitivity of viable counting, but this technique assumes we know which organisms to culture and how to culture them. Using the new DNA amplification techniques, it should be possible to detect the continued presence of DNA in the environment below the limits of viable counts. If that DNA is contained in a living organism, it will be more difficult to determine.

16.4.2 HABITAT SIMULATION

Artificial habitats (i.e., microcosms) are an essential intermediate step to ecologically sound analyses leading to environmental release. The problem with microcosms is that they never adequately duplicate natural ecosystems (Klute, 1985; Morley et al., 1983; Page et al., 1982). Simplistic microcosms may be replicated, but they do not mimic environmental fluctuations in temperature, humidity, pH, light intensity, and other multitudinous variables. Complex microcosms simulate some, but not all, environmental parameters, and are difficult to replicate because of the cost and complexity (Roberge, 1978). Larger microcosms (mesocosms) circumvent some of the problems by increasing the number of samples taken within one mesocosm. Mesocosms are difficult and expensive to duplicate; hence, treatment numbers and replicates are limited (Brown, 1987). Research is needed to increase basic understanding of environmental interactions and cycles. Currently, our understanding of these phenomena is so scanty that our ability to engineer adequate microcosms and to interpret results derived from them is restricted (Armstrong et al., 1987; DeAngelis, 1973; Kool et al., 1987). Scientific understanding and appropriate methodologies are absent to examine the effects of GEMS on ecosystem structure and function. Studies that add a GEM to a natural ecosystem usually require extensive microcosm simulations. Three principal research approaches that appear to be useful include: expanded use of microcosms as representations of various ecosystems, including appropriate design features, characterization of the most critical parameters, calibration to the field, and standardization of protocols; establishment of ecosystem baseline data for any particular habitat to permit confident analysis of ecological effects (and the degree of change required to detect a significant effect); and the generation of positive controls to ensure that test systems are appropriate to assess ecosystem effects (Brezhnev et al., 1974; Hunter et al., 1986).

16.4.3 MODELLING FOR PREDICTIVE CAPACITY

Modelling procedures are suggested by any use of microcosms. Three types of models exist: descriptive, analytical, and predictive. To date, most models are descriptive, since our knowledge of microbial survival, movement, and colonization is very limited. Most models are reactive, since they are usually based on existing data. Active models are needed that indicate what data are required to satisfy the questions being posed. Evolution of modelling is related directly to the amount of accurate descriptive data available. At this time, modelling becomes more precise and analytical as the quality and amount of data improve. Eventually, models may develop predictive capacity based on the precision of the analytical forebears. Numerous modelling approaches are available to encompass habitat and organism inventories, population dynamics, nutrient flow and cycling, and interrelationships among environmental factors (Corapciolu and Haridas, 1984). Such simulation models are useful to characterize parameters, establish baseline data points, and identify and prioritize the most important processes for investigation (Brezhnev et al., 1974; King and DeAngelis, 1986). Research trends in modelling must include some element of experimental procedure that incorporates: the design of a microcosm to simulate some important aspects of the proposed release area; evaluation of the GEM by monitoring a set of parameters deemed `most critical' for the particular microcosm; and development of a descriptive model of the microcosm that can account for process rates, perturbations, and analyses (Hicks and Newell, 1984; Minogue and Fry, 1983; Thingstad and Pengerud, 1985). Lastly, numerical analyses of data assume a uniquely important aspect in ecosystem studies because of the large variation inherent in the components of the system and the difficulty of establishing cause and effect relationships (Rouse, 1985). It will be necessary to define normal ranges of variability and function for specific ecological populations and/or processes and naturally occurring levels of change (Gladden and Ginzburg, 1982). Only then can perturbations be detected and the probability be determined that the alteration was associated with the presence and/or activities of a GEM.

16.4.4 ASSESSING ORGANISMAL FITNESS

Organismal fitness is a measure of an engineered biological control agent and its ability to enter into, survive, and colonize a particular habitat whether that habitat is the target or a non-target habitat. Fitness, therefore, is related to a complex assortment of traits for competitiveness, substrate utilization, rate of replication, optimum growth temperature of pH, and a myriad of other abilities. Since the components of fitness are so many and complex, testing them separately, although of academic interest, would be the least efficient method for determining fitness (Silver et al., 1986). Currently, direct evalution by competition of the engineered organism is the most efficient method for testing fitness. In this method, a series of mixed inocula are utilized so that the wild-type parent is the dominant population in some inocula, and the engineered strain progeny is dominant in others. By measuring relative changes in the ratios of colonization, the competitive ability of the engineered strain may be measured directly against the parental strain.

16.4.5 ESTIMATING RISK

Currently, risk assessment is an imprecise art that involves determining exposure combined with potential hazard to arrive at a level of relative risk or probability of a specific adverse effect. Risk assessment involves using available background data to design an effective artificial habitat that will generate enough additional data to construct a predictive model. This model can then be used to estimate the probability that a specific adverse effect might occur. The limitations of risk assessment are that non-specific effects are difficult to assess, the reliability of predictive systems is poor, and negative results (i.e., no adverse effect is detected) do not provide adequate statistical confidence (Cairns and Pratt, 1987). 

16.5 ENGINEERED BIOCONTROL AGENTS

16.5.1 MICROBIAL INSECTICIDES

Bacillus thuringiensis produces a group of closely related potent polypeptide toxins (BT toxins) that are active against several chewing insects, including various beetles and moths (Bernhard, 1986). Some of these insects are serious pests of economically important crop plants. The activity spectrum of species against which the toxin is effective may be manipulated by selecting strains of B. thuringiensis that produce toxins with different activities (Krieg et al., 1983). Since the gene responsible for the production of this toxin has been cloned and characterized at the molecular level, the possibility for genetic engineering to further restrict or extend the activity spectrum of the toxins is a distinct possibility (Hernstadt et al., 1987). BT toxins have been utilized very successfully in biological control for many years. Their effectiveness is due in part to the ability of B. thuringiensis to produce very resistant structures called endospores, which contain protein crystals of the toxin (Bulla et al., 1979; Whitely and Schnepf, 1986). Sprayed on aerial surfaces of plants, these spores are ingested by chewing insects, releasing the toxin into the gut of the insect. If the insect is susceptible to its toxic effects, it dies. Below ground applications of the endospores have not been effective due to the relatively low cost of treatment and to their rapid deterioration probably by microbial degradation (Hernstadt et al., 1986).

Biotechnological research has indicated other ways to utilize BT toxins for the biological control of pests. For example, the genes mediating production of BT toxin have been moved into strains of fluorescent pseudomonads, bacteria that inhabit the rhizoplane and rhizosphere of plants. The biological control tactic was that the BT toxin, now produced by the pseudomonads, would be ingested by root-grazing insects, thus extending the use of BT toxins below the surface of the ground (Schroth and Hancock, 1985). The toxins were found to be effective in laboratory tests for controlling some root pests. The development of this method for biological control has apparently been suspended, since the system did not offer methods for controlling the movement of the root-inhabiting pseudomonads to other plants and possible environmental impacts on the distribution of non-target insects. A second tactic incorporates the genes for BT toxin production into the genome of a crop plant (Horsch et al., 1984). Chewing insects feeding on the plant tissue ingest BT toxin. Ecologically, this is a better strategy, since the BT toxin will only be active against insects that are pests on that particular plant. However, the movement of the BT toxin gene from one plant to others through sexual exchanges may allow the gene to escape into weed plants affecting those insects that are pests on the weeds (Tomashow et al., 1980).

Baculoviruses are among the most promising alternatives to chemical pesticide control of major agricultural and forest insect pests. These viruses are presently being genetically engineered by commercial companies to enhance their pathogenicity and for their expression in vector systems. Within the next five years, commercially desirable, genetically engineered baculovirus (GEV) pesticides will probably be available for release into many types of environmental settings. These GEVs will contain a variety of foreign genes (Biever et al. 1982; Falcon et al., 1968).

Several strategies can be used for inserting `pesticidal' genes into baculoviruses. The first calls for early insertion of the foreign gene under the control of a duplicated immediate or early viral promoter (Summers et al. 1980). The second calls for insertion of the foreign gene under the control of a very late viral promoter. The third strategy for construction of a GEV is to remove the polyhedron gene (poly-GEV) and to replace it with a `pesticidal' gene. The polyhedron gene product is responsible for occlusion of virions within polyhedra, which is required for survival and transmission of baculoviruses in nature. For field application, the poly-GEV will need to be occluded by the polyhedron gene product provided by the wild type (poly + virus). This can be achieved by coinfection with both virus types. Accordingly, the poly-GEV can only persist following co-infection of larval cells, and will rapidly be lost from the environment. The only avenue available for persistance of the GEV is through illegitimate recombination (foreign gene insertion into the poly + genome without displacement or inactivation of viral genes).

Other microbial insecticides being actively investigated include both fungi and protozoa (Falcon, 1985), although these entomopathogens are at present of little importance compared to BT and the baculoviruses.

16.5.2 FROST CONTROL

Ice crystal formation in plant cells ruptures cell and organelle membranes, and often results in cell death. Plants, especially those that have evolved in temperate climates, have developed several methods to deal with ice formation. Among these are cytoplasm concentration to reduce the formation of ice and nucleation of ice at warmer temperatures, so that ice forms intercellularly rather than intracellularly. Many economically important crop species evolved in warmer climates and do not have obvious methods for protection from ice formation. However, a paradox exists; plants are often damaged by frost at temperatures of 2 to 5 °C, yet plant tissue freezes at much lower temperatures of 7 to 10 °C; because the cytoplasm is not pure water. This difference is attributed to the presence of ice-catalysing epiphytic bacteria such as members of the almost universally present genera of Pseudomonas, Erwinia and Xanthomonas (Kim et al., 1987; Lindow, 1982). Molecular studies indicate that ice-nucleation-active bacteria contain a single gene coding for a protein that seems to be responsible for catalysing ice (Ou et al. 1988). Using natural strains of non-ice-nucleating bacteria or chemical mutants deficient in the ability to catalyse ice formation, it was discovered that plants may be protected to some extent from ice damage by manipulating their epiphytic flora to favour the non-ice-nucleating strains (Lindow, 1982; Orser et al., 1985). Since strains of ice-nucleating bacteria demonstrate more or less host specificity in their epiphytic relationships with plants, greater protection is afforded by constructing non-ice-nucleating strains from the wild-type ice-nucleating strains by deleting the ice-nucleation gene (Lindow, 1985). Since chemical mutants usually have secondary mutations often affecting growth rate and nutrition, genetic engineering was selected as the technique of choice to modify wild-type, highly strain-specific bacteria into non-ice-nucleating biological control agents for protection from warm temperature frost damage (2 to 5 °C). One potential environmental impact would be that non-ice-nucleating engineered strains might affect weather patterns by perturbing any role ice nucleators might have in rain nucleation. Since natural non-ice-nucleation-active strains occur in nature, this seems unlikely. Limited field trials with engineered strains indicate that some control of warm temperature frost damage may be expected without apparent environmental impact.

16.5.3 RESISTANCE TO VIRUS-CAUSED DISEASES

Molecular biology of plantviral interactions have provided much information about plant infection, viral replication within the plant, and recognition of viruses by plants. Based on this information, several strategies have been developed to make plants more resistant to viral diseases (Ponz and Bruening, 1986). For the first strategy, plants resistant to virus-caused diseases often recognize some component of the virion, thus triggering a resistant response (Nelson et al., 1988). To construct plants resistant to virus-caused diseases, genetic engineers have introduced a gene for viral coat proteins into the plant genome. In this manner, the plant produces a coat protein that triggers a plant's own defences, so that it is already in the resistant mode before contact with the virus.

The second tactic is to inactivate replicons of the viral genome as they are produced. Many plant viruses are single-stranded RNA replicated from a complementary strand of RNA that is in much lower concentration than the single-stranded virus itself. Observations indicate that the virus could actually be inactivated if the complementary strand were in higher concentration than the virus, perhaps by binding the complementary strand to the virus, thereby blocking the translocation of viral products required for disease initiation (Palukaitis and Zaitlin, 1984).

The third tactic is to engineer plants with a DNA copy of a satellite virus genome that causes attenuation of viral-caused disease. This system has an added advantage in that it is induced only when complementary viral particles are present; that is, resistance is `turned on' only when needed and does not form a continuous energy drag on the host plant. Tolerance, rather than complete resistance, is mediated by this form of disease control (Harrison et al., 1987).

The fourth tactic takes advantage of the newly discovered class of enzymatic nucleic acidsribozymes, or catalytic RNAs. In this system, a sequence of DNA containing an inserted template for a ribozyme produces an mRNA complementary to the viral genome and is introduced into the plant genome. Upon complementation of the viral genome, the ribozyme catalytically cleaves the viral genome in such a manner as to prevent further viral replication (Gerlach, 1988; Gerlach et al., 1987).

16.5.4 HERBICIDE RESISTANCE

Genetic improvement of crop tolerance to herbicides would provide the possibility for weed control using fewer herbicides, since crop plants now susceptible to those herbicides would be resistant to commonly employed herbicides (Cocking et al., 1981; Hatzios, 1987).

16.5.4.1 Triazine resistance

Triazines are photosynthesis-inhibiting herbicides known to act on the pastoquinone-binding chloroplast-membrane Q_b protein encoded by the psbA gene. Genes from triazine-resistant and sensitive weeds are found to differ by as little as a single nucleotide base in their sequences (Arntzen, 1986; Mets et al., 1986). Triazine-resistant plants have been constructed by moving a gene from bacteria into plants using the molecular vectors developed from the plant pathogenic bacterium Agrobacterium tumefaciens and its plasmid (PTi) which is the vector of oncogenicity genes.

16.5.4.2 Glyphosate resistance

Glyphosate inhibits 5-enolpyruvyl-shikimate--phosphate synthetase (EPSP) in the shikimate pathway for biosynthesis of amino acids, plant growth hormones, phytoalexins, and other phenolic compounds (Amrhein et al., 1983). Alleles of the aroA gene encoding EPSP that provide a resistant phenotype have been described for species of bacteria. An aroA gene from the human pathogen Salmonella typhimurium was moved using the Agrobacterium pTi system into plants and conferred some tolerance to the herbicide (Comai et al., 1986; Shah et al., 1986).

16.5.4.3 Sulphonylurea and imidazolinone resistance

Susceptibility to the structurally unrelated sulphonylurea and imidazolinone herbicides is due to inhibition of a single enzyme, acetolactate synthase (ALS) (Chaleff and Mauvais, 1984; Shaner et al., 1984). This enzyme is the first specific one in the pathway to branched-chain amino acids such as valine, leucine, and isoleucine. Resistance to chlorsulphuron (a sulphonylurea herbicide) was engineered into tobacco by first fusing the open reading frame of a chlorsulphuron-resistant ALS gene from yeast to the sequence coding for the lead peptide of a small subunit gene for ribulose biphosphate carboxylase in tobacco, and then transferring it into the plant using a pTi vector (Falco et al., 1985). The chimeric gene increased threefold the resistance to a herbicide, indicating that further refinement of the system will be necessary before field use is practicable (Fraley et al., 1986). 

16.5.4.4 Herbicide degradation

Biodegradation is an important factor affecting environmental persistence of most organic herbicides used in contemporary agriculture. Microbes are the chief agents of herbicide degradation, most probably through the action of selected constitutive or inducible enzyme systems (Hatzios, 1987). Often components of the herbicides may be used as microbial nutrients. The rate of microbial degradation is influenced by physicochemical characteristics of soils, rate and frequency of herbicide application, the cropping system, and the presence of other pesticides. Repeated applications of herbicides may lead to increased levels of microbial degradation to the point where succeeding applications are ineffective on target pests.

Currently, the mechanisms of biodegradation in soils are poorly understood. However, over thirteen genera of bacteria and five genera of fungi have been associated with biodegradation of herbicides to date. Elucidation of the mechanisms for biodegradation using molecular techniques will aid in the design of herbicides increasingly resistant to microbial degradation.

Environmental benefit may be hidden within the problem of herbicide biodegradation. Improved strain selection and biological engineering may provide strains to be used to ameliorate environmental damage due to pesticide spills or to detoxify pesticides in the soil column to protect groundwater from contamination.

16.5.4.5 Herbicide `safeners' 

A `safener' is a term used to describe a compound that protects crop plants from herbicide injury. Antidote, protectant, and antagonist are synonyms. Microbial strain selection and genetic engineering may result in the development of improved herbicide safeners. Safeners allow the extended use of herbicides on crops that would ordinarily be damaged by those herbicides. In IPM schemes, safeners simplify crop protection by reducing the numbers of pesticides in the environment.

Microbial safeners may be developed by using herbicide-biodegrading microbial strains to protect susceptible crop plants from these herbicides. Microbial safeners will probably be most effective against soil-applied herbicides (Hatzios, 1987). The crop plants may be specifically protected by applying the microbial safeners directly to roots of transplants or to seeds before sowing. 

16.5.5 BIOLOGICAL CONTROL

Specific strains of Pseudomonas fluorescens and Pseudomonas putida provide biological control of fungal plant pathogens and deleterious rhizobacteria with concomitant positive growth responses above and beyond simple amelioration of disease. The ability of these strains to act as biological control agents has been attributed to their ability to rapidly and competitively colonize roots and to produce siderophores (iron-binding compounds) and antibiotics. The first of these biological control formulations based on Pseudomonas has been placed into the field for commercial control of damping-off fungi on cotton by Ecogen Inc. (Makulowich, 1988). Not much is understood about the competitive ability to colonize plant roots or the host plant specificity demonstrated by some of the bacterial strains. However, some information is known about siderophore and antibiotic production and is discussed below.

16.5.5.1 Siderophores

Most microorganisms respond to low-iron stress by producing extracellular, lowmolecular-weight (5001000 Daltons), iron transport compounds (or siderophones) which bind iron selectively and with great binding power. These compounds are bound with iron for microbial metabolism. For instance, oxidative phosphorylationprobably the most important energy-producing process in a cellrequires iron-containing proteins for electron capture and transfer. Among fluorescent pseudomonads, such as P. fluorescens and P. putida, yellow-green fluorescent pyoverdines act as siderophores. The best known of these is pseudobactin, a hexapeptide cyclized through a fluorescent quinoline derivative with a hydroxamic iron-chelating group derived from moieties of ornithin, aspartic acid, and aniline residues (Leong, 1986). Pseudobactin requires several genes for its biosynthesis; at least five gene clusters and a minimum of seven genes are required for biosynthesis of pyoverdine siderophores (Marugg et al., 1985; Weisbeek et al., 1986).

Siderophores evidently have a role in disease control, since the incidence and severity of disease is reduced in soils containing either siderophore-producing microbes or deliberately added pyoverdine siderophores. Finally, in soils containing excess iron, neither siderophore-producing bacteria nor added pyoverdine siderophores were as effective in reducing disease incidence or severity (Leong, 1986). Unknown is whether siderophores act by chelating iron so that it is unavailable to fungal plant pathogens (Leong, 1986), act as antibiotics (Ahl et al., 1986), or act by a combination of mechanisms. Genetic engineers may manipulate siderophore production to produce greater efficiency, higher levels of production, and constitutive variations of siderophores. Until these experimental constructions are made and tested, the role of siderophores in biological control and integrated pest management remains to be fully understood.

16.5.5.2 Antibiotics

Pseudomonads produce a host of antibiotics including phenazines, pyrroles, pseudomonic acid, pyo compounds, and amino acid-containing antibiotics (Leisinger and Margraff, 1979). By correlating the activity spectrum of each antibiotic in vitro with the control of target pathogens using purified antibiotics, the antibiotics pyrrolnitrin, pyoluteorin, and phenazine-1-carboxylate have been found to play roles in biological control (Gurusiddaiah et al., 1986; Howell and Stipanovic, 1979, 1980). Like the pyoverdine siderophores, roles for the antibiotics in the biological control phenomonenon may be contributory or central, but have not been established firmly. Genetic engineers may manipulate antibiotic production to produce more efficient antibiotics and higher levels of production.

16.5.6 COPPER AND FUNGICIDE RESISTANCE

Manipulating copper and/or fungicide resistance in non-pathogenic bacteria to enhance biological control of pathogens is a unique scheme under consideration for IPM. Copper resistance was discovered on conjugal plasmids in Pseudomonas syringae pv. tomato, the causal agent of bacterial speck of tomato (Bender and Cooksey, 1987). The presence of copper resistance interferes with the control of bacterial speck using copper sprays; however, the use of copper sprays interferes with the saprophytic barrier to disease established by non-pathogenic bacteria on the surface of tomatoes. By moving copper-resistance plasmids into non-pathogenic epiphytic bacteria, glass-house experiments have demonstrated that a saprophytic barrier to the non-copper-resistant pathogen may be established in the presence of copper sprays, thus reducing disease incidence and severity to a greater extent than that with either the biological control strain or the copper treatment separately (Cooksey, 1988). Most likely, this combined control would reduce the incidence and severity of disease caused by the copper-resistant pathogen, since the epiphytic population of the biological control agent remains to provide the saprophytic barrier. Similar studies have been conducted with benomyl (TeBeest, 1984) and other fungicidal compounds (Delp, 1980).

16.6 SUMMARY

The use of techniques developed for recombinant DNA in the biological control of plant pathogens is currently a very active area of research in both academic and industrial institutions. The application of molecular genetics to selected biocontrol systems can potentially provide an understanding of the genes involved in biological control phenomena. An establishment of a genetic basis for biological control is a prerequisite to further molecular studies that centre on the enhancement of biological control phenomena. It is not necessary to completely understand a biological control system to use it successfully in the field. However, this basic understanding is necessary to apply molecular genetics as a means to improve a biological control system. Initially, a model system that consistently works under field conditions should be exploited and genetically dissected. Biological control can be best approached by collaborations between molecular biologists and plant pathologists. Thus, the plant pathologist can provide the biological material and the practical experience, while the molecular biologist can manipulate the specific genes of interest.

An obstacle to the effective use of biocontrol agents is the complex microscopic ecosystem in which biological control agents for plant disease must survive, multiply, and decrease disease. Disease can be reduced by: a reduction in inoculum density of the pest, as with insects; protection of infection sites; or modification of the response of the host to the pathogen (i.e., induced resistance and cross protection). The latter two cases appear to be more common in plant pathology. Whether on leaves (Lindow, 1982; Blakeman, 1981) or roots (Schroth and Hancock, 1985), the control agent must be highly competitive.

The delivery of control agents is critical in plant disease control. The beneficial microorganisms must be placed in an environment where they can survive and interact with a pathogen or plant in such a manner as to reduce disease. Favourable environments are usually found where colonizaton by other microorganisms is reduced, such as young leaves (Lindow, 1982), fumigated soils (Martin et al., 1985), roots of transplants, or fresh wounds (Carroll, 1988; Schroth and Hancock, 1985).

Due to the difficulty of establishing organisms in complex communities it is important to consider the agroecosystem as a continuum: from relatively simple greenhouse systems, where large amounts of energy and chemical inputs are used to maintain a constant agroecosystem, to complex natural forest systems, where little or no inputs are available (Shikano and Kurihara, 1985). The simpler, more intensive greenhouse systems are more amenable to biological control, because control agents can be applied to steamed or fumigated soils, where competition from other microorganisms is low. The agroecosystem continuum can be extended to include annual crops transplanted to fumigated soils, annual crops seeded into fumigated soils, annual crops seeded into non-fumigated soils, perennial fruit and nut crops intensively managed, perennial crops that receive little or no management, and finally unmanaged natural systems. Due to the increase in complexity and decrease in areas available for introduction of control agents, the implementation of biological control is more difficult as one moves along the continuum (Cairns, 1986, 1987; Cairns and Pratt, 1987).

The release of engineered microorganisms into the environment has been the focus of much public and scientific concern (Rissler, 1984). Before a release is allowed, the construction of effective and safe biological control agents should be carefully analysed and the risks and benefits assessed. Collaboration among pathologists, molecular biologists and microbial ecologists is essential at this point. Such policies are already in place or are in the developmental stage in many countries throughout the world.

Lastly, to implement biological control programs based on biotechnology within IPM, the following need to be improved: understanding of total ecosystems; mass production and delivery of control agents; selection and enhancement of control agents; and the specificity of control agents to only one disease on the crop in a small geographic area. Those researching the biological control of insects share many of these same problems. The success rate of recently developed biocontrol agents (naturally occurring and genetically altered) is impressive, and demonstrates the potential that genetic engineering has on biocontrol systems. The future appears filled with exciting possibilities for applying effective biologicals to help reduce the chemical dependency of agricultural production systems.

16.7 REFERENCES

Ahl, P., Voisard, C. and Defago, G. (1986) Iron-bound siderophores, cyanic acid, and antibiotics involved in suppression of Thielaviopsis basicola by a Pseudomonas fluorescens strain. J. Phytopathol. 116, 121-124.

Amrhein, N., Johanning, D., Schab, J. and Schultz, A. (1983) FEBS Lett. 157, 191-196. 

Armstrong, J. L., Knudsen, G. R. and Seider, R. J. (1987) Microcosm method to assess survival of recombinant bacteria associated with plants and herbivorous insects. Curr. Microbiol. 15(4), 229-232.

Arntzen, C. J. (1986) Protein engineering of herbicide target sites. Abstracts of the International Congress on Pesticide Chemistry, 6th IUPAC, Ottawa 3S-04. 

Atlas, R. M. (1982) Enumeration and estimation of biomass of components in the biosphere. In: Burns, R. G. and Slater, J. H. (Eds) Experimental Microbial Ecology, Blackwell Scientific, Oxford, UK, pp. 84-102.

Baross, J. H., Liston, J. and Morita, R. Y. (1967) Some implications of genetic exchange among marine vibrios, including Vibrio parahaemolyticus, naturally occurring in the Pacific oyster. In: Fujino, T., Sakaguchi, G., Sakazaki, R. and Takeda, Y. (Eds) International Symposium on Vibrio parahaemolyticus, Saikon Publishing, Tokyo, pp.126-137.

Bender, C. L. and Cooksey, D. A. (1987) Molecular cloning of copper resistance genes from Pseudomonas syringae pv. tomato. J. Bacteriol. 169, 470-474.

Bernhard, K. (1986) Studies on the delta-endotoxin of Bacillus thuringiensis var. tenebrionis. FEMS Microbiol. Lett. 3, 261-265.

Biever, K. D., Andrews, P. L. and Andrews, P. A. (1982) Use of a predator, Podusus maculiventris, to distribute virus and initiate epizootics. J. Econ. Entomol. 75, 150-152.

Blakeman, J. P. (Ed.) (1981) Microbial Ecology of the Phylloplane, Academic Press, New York.

Brezhnev, A. I., Ginzburg, L. R., Pluektov, R. A., and Poluektov, R. A. (1974) The structural optimization of exploited populations. Econ. Math. Meth. 4, 808-811. 

Brown, J. R. (Ed.) (1987) Soil Testing. Sampling, Correlation, Calibration and Interpretation, Special Publication No. 21, Soil Science Society of America, Madison, Wisconsin, 144pp.

Bulla, L. A., Davidson, L. I, Kramer, K. J. and Jones, B. L. (1979) Purification of the insecticidal toxin form the parasporal crystal of Bacillus thuringiensis subsp. Kurstaki. Biochem. Biophys. Res. Commun. 91, 1121-1130.

Cairns, J. Jr (1986) The myth of the most sensitive species. BioScience 36, 670-672. 

Cairns, J. Jr (1987) Problems associated with selecting the most sensitive species for toxicity testing. Hydrobiologia 153, 87-94.

Cairns, J. Jr and Pratt, J. R. (1987) Ecotoxial effect indices: a rapidly evolving system. Water Sci. Tech. 19(11)1, 1-12.

Carroll, G. (1988) Fungal endophytes in stems and leaves: from latent pathogen to mutalistic symbiont. Ecology 69, 2-9.

Chaleff, R. S. and Mauvais, C. J. (1984) Acetolactate synthase is the site of action of two sulfonylurea herbicides in higher plants. Science 224, 1443-1445.

Cocking, E. C., Davey, M. R., Pental, D. and Power, J. B. (1981) Aspects of plant genetic manipulation. Nature 293, 265-270.

Comai, L., Facciotti, D., Stalker, D. M., Thompson, G. A. and Hiatt, W. R. (1986) Expression in plants of a bacterial gene coding for glyphosate resistance. In: Zaitlin, M., Day, P. and Hollaender, A. (Eds) Biotechnology in Plant Science. Relevance to Agriculture in the Eighties, Academic Press, New York, pp. 329-338.

Cooksey, D. A. (1988) Reduction of infection by Pseudomonas syringae pb. tomato using a nonpathogenic, copper-resistant strain combined with a copper bactericide. Phytopathology 78, 601-603.

Corapciolu, M. Y. and Haridas, A. (1984) Transport and fate of microorganisms in porous media: a theoretical investigation. J. Hydrol. (Amsterdam) 72, 149-169. 

DeAngelis, D. L. (1973) A general code for ecosystem models. Gött Bodenkdl. Ber. 29, 55-92.

Delp, C. J. (1980) Coping with resistance to plant disease control agents. Plant Dis. 64, 652-657.

Falco, S. C., Dumas, S. K. and McDevitt, R. (1985) Molecular genetic analysis of sulfonylurea herbicide action and resistance in yeast. In: van Vloten-Doting, L., Groot, G. S. P. and Hall, T. C. (Eds) Molecular Form and Function of the Plant Genome, NATO ASI Series A, Life Sciences Vol. 83, Plenum, New York, pp. 467-478. 

Falcon, L. A. (1985) Development and use of microbial insecticides. In: Hoy, M. A. and Herzog, D. C. (Eds) Biological Control in Agricultrual IPM Systems, Academic Press, New York, pp. 229-242.

Falcon, L. A., Kane, W. R. and Bethell, R. S. (1968) Preliminary evaluation of a granulosis virus for control of the codling moth. J. Econ. Entomol. 61, 1208-1213.

Fraley, R. T., Rogers, S. G. and Horsch, R. B. (1986) Genetic transformation on higher plants. CRC Crit. Rev. Plant Sci. 4, 1-46.

Gerlach, W. (1988) Plant virus resistance based on the satellite RNA of tobacco ringspot virus. J. Cell. Biochem. 12C, 239 (abstr.).

Gerlach, W. L., Llewellyn, D. and Haseloff, J. (1987) Construction of a plant disease resistance gene from the satellite RNA of tobacco ringspot virus. Nature 328, 802-805.

Gladden, J. and Ginzburg, L. R. (1982) Variation and persistence in stochastic population models with differing age structures. Bull. Ecol. Soc. Am., 63, 139.

Graham, J. B. and Istock, C. A. (1978) Genetic exchange in Bacillus subtillis in soil. Mol. Gen. Genet. 166, 287-290.

Gurusiddaiah, S., Weller, D. M., Sarkar, A. and Cook, R. J. (1986) Characterization of an antibiotic produced by a strain of Pseudomonas fuorescens inhibitory to Gaeumannomyces graminis var. tritici and Pythium spp. Antimicrob. Agents Chemother. 29, 488-495.

Hagedorn, C. (1986) Role of genetic variants in autecological research. In: Tate, R. L. (Ed.) Microbial Autecology: A Method for Environmental Studies, John Wiley & Sons, New York, pp. 61-74.

Harrison, B. D., Mayo, M. A. and Baulcombe, D. C. (1987) Virus resistance in transgenic plants that express cucumber mosaic virus satellite RNA. Nature 328, 799-802. 

Hatzios, K. K. (1987) Biotechnology applications in weed management: now and in the future. Adv. Agron. 41, 325-375.

Hernstadt, C., Soares, G., Wilcox, E. and Edward, D. (1986) A new strain of Bacillus thuringiensis with activity against coleopteran insects. BioTechnology 4, 305-308. 

Hernstadt, C., Gilroy, T. E., Sobieski, D. A., Bennett, B. D. and Gaertner, F. H. (1987) Nucleotide sequence and deduced amino acid sequence of a coleopteran active -endotoxin gene from Bacillus thuringiensis subsp. San Diego. Gene 57, 37-46. 

Hicks, R. E. and Newell, S. Y. (1984) The growth of bacteria and the fungus Phaesphaeria-typharum Eumycota Ascomycotina in salt marsh microcosms in the presence and absence of mercury. J. Exp. Mar. Biol. Ecol. 78(1-2), 143-156. 

Horsch, R. B., Fraley, R. T., Rogers, S. G., Sanders, P. R., Lloyd, A. and Hoffman, N. (1984) Inheritance of functional foreign genes in plants. Genetic engineering with T-DNA from Agrobacterium tumefaciens, Nicotiana plumbaginifolia. Science 223, 496-498.

Howell, C. R. and Stipanovic, R. D. (1979) Control of Rhizoctonia solani on cotton seedlings with Pseudomonas fluorescens and with an antibiotic produced by the bacterium. Phytopathology 69, 480-482.

Howell, C. R. and Stipanovic, R. D. (1980) Suppression of Pythium ultimum-induced damping-off of cotton seedlings by Pseudomonas fluorescens and its antibiotic, pyoluteorin. Phytopathology 70, 712-715.

Hunter, M., Stephenson, T., Kirk, P. W. W., Perry, R. and Lester, J. N. (1986) Effect of salinity gradients and heterotrophic microbial activity on biodegradation of nitrilotriacetic-acid in laboratory simulations of the estuarine environment. Appl. Environ. Microbiol. 51(5), 919-925.

Jansson, J. K. and Tiedje, J. M. (1988) Detection of gene rearrangement in soil using gene probes. In: M. Sussman et al. (Eds), The Release of Genetically Engineered Microorganisms, Academic Press, New York.

Kelman, A., Anderson, W., Falkow, S., Federoff, N. V. and Levin, S. (1987) Introduction of Recombinant DNA-Engineered Organisms into the Environment: Key Issues. Prepared for the National Research Council of the National Academy of Sciences by the Committee on the Introduction of Genetically Engineered Organisms into the Environment, National Research Council, Washington DC.

Kerr, A. Manigault, P. and Tempe, J. (1977) Transfer of virulence in vitro and in vivo in Agrobacterium. Nature 265, 560-561.

Kim, H. K., Orser, C., Lindow, S. E. and Sands, D. C. (1987) Xanthomonas campestris pv. translucens strains active in ice nucleation. Plant Dis. 71, 994-997.

King, A. W. and DeAngelis, D. L. (1986) Site-Specific Models of Seasonal Carbon Fluxes in Terrestrial Biomes, Report No. ORNL/TM-9749, Oak Ridge National Laboratory, Oak Ridge, Tennessee.

Klute, A. (Ed.) (1985) Methods of Soil Analysis, Part 1 (2nd edn), Special Publications No. 9, American Society of Agronomy, Madison, Wisconsin, 1188pp.

Kool, J. B., Parker, J. C. and Van Genuchten, M. T. H. (1987) Parameter estimation for unsaturated flow and transport modelsa review. J. Hydrol. 91, 255-293. 

Krieg, V. A., Huger, A. M., Langenbruch, G. A. and Schnetter, W. (1983) Bacillus thuringiensis var. tenebrionis, a pathotype effective against larvae of Coleoptera. Z. Angew. Entomol. 96, 500-508.

Lacy, G. H. (1978) Genetic studies with plasmic RPI in Erwinia chrysanthemi strains pathogenic on maize. Phytopathology 68, 1323-1330.

Lacy, G. H. and Leary, J. V. (1975) Transfer of antibiotic resistance plasmid RPI into Pseudomonas glycinea and Pseudomonas phaseolicola in vitro and in plants. J. Gen. Microbiol. 88, 49-57.

Lacy, G. H., Stromberg, V. K. and Cannon, N. P. (1984) Erwinia amylovora mutants and in planta-derived transconjugants resistant to oxytetracycline. Can. J. Plant Pathol. 6, 33-39.

Leisinger, T. and Margraff, R. (1979) Secondary metabolites of the fluorescent pseudomonads. Microbiol. Rev. 43, 422-442.

Leong, J. (1986) Siderophores: their biochemistry and possible role in the biocontrol of plant pathogens. Annual Rev. Phytopathol. 24, 187-209.

Lincoln, D. R., Fisher, E. S. and Lambert, D. C. (1985) Release and containment of microorganisms from applied genetics activities. Enzyme Microb. Technol. 7(7), 314-321.

Lindow, S. E. (1982) Epiphytic ice nucleation-active bacteria. In: Mount, M. S. and Lacy, G. H. (Eds) Phytopathogenic Prokaryotes, Vol. 1, Academic Press, New York, pp.335-362.

Lindow, S. E. (1985) Ecology of Pseudomonas syringae relevant to the field use of ice-deletion mutants constructed in vitro for plant frost control. In: Halvorson, H. O., et al. (Eds) Engineered Organisms in the Environment: Scientific Issues, American Society for Microbiology, Washington DC, pp. 23-35.

Makulowich, J. S. (1988) Agbio suddenly captures industry attention. Genet. Engin. News 8(47), 1, 28-29.

Martin, S. B., Abawi, G. S. and Hoch, H. C. (1985) Biological control of soil-borne pathogens and antagonists. In: Hoy, M. A. and Herzog, D. C. (Eds) Biological Control in Agricultural IPM Systems, Academic Press, New York, pp. 433-454.

Marugg, J. D., van Spanje, M., Hoekstra, W. P. M., Schippers, B. and Weisbeek, P. J. (1985) Isolation and analysis of genes involved in siderophore biosynthesis in plant-growth-stimulating Pseudomonas putida WCS358. J. Bacteriol. 164, 563-570.

Mets, L. J., Galloway, R. E. and Erickson, J. M. (1986) Prospects for genetic modification of plants for resistance to triazine herbicides. In: Zaitlin, M., Day, P. and Hollaender, A. (Eds) Biotechnology in Plant Science: Relevance to Agriculture in the Eighties, Academic Press, New York, pp. 301-312.

Milewski, E. A. (1985) Field testing of microorganisms modified by recombinant techniques: applications, issues, and development of `points to consider'. Recomb. DNA Tech. Bull. 8, 102-108.

Minogue, K. P. and Fry, W. E. (1983) Models for the spread of disease: model description. Phytopathology 73, 1168-1173.

Morley, C. R., Trofymow, J. A., Coleman, D. C. and Cambardella, C. (1983) Effects of freeze thaw stress on bacterial populations in soil microcosms. Microb. Ecol. 9(4), 329-340.

Morrison, W. D., Miller, R. V. and Saylor, G. S. (1978) Frequency of F-116 mediated transduction of Pseudomonas aeruginosa in a freshwater environment. Appl. Environ. Microbiol. 36, 724-730.

Napoli, C. and Staskawicz, B. J. (1985) Molecular genetics of biological control agents of plant pathogens; status and prospects. In: Hoy, M. A. and Herzog, D. C. (Eds) Biological Control in Agricultural IPM Systems, Academic Press, New York, pp. 455-461. 

Nelson, R. S., McCormick, S. H., Delannay, X., Dube, P., Layton, J., Anderson, E. J., Kaniewska, M., Proksch, R. K., Horsch, R. B., Rogers, S. G., Fraley, R. T. and Beachy, R. N. (1988) Virus tolerance, plant growth, and field performance of transgenic tomato plants expressing coat protein from tobacco mosaic virus. Biotechnology 6, 403-409.

Orser, C., Staskawicz, B. J., Panopoulos, N. J., Dahlbeck, D. and Lindow, S. E. (1985) Cloning and expression of bacterial ice nucleation genes in Escherichia coli. J. Bacteriol. 164, 359-366.

Orvos, D. R., Lacy, G. H. and Cairns, J. Jr (1988) Origin of antibiotic resistant strains isolated from microcosms contaminated with genetically-engineered microorganisms. In: M. Sussman et al. (Eds) The Release of Genetically-engineered Microorganisms, Academic Press, New York.

Ou, C. Y., Kwok, S., Mitchell, S. W., Mack, D. H., Sninsky, J. J., Krebs, J. W., Feorino, P., Warfield, D. and Schochetman, G. (1988) DNA amplification for direct detection of HIV-1 in DNA of peripheral blood mononuclear cells. Science 239, 295-297. 

Ouellette, R. P. and Cheremisinoff, P. N. (1985) Applications of Biotechnology, Technomic Publishing, Lancaster, Pennsylvania, 247pp.

Page, A. L., Miller, R. H. and Keeney, D. R. (Eds) (1982) Methods of Soil Analysis, Part 2 (2nd edn), Special Publication No. 9, American Society of Agronomy, Madison, Wisconsin, 1159pp.

Palukaitis, P. and Zaitlin, M. (1984) A model to explain the 'cross-protection' phenomenon shown by plant viruses and viroids. In: Kosuge, T. and Nester, E. (Eds) PlantMicrobe Interactions. Molecular and Genetic Perspectives, Macmillan, New York, pp. 420-429.

Ponz, F. and Bruening, G. (1986) Mechanisms of resistance to plant viruses. Annual Rev. Phytopathol. 24, 355-381.

Rissler, J. (1984) Research needs for detecting environmental effects of genetically engineered microorganisms. Recomb. DNA Tech. Bull. 7, 20-30.

Roberge, M. R. (1978) Methodology of soil enzyme measurement and extraction. In: Burns, R. G. (Ed.) Soil Enzymes, Academic Press, New York, pp. 341-375. 

Roszak, D. B. and Colwell, R. R. (1987) Survival strategies of bacteria in the natural environment. Microb. Rev. 51, 365-379.

Rouse, D. I. (1985) Construction of temporal models: I. Disease progress of air-borne pathogens. Adv. Plant Pathol. 3, 11-29.

Saiki, R. K., Glenfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B. and Erlich, H. A. (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487-494.

Schroth, M. N. and Hancock, J. G. (1985) Soil antagonists in IPM systems. In: Hoy, M. A. and Herzog, D. C. (Eds) Biological Control in Agricultural IPM Systems, Academic Press, New York, pp. 415-432.

Shah, D. M., Horsch, R. B., Klee, H. J., Kishore, G. M., Winter, J. A., Turner, N. E., Horinaka, C. M., Sanders, P. R., Glasser, C. S., Aykent, S., Siegel, N. R., Rogers, S. G. and Fraley, R. T. (1986) Engineering herbicide tolerance in transgenic plants. Science 233, 478-481.

Shaner, D. L., Anderson, P. C. and Stidham, M. A. (1984) Imidazolinones: potent inhibitors of acetohydroxyacid synthase. Plant Physiol. 76, 545-546.

Shikano, S. and Kurihara, Y. (1985) Community responses to organic loading in a microcosm. Jpn. J. Ecol. 35(3), 297-306.

Silver, M., Ehrolich, H. L. and Ivarson, K. C. (1986) Soil mineral transformation mediated by soil microbes. In: Huang, P. M. and Schnitzer, M. (Eds) Interactions of Soil Minerals with Natural Organics and Microbes, Special Publication No. 17, Soil Science Society of America, Madison, Wisconsin, pp. 497-514.

Somerville, C. C., Knight, I. T., Straube, W. L. and Colwell, R. R. (1988) Probe-directed polymerization-enhanced detection of specific gene sequences in the environment. In: M. Sussman et al. (Eds), The Release of Genetically Engineered Microorganisms, Academic Press, New York.

Summers, M., Engler, R., Falcon, L. A. and Vail, P. V. (1980) Antibiotic resistance and its transfer among clinical and nonclinical Klebsiella strains in botanical environments. Appl. Environ. Microbiol. 39, 97-104.

Talbot, H. W., Yamamoto, D. Y., Smith, M. W. and Seidler, R. J. (1980) Antibiotic resistance and its transfer among clinical and nonclinical Klebsiella strains in botannical environments. Appl. Environ. Microbiol. 39, 97-104.

Tauber, H. J., Hoy, M. A. and Herzog, D. C. (1985) Biological control in agricultural IPM systems: a brief overview of the current status and future prospects. In: Hoy, M. A. and Herzog, D. C. (Eds) Biological Control in Agricultural IPM Systems, Academic Press, New York, pp. 3-12.

TeBeest, D. O. (1984) Induction of tolerance to benomyl in Colletotrichum gloeosporioides f. sp. aeschynomene by ethyl methanesulfonate. Phytopathology 7-4, 864 (Abstract).

Thingstad, T. F. and Pengerud, B. (1985) Fate and effect of allochthonous organic material in aquatic microbial ecosystems: an analysis based on chemostat theory. Mar. Ecol. Prog. Ser. 21(1-2), 47-62.

Tomashow, M. F., Nutter, R., Postle, K., Chilton, M. D., Blattner, F. R., Powell, A., Gordon, M. P. and Nester, E. W. (1980) Recombination between higher plant DNA and the Ti plasmid of Agrobacterium tumefaciens. Proc. Nat. Acad. Sci. USA 77, 6448-6452.

Weinberg, S. R. and Stotzky, G. (1972) Conjugation and genetic recombination of Escherichia coli in soil. Soil Biol. Biochem. 4, 171-180.

Weisbeek, P. I. J., Schippers, B., van der Hofstad, G. and Marugg, J. (1986) Molecular genetics of siderophore biosynthesis in Pseudomonas putida WCS358. In: Swinburne, T. R. (Ed.) Iron, Siderophores and Plant Diseases, NATO ASI Series A, Life Sciences Vol. 117, Plenum Press, New York.

Whitley, H. R. and Schnepf, H. E. (1986) The molecular biology of parasporal crystal body formation in Bacillus thuringiensis. Ann. Rev. Microbiol. 40, 549-576.