For centuries, nonindigenous plant species have been introduced into global ecosystems. Some have remained relatively harmless and some have spread at alarming rates. Such invasive species have been known to outcompete wild plants where they spread. A central concern of many opponents of genetic engineering is that GE plants will invade their surroundings and choke out wild plants because of their favorable engineered traits. A recent ten year study of GE crops in natural habitats demonstrated that the studied GE crops were never more invasive or persistent than their concentionally bred counterparts. Both plants engineered for herbicide tolerance and for insect tolerance were studied in a total of twelve different habitats. The researchers for this study were reluctant to extrapolate what these findings could mean for all GE crops, though. They noted that these specific traits were not intended to make the engineered plants more fit in their habitats. The researchers stated that studies on GE plants with other introduced traits and in more natural habitats would have to be conducted in order to understand the potential for GE plants to become successful weeds in their surroundings (Crawley et al., 2001). Another study by Bergelson demonstrated that more factors than the specific selected for traits need to be considered when determining the likelihood of a GE plant to become a weed. These factors include dynamic environmental factors and specific methods of cultivation (Bergelson, 1994). These findings demonstrate the lack of overall knowledge regarding the invasiveness of GE plants. They also demonstrate the unpredictability of crop invasiveness beyond testing situations. For these reasons, more comprehensive and varied research is necessary to determine whether GE plants are more likely to become invasive weeds than their conventionally bred counterparts.

Critcs of the commercial cultivation of GE crops also worry that the GE plants will hybridize with their wild neighbors and so pass on their engineered traits to these wild plants. The consequences of such cross-pollination could, in theory, be profound. Wild neighbors could become the weeds that choke out other plants or they could resist insects and so disturb nontarget organisms. A 1999 study of these concerns demonstrates that wild plants can, in fact, hybridize with GE plants and then produce progeny with the selected trait of herbicide-resistance. These hybridized plants are then able to persist and reproduce -- passing on this trait. The study found that the hybridized plabnts fared only aas well as their conventional, wild counterparts in laboratory conitions (Snow et al., 1999). Ths study was done in laboratory conditions and with one type of GE plant. The conclusions of the study are therefore limited. Clearly, the potential for the passage of engineered traits to wild plants exists and must be a concern for both researchers and farmers. It must play a role in how GE plants are cultivated and what types of engineered traits are released into certain areas of the world. More research must be conducted to determine exactly how such hybridization could affect wild plants in natural habitats.

Because scientists combine an antibiotic-resistant gene with the selected trait gene when they introduce new genes into a plant, many people are concerned that this widespread use of antibiotic-resistance will contribute to the global resistance of pathogens to antibiotics. This fear may be largely unwarranted because the marker that is used is usually a common marker used in many different types of genetic testing. This marker is resistant to "old" antibiotics. Most human pathogens have already evolved strains of resistance to these "old" antibiotics. Researcers contend that the use of this marker therefore is not a direct danger to human health. Moreover, reseacrhers have found that the transfer of this resistant gene to a human pathogen that does not already have it is highly unlikely (Gassen). Once again, these findings relate only to the comon markers that are used. If new markers are to be considered for use in gene insertion, more research needs to be conducted.

An important concern of opponents to genetic engineering is that the selected traits expressed in GE plants will have strong effects on nontarget organisms. A 1998 study found that 62% of chrysopid larvae raised on diets of prey that had been fed Bt insect-resistant corn died. In contrast, 37% of the larvae that were raised on prey that had been fed corn without the Bt insect-resistant gene died (National Research Council, 2000). Another study found that these Bt genes that were expressed through toxicity to target insects could persist in soil after the decomposition or harvesting of the plants that conained them (National Research Council, 2000). Increased mortality of nontarget monarc butterflies was also witnessed when they were fed milkweed covered in the pollen of GE plants containing the Bt insect-resisance gene (Losey et al., 1999). These results demonstrate that nontarget organisms may be affected by the cultivation of GE plants. Such nontarget effects could extend throughout the food web and so multiply into large-scale effects on many organisms. The potential for these effects is frightening, but the scientific findings thus far must be carefully considered. All of the above studies have been critiized for simulating non-natural conditions such as proximity of plants or unrepresentative concentrations of GE gene products in their work. For these reasons, efforts must be made to improve the research regarding nontarget effects of the genetic engineering of plants.