Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
10 Properties of the Environment Relevant to the Introduction of Genetically Modified Microorganisms Successful establishment of a specific population depends on two components: the organism and the environment. Although most discussion of genetically modified microorganisms has focused on the properties of the organisms, properties of the environment are equally important. Microbial environments often are less well understood than those of higher organisms because of the difficulties in defining, measur- ing, and controlling various physical and chemical details of the microenvironment important in establishing introduced microorgan- isms. More studies of microbial interactions with the environment are needed. However, we have considerable knowledge about cer- ta~n microorganisms, such as Rhizobium and mycorrh~zal fungi that provide import ant insights into the interactions in question. TYPES OF ENVIRONMENTS Environments that are to receive introduced m~croorg~sms may vary considerably with respect to their biological, chemical, and geological properties, and these properties may vary with physico- chemical changes. Hence, our ability to establish a level of certainty about the risks and benefits varies with our knowledge and expe- rience with the particular site where the microorganisms are to be 113
114 introduced. Scientists have extensive experience with major agri- cultural soil-crop environments, much less experience with tropical forests, and little experience with the open ocean. Predictability about the fate and effects of introduced microorganisms increases with experience. The ease and reliability with which the introduced m~croorgan- ism can be confined also depend on the environment into which it is introduced. Microorganisms introduced onto surface soil not sum ject to excessive wind and rainfall are obviously more easily confined to the test site than are microorganisms added to a site subject to flooding and excessive wind erosion. Special features of the environment may be important in decid- ing whether an introduction is advisable. Most introductions are intended for sites that are far from pristine. At hazardous waste sites or In streams made acid by mine drainage, a microbial introduction will promise far more benefit than risk. Microorganisms designed for removal of toxic pollutants would be unlikely to flourish outside the site of introduction. HABITABILITY OF ENVIRONMENTS Four characteristics of the environment that determine habitabil- ity of an area for introduced microorganisms are (1) nutrient status, (2) toxic chemicals and metabolites, (3) physicochem~cal factors, and (4) biological factors. Nutrient Status Energy supply often limits the growth of microbial populations. Organic compounds represent the major energy source for most ge- netically modified microorganisms currently berg studied. Light or reduced inorganic compounds also can supply energy for some microorganisms. Although population density and community di- versity usually parallel the organic carbon concentration in a habi- tat, the competition for these carbon substrates and the diversity of substances are also important. Whereas energy usually limits het- erotrophic populations, inorganic nutrients may limit others. For example, algal blooms are typically limited by phosphate in fresh- water environments and by nitrogen in the open ocean. Carbon and nutrient resources are Innited in nature, and the growth of introduced microorganisms cannot exceed these resource limits.
115 Tactic Chemicals and Metabolites Extreme concentrations of heavy metals, acids, and organic pollutants can be toxic to microbial cells through their effects on metabolic processes. Thus, toxic compounds may decrease microbial population density and limit community diversity to microorganisms resistant to elevated concentrations of heavy metals and organic pol- lutants. Bacteria resistance to heavy metal toxicity can indicate whether certain metals are or have been present in a given environ- ment (Olson and Thornton, 1982; Olson and Barkay, 1986; Zelibor et al., 1987~. Specialized rn~croorganisms often successfully colonize these stressed habitats, but may be relatively less competitive in nonstressed environments (Konings and Vel~kamp, 1980~. Physicocherrucal Factors Environmental chemucal variables (pit, oxidation/reduction po- tential, nutrients, toxicants, salinity) and physical variables (light, surfaces, temperature) influence the diversity of microbial commu- nities (Stotzky and Babich, 1986~. Environmental factors such as moisture, temperature, and oxygen can vary over wide ranges. Biological Factors Microbial predators, parasites, symbionts, and competitors con- tribute significantly to microbial community structure. Experience from microbial introductions into soil and water environments typi- cally has shown that it is difficult to establish introduced populations at densities sufficient to achieve the desired effect, such as nitrogen fixation in rice paddies (Ready and Roger, 1988) and biocontro! of pests (Bahme and Schroth, 1987~. The lack of success of some micro- bial introductions may result from competition with the indigenous community or introduction into unfavorable habitats. DISPERSAL An important issue for the environmental introduction of m~- croorgan~sms is the extent to which the microorganism and its progeny are dispersed from the application site. Dispersal provides a route of entry for rn~croorganisms to new habitats. Prior experi- ence has revealed no problems that have arisen because introduced
116 microorganisms became established beyond their area of mtroduc- tion, with the exception of pathogens. Even with plant pathogens, carefully designed field tests are routinely performed which result in no or negligible damage to neighboring crops (Tolin and Vidaver, 1989~. Exceptions have occurred, usually in the field testing or in- troduction of plant pathogens that are not indigenous to the area. Microorganisms can be dispersed from terrestrial field plots during their application, or by leaching through soil, runoff in surface wa- ter or on soil particles, dissemination by Anna (dust particles), and transport from the plot by Sumacs, humans, and field machinery. Soil generally filters microorganisms effectively, and the motility of soil rn~croorganisms does not support extensive movement. Most rn~croorganisms in soil are firmly attached to soil particles, so that any movement of the soil particle moves the attached organ- ism as well (Faust, 1982~. Attachment often aids microbial survival, because particles can protect organisms from ultraviolet radiation during aerial transport (Stetzenbach, 1989) and from predation in soil and water (Roper and Marshall, 1974~. Recently introduced organisms may be less likely to attach to the soil, and they may move by saturated flow (Rake et al., 1978; Smith et al., 1985~. Laboratory studies with sieved and repacked soil cores tend to underestimate microbial movement by leaching compared with that found in undisturbed soil cores, because natural soil structure has more macrop ores and connecting channeb that reduce its effectiveness as a filter. Groundwater and surface water environments may furnish sim- ilar habitats and harbor similar heterotrophic microorganisms. Di- gestive tracts of insects, birds, and other an~rnab provide a habitat quite different from soil, but they may support soil organisms and disperse them with fecal material (Reyes and Tie~je, 1976~. If the major dispersal mechanisms are known, dispersal can often be effectively ev~uatec! and controlled. Selection of level sites for field tests and construction of terraces should virtually elirn~nate surface runoff. Selection of a site distant from groundwater and application of the organisrrLs to minimize individual cell movement will reduce leaching to groundwater. Fencing and netting can be used to control animals. Controlling insect dispersal of the introduced mucroorganism may be difficult, but such dispersal often can be minimized by the choice of suitable plants and through the use of appropriate insecticides.
117 SmalI-scale tests ~ estuarine, marine, and other aquatic environ- ments pose special problems of confinement. The use of membrane chambers offers confinement (McFeters and Stuart, 1972; Mach and Barnes, 1982; Grimes and ColweD, 1986), but the ability of ma- rine microorganisms to shrink during starvation (Morita, 1982) or to enter a dormant state (R. R. Colwell et al., 1985; Grimes et al., 1986; Roszak and ColweD, 1987) requires that membrane chambers have sufficiently small pore size to retain cells but still permit nutri- ents to enter the chamber. Bubble containment devices placed in the aquatic environment are not reliable protection against dispersal (Grice and Reeve, 1982), and tests of genetically modified rn~croor- ganisms that Ought be performed using them should be carefully eval- uated. Laboratory microcosms and mesocosms are the most suitable compromise (Pritchard and Bourquin, 1984; Cripe and Pritchard, 1989~. In estuarine, coastal, and open ocean systems, the ejects of wind, tides, and currents, as well as dispersal by fish, birds, aquatic plants, ~d other organ~srns, must be considered. SUITABILITY OF MICROCOSMS FOR TESTING OF MICROBIAL INTRODUCTIONS It is widely accepted that aquatic and terrestrial laboratory microcosms are useful for examining the fate and effects of introduced microorganisms as wed as their survival and persistence in specific environments. The definition of a rn~crocosm which can be adapted for purposes of discussion here is I. . . an intact, m~nnnally disturbed piece of an ecosystem brought Alto the laboratory for studyn (Cripe and Pritchard, 1989, p. 1~. Thus, a microcosm can be used to relate laboratory data to the site where the environmental samples were taken (Greenberg et al., 1988) as it is a site- and system-specific construct. In principle, the microcosm is an intact piece of the field that behaves ecologically like its counterpart in the actual field (Pritchard and Bourquin, 1984~. A variety of laboratory test systems have been designed to mode! the environment. These include synthetic communities, with weD- character~zed organisms placed In sterile media under defined en- vironmental conditions. In other systems, natural samples may be incubated over long periods such that a unique and sustaining ecosys- tem evolves (Greenberg et al., 19883. Results of several studies attest
118 to the value of studying ecological processes In microcosms and ex- trapolating the information to ecosystems in nature (Livingston et al., 1985; Diaz et al., 1987~. Most use of microcosms has been applied to the study of toxic substances in the environment (Gillett and Witt, 1979; Giesy, 1980; Hammonds, 1981; Cairns et al., 19813. Synthetic communities have proven useful in studying the fate of xenobiotic compounds (foreign chemicals) in aquatic systems (Isensee and Tayaputch, 1986; Metcalf et al., 1971) as well as their effects on biological communities (Crow and Taub, 1979; Levier, 1984~. Artificial communities may be lim- ited, however, ~ that they lack complex population structures and may not function like those occurring in natural ecosystems (Cripe and Pritchard, 1989~. Some of the ecological processes quantified in microcosms include nutrient leaching (Van Voris et al., 1980, 1983), nutrient cycling (Harte et al., 1980), predator-prey interaction (Gillett et al., 1983), primary production (Harte et al., 1980), and microbial respiration (I.ighthart et al., 1982; Taub and Crow, 1980~. Mode] ecosystems, although often criticized as ecologically sim- plistic, have been heavily used in assessments of pesticides and toxic substances, as In the constructed ~farm-pond" system of Metcalf et al. (1971~. If it had been used decades earlier, such a mode! might well have suggested possible adverse ecological consequences of the use of chlorinated hydrocarbon pesticides (Gillett et al., 1985~. Mocle} ecosystems also may be useful for monitoring the colonization and persistence of genetically modified microorganisms. S}ngle-species tests, synthetic communities, and rn~crocosms provide three pre- liminary field-trial assessments of ecological effects. Microcosms, if operated in a manner sunulating the field site, sometimes may be used as surrogates for field research, with reduced effort and cost (Cripe and Pritchard, 1989~. If the measurements taken to analyze a microcosm are the same as those used in the field, rn~crocosms can be used to establish the sensitivity and appropriateness of analytical methods pertinent to field tests. Microcosm studies are most useful ~ situations when their per- formance is needed to clarify questions about unfamiliar irtroduc- tions. They are not necessary in cases where experience and scientific inference provide enough information to permit scientists to under- stand and be familiar with the intended introduction. Furthermore, the microcosm is a site- and system-specific model; specific processes
119 may be of greater or lesser importance depending on the site and eco- logical system (Cripe and Pritchard, 1989~. Effects that are strongly scale-dependent may be overlooked in microcosm studies. In instances when such studies are deemed to be appropriate, questions of biological containment and certain environmental effects can be addressed. These include persistence in a given environment, transfer of genetic material to other organisms, population density and community structure, changes in heterotrophic activity, and nutrient cycling (Cripe and Pritchard, 1989~. Gillett et al. (1985) concluded that microcosm technology should receive a high priority for assessment of both hazard and exposure. SCALE AND FREQUENCY OF INTRODUCTION ~ this report we cover small-scale field tests and not large-scale introductions of organisms. The report sponsored by the Ecological Society of America (Tie~je et al., 1989), documents the importance of scale and encourages the use of smaD-scale field tests, when am propriate, to evaluate the potential for larger scale environmental ejects. POTENTIAL EFFECTIVENESS OF MONITORING Many planned introductions of genetically modified m~croor- ganisms should include appropriate methodology for monitoring the released microorganisms ~ and around the test site. Monitoring is rnportant for several reasons: (1) understanding the basis for the organism's effectiveness, (2) detecting any unexpected spread, and (3) building a data base on survival, spread, genetic stability, and ecological effects of genetically modified rn~croorganisms in nature. Lack of efficient recovery of the microorganisms and insensitive assays are often obstacles to monitoring microbial populations in- troduced into the environment. The classical plate-count method or similar culture methods are still the mainstays of monitoring pro- tocols. Stressed or dormant organisms may not be recovered by culturing (Roszak and Colwell, 1987~. The most common markers used for tracking microbial populations are antibiotic resistances; a typical lower limit for detection Is In the range of 103 organisms per grain of soil. At this level of detection, for a 1-hectare field site, to a depth of 10 centimeters, about 10~2 microorganisms could survive and yet be undetectable. Thus, it often has little meaning to ar- gue whether an introduced organism dies out completely. Rather,
120 it is more important to focus on whether a residual population can multiply under the environmental conditions expected at the site of testing. More sensitive and less costly monitoring methods are needed. Reliance on antibiotic-res~stance as ~ selective marker has been ex- tensive, but spreading antibiotic-resistant strains should be discour- aged. The advantages and disadvantages of selective culturing as well as antibody and nucleic acid hybridization methods for monitor- ing introduced transgenic organisms in the environment have been sumrnar~zed elsewhere (Tie~je, 1987; R. K. Colwell et al., 1988~. The nucleic acid methods offer the most specificity, require no prior culturing, have high potential sensitivity, and thus, recently, have received the most attention. These methods include (1) detec- tion by DNA-DNA hybridization of unique ribosomal sequences in total DNA extracted from communities (Attwood et al., 1988~; (2) detection by microscopy of cells containing a DNA floor hybridized to unique ribosomal RNA sequencer in cells (Giovianni et al., 19883; (3) detection by DNA-DNA hybridization of cloned genes in the total DNA extracted from soil and seawater communities (Holben et al., 1988; R. R. ColweH et al., 1988; Somerville et al., 1989~; (4) detec- tion of unique but native sequences in the tote DNA of communities (SteBan et al., 1988~; and (5) detection by DNA hybridization after DNA amplification by polymerase chain reaction to improve sensi- tivity (Steffan and Atlas, 1988~. Nucleic acid methods have been combined with culturing methods, such as the most probable num- ber method (Fredrickson et al., 1988) and plate counts (Ogram and Sayler, 1988) to improve specificity and sensitivity when culturing is not a limitation. Polyclonal antibody methods have long been used in micro- bial ecology (BohIoo! and Schmidt, 1980), and recently monoclonal methods have been used to improve specificity (DeMaag] et al., 1989; Wright et al., 1986~. Although these methods are excellent when used to study the ecology of the indigenous community, they may not always distinguish between the introduced organism and its indigenous close relatives unless combined with nucleic acid method- ology to track the specific genetic material. MITIGATION Microbial habitats vary in the ease and effectiveness with which unwanted effects from introduced organisms can be mitigated. In
121 most natural environments it is possible to reduce populations but Biscuit to el~rninate an introduced rn~croorgan~sm completely if it becomes established. Control methods have been summarized by Vidaver and Stotzky (1989~. For terrestrial environments, the methods include furn~gation, sealing the soil surface, or otherwise making the Introduced rn~croorganism's environment unfavorable for survival. Some fumigants are selective for fungi and do not kill bact~ ria, but methyl bromide has proven effective in controlling plant pathogenic bacteria as well as fungi. Its effectiveness depends on soil texture, moisture, and the depth at which the organisms are located, proper use of methyl brorn~de under a tarpaulin should control most introduced fungi ~d bacteria. Many antibiotics and nonvolatile or- ganic biocides are ineffective in soil because they are readily bound to soil material and cannot be minced electively throughout the soil volume. The environment may be made inhospitable to the introduced organisms by flooding the soil to create anaerobic conditions, altering the pH by adding lime or suiphur, or destroying the plant vegetation by burning or other means. Effort should put into characterizing and preventing risk, so that mitigation plans are only a secondary means of environmental protection. Unfortunately, few if any of these methods are applicable to aquatic ecosystems, as estuarme and marine systems are driven by tides and currents, and Ekes are subject to substantial mung; material cannot be confined after its introduction. SIJ~A:RY POINTS 1. The persistence and effects of an introduced microorga~- ism depend on features of the environment as well as on phenotypic properties of the organism. Some environments are better under- stood than others; for example, we know more about agricultural fields than natural ecosystems. 2. The ease and reliability with which a particular introduced rn~croorganism can be confined are important considerations in choosing a target environment for a field test. The potential for dispersal of introduced organisms, their progeny, or their genes by exchange with indigenous organisms must be considered.
122 3. Environmental features that will affect the likelihood of per- sistence of an introduced microorganism include nutrient avaalabil- ity; physicochemical factors such as pH, temperature, and inhibitory chemicals; and biological factors such as competitors and predators. 4. Although documented examples of introduced microorgan- isms that have measurably or adversely altered ecosystem processes are not available, unfamiliar microorganisms should be studied care- fully first in the laboratory and then ~ smaD-scale field tests before they are introduced on a broader scale. 5. Microcosms are minimally disturbed Pieces of natural ecosystems that are brought into the laboratory, and their use in appropriate cases may provide useful information for evaluating the survival and impact of proposed ~rucrobial field tests. 6. Small-scale field research is an important step in the inves- tigation of properties of a particular microorganism intended even- tuaDy for environmental application. Hence, field research should be encouraged after appropriate investigations have been conducted in physically confined settings and after appropriate methods have c been considered for monitoring and controlling the introduced m~- croorg~ism.