Biosicherheit & Nachhaltigkeit




Zentrum BATS


Zentrum für Biosicherheit und Nachhaltigkeit

Technikfolgen Risikoanalyse
Home Home  |   Language: English English  |  

Dieser Artikel ist leider nicht in deutscher Sprache verfügbar.

Herbicide Tolerant Crops - Impacts on agriculture and environment

Table of Contents:

Download this article as pdf here: Herbicide Tolerant Crops ( pdf 500 Kb) E. Rasche, G. Donn


Crop growing is not feasible without measures to control weeds. Weeds are plants which compete with the planted crop for space, light water and nutrients. They could furthermore cause difficulties during harvest or in the harvested product.
Weeds can be undesirable wild plants as well as plants of the crop(s) previously planted on the same field. They diminish yield of crops and quality of products. Therefore weeds must be controlled to exploit the yield potential of crops.
However, all systems of weed control be they mechanical, thermal, biological or chemical, have some side effects. For instance mechanical tillage can accelerate soil erosion under certain conditions and flaming can additionally kill the accompaiying fauna while chemical herbicides can cause concerns in regard to possible residues in and on the harvested crop and to environmental impacts. These drawbacks can not be completely eliminated but they can be minimized.
The first requirement for widespread use of a herbicide after the emergence of a crop is its tolerance by the crop. Most crop plants display various degrees of resistance to many herbicides. Such natural resistance is generally based on morphological or physiological differences between weeds and crop plants. In some crops selectivity is also achieved by appropriate application techniques (soil application before sowing or emergence of the crop, spray shields to mask the crop).
These systems are sometimes poorly effective, and damage may result from even slight overdosing. The application rates per hectare as well as the stage of development of the crop plants, play an important role in avoiding this damage when selective herbicides are applied.
Modern herbicide research aims at the development of products which meet the farmers requirements and have the lowest possible side effects. As a result, more environmentally acceptable herbicides are available. However, some of these herbidices can not be used on major field crops.

Methods to generate herbicide tolerant plants

In recent years several approaches have proved practicable of circumventing this problem of selectivity in crops. One way is through biotechnology. It is now possible for instance to transfer herbicide inactivating traits into important crop plants. The responsible genes have been shown to be stably integrated and inherited.
Currently seed companies are developing crop plants which are tolerant to different herbicides. One of those herbicides which meet requirements of farmers, regulators and the informed public is Glufosinate Ammonium. It is taken as case for further considerations, because herbicide tolerance achieved by genetic modification is tight to a specific herbicide or a group of herbicides with the same mode of action and not existing per se.

Glufosinate Ammonium - a modern herbicide

Glufosinate Ammonium is the ammonium salt of the amino acid Phosphinothricin which has been derived from the natural compound L-phosphinothricyl -L-alanine-L-alanine. This tripetide was obtained from Streptomyces viridochromogenes by Bayer et al (1972) in Germany and from Streptomyces hygroscopicus by Kondo et al (1973) in Japan and got the common name Bialaphos. The amino acid Phosphinothricin was found to be the biologically active moiety of the tripetide. It proved to have strong herbicidal efficacy on mono- and dicotyledonons plants and was patented by Hoechst AG (Rupp et al, 1977).

Mode of herbicidal efficacy

The broad herbicidal activity of Glufosinate Ammonium is the result of a specific inhibition of glutamine synthetase. In plants this is an essential enzyme for the assimilation of primary ammonia as well as for the reassimilation of ammonia released by metabolic processes. The inhibition of glutamine synthetase causes increasing concentrations of ammonia in plant cells. Finally they reach a level which is phytotoxic and destroys the plant.

Mechanism of Tolerance to Glufosinate Ammonium

The Streptomyces species producing the tripeptide Bialaphos containing the herbicidal active amino acid Phosphinothricin possess an enzyme - and thus a gene - which acetylates the amino group of Phosphinothricin. It protects the producer strain against intoxication by the own metabolite.
De Block et al (1987) proved that a Bialaphos resistance gene (named BAR gene) isolated from Streptomyces hygroscopicus was expressed in plants as well and protected them from herbicidal effects of Glufosinate Ammonium. Wohlleben et al (1988) isolated and characterized a Phosphinothricin resistance gene from Streptomyces viridochromogenes. This gene codes for an enzyme named Phosphinothricin-Acetyl-Transferase (PAT). Correspondingly the gene is called PAT gene. Although nucleotide sequences of BAR and PAT genes are slightly different from each other they code for similar enzymes which inactivate Glufosinate Ammonium by specific acetylation of its amino group. N-Acetyl-Glufosinate is formed which is no longer inhibiting glutamine synthetase and does not have any herbicidal activity.

Metabolism and Residues

Glufosinate tolerant plants

The metabolism of Glufosinate-Ammonium was tested in the following genetically modified crops: maize, soybean, rape, tomato. Due to the presence of the PAT gene in these plants the herbicidal active isomer (L-Glufosinate) is rapidly acetylated within as fast as it penetrates into the plant. The high initial residue level (residues in the transgenic plants normally consist of N-acetyl-L-Glufosinate, D-Glufosinate and with lesser amounts 3-methylphosphinico-propionic acid, Hoe061517) decreases when the plants are growing. From residue trials it is documented that at the silage or harvest stage no residues remain (below limit of quantification) in maize and oilseed rape.


In soil the active substance Glufosinate-ammonium is rapidly degraded with half-life (DT-50) of 3-20 days and DT-90 values of 10-30 days. The new metabolite from the transgenic plants N-acetyl-L-Glufosinate is also rapidly degraded. The deacetylation (first degradation step) takes place within one day after which the degradation pathway and velocity is the same as for the parent substance Glufosinate-Ammonium. Due to this rapid degradation, the risk of leaching and groundwater contamination is minimized. In lysimeter studies neither Glufosinate nor its metabolites were detectable in the leachate.


Mammals excrete both Glufosinate-Ammonium and N-acetyl-Glufosinate very rapidly. More than 90% of the administered dose are excreted within the first 48-h intervall after admininstration. (90% of that amount via faeces and normally less than 10% with the urine after oral dose). N-acetyl-L-Glufosinate is stable in the animal: it passes the stomach of rats without deacetylation, in the urine only the N-acetyl compound is excreted. Deacetylation takes place in the faeces presumably by the intestinal bacteria. Depending on the initial dose between 1 and 10% are deactylated. This process is reversible. After dosage of Glufosinate-Ammonium, the acetylated compound was found in the rat faeces.

Product safety and environmental effects

Toxicological properties

Acute, subchronic and chronic toxicity studies showed very low toxicity of Glufosinate-Ammonium. As a manufacturing use product Glufosinate-Ammonium possesses no toxic properties which would render it a dangerous substance as defined in the Regulation on Hazardous Materials.
Tests for sensitizing properties yielded no indications of allergenic effects.
Mutagenicity assays showed Glufosinate-Ammonium to be non - mutagenic.
N-Acetyl-L-Glufosinate, which is formed in Glufosinate-Ammonium tolerant plants, is per se inactive and did not show toxic properties either.

Assessment of hazard to typical non-target organisms

Assessments of possible ecological risks for the use of Glufosinate-Ammonium were conducted by Dorn et al (1992). They have shown that as far as can be known at present no hazard is to be expected neither to aquatic organisms as algae, waterfleas and fishes nor to terrestrial organisms as soil microorganisms, earthworms, honeybees, beneficial arthropods, birds and mammals.


Glufosinate Ammonium is registered as a non selective herbicide in all major countries of the world. It is known for its favorable toxicological and ecotoxicological profile. In spring 1995 Canadian authorities granted the first registration for use as a selective herbicide in Glufosinate tolerant oilseed rape.

Safety Assessments of PAT Gene and Protein

Specificity of the PAT Protein

Incubation of purified enzyme (PAT Protein) with 14C Glutamate did not result in acetylation of Glutamate. A 1000 fold higher concentration of Glutamate (or other proteingenic amino acids) could not outcompete 14C labelled Glufosinate as a substrate. The extremely high substrate specifity of the PAT protein indicates that no other proteinogenic amino acid can be acetylated.


Degradation studies of the PAT gene in digestive fluids from pork, chicken and cow showed that the DNA is completely degraded within 1 hour (at 37°C and pH 1,5). Degradation studies of the PAT Protein in digestive fluids from pork, chicken, cow and in simulated human gastric fluid resulted in an immediate breakdown of the protein and its enzymatic activity within seconds.
These results demonstrate an equivalent behaviour of PAT gene and protein with other DNA and proteins of our diet.


A thorough comparison of the PAT protein with other known protein sequences was conducted. No homology to known allergenic or toxic protein could be detected.
All results suggest that the PAT gene and PAT protein do not pose any hazard to consumers.

Safety Assessment of transgene dispersal

With the advent of transgenic crops, the danger of gene introgression from cultivated plant species into the natural flora is widely discussed. Herbicide resistance genes are in this respect an ideal approach for ecological monitoring. If the postulated gene flux from cultivated crops towards the wild flora would occur it would be easily detectable. Herbicide tolerant crops therefore can help to answer the questions:

  • Does gene flux from a given crop towards wild relatives occur under natural conditions?
  • What is the frequency of such events?
  • Which consequences has the gene transfer for the fitness of wild plants?
  • Which effects on agricultural ecosystems has such a postulated gene flow?

It is evident that no general answer can be drawn on this topic. Therefore a species by species consideration is necessary. It is accepted that a transgene in maize cannot be transmitted via cross pollination to other plant species in most parts of the world with the exception of Southern Mexico and Central America where wild relatives occur.

Soybean is also a safe crop in this respect. Only in Northern China wild relatives occur and the plant is a notorius self pollinator.

Sugarbeets can cross freely with wild beets (Beta Vulgaris ssp maritima). This wild plant grows along the coasts of Western Europe and the Mediterranean bassin. In these areas outcrossing is posssible. Using transgenic beets as females or by avoiding seed production areas where wild beets occur, the outcrossing into wild beet populations can be avoided. In farmers fields sugar beets do normally not flower. The few flowering contaminants which arose either from weed beets or from crosses with wild beets in some seed production areas (Broomberg et al 1995) should be erased before seed setting anyway.

Rapeseed is considered as a crop which interbreeds with related wild species from the Brassicaceae family. Therefore in this crop the probability for outcrossing is higher than in all other agricultural crops.
Rapeseed (Brassica napus) arose from a cross between B campestris (B. rapa) and its B. oleracea. It still can be backcrossed with its ancestors. Especially crosses between B. napus and B. campestris are well documented, (Jürgensen et al 1994) whereas outcrossings from rapeseed into B. oleracea under field conditions do not occur at a detectable frequency (K. Hild, personal communication). In fields in which B. campestris grows as a weed besides B. napus and the rapeseed variety confers a herbicide resistance gene, the weedy relative will be eliminated by spraying the complementary herbicide before an outcrossing can occur. B. campestris does not occur in natural habitats. Its ecological preferences are similar to rapeseed, which only can establish on sites which are free of competing perennial plants (Crawly et al 1992).

If the rapeseed field will not be treated with the complementary herbicide and outcrossing into B. campestris can occur, the consequence is similar to the situation of a non contaminated rapeseed field. Good agricultural practice will suppress and eliminate the emerging seedlings in the next crop.

Crosses of rapeseed and the two most common related weeds Sinapis arvensis and Raphanus raphanistrum do not occur under agricultural conditions even if both weeds grow in close vicinity of B. napus. If pollen from the wild species are brought on a stigma o B. napus together with B. napus pollen, the B. napus pollen grows through the pistill faster and will fertilize the egg cell. As a consequence only non hybrid seed will develop The same is true in reciprocal crosses where the weeds are used as females. Also there the pollen of the same species fertilizes the ovaries (Kerlan et al 1992).
Therefore hybrids between Raphanus raphanistrum or Sinapis arvensis and rapeseed are not detected even in rapeseed fields heavily infested with both weeds. Recently it was shown that in fields were male sterile rapeseed genotypes were grown in close vicinity of Raphanus raphanistrum in the absence of male fertile B. napus plants, a few seeds developed on the sterile rapeseed plants.
The seedlings were analysed. Besides dihaploid rapeseed plants, intergeneric hybrids were found, amphitriploids as well as trihaploids and plants with irregular chromosome numbers (Baranger, Ph D. Thesis 1995).

The obtained interspecific F1-hybrids showed besides morphological abnormalities a reduced fertility. It is extremely unlikely that under agricultural conditions these plants can compete with well adopted fully fertile species.

These mentioned artificial conditions under which the Raphanus-Brassica hybrids were created do not correspond to conditions in agriculture, because a pollinator with B napus pollen was completely excluded. In presence of B. napus pollen this would have outcompeted the pollen of the related species as described above.

In order to set the remote probability of intergeneric crosses in the right relation to agricultural reality it is helpful to remember that in a rapeseed field at harvest 3-5% of the seeds are falling on the ground. This corresponds to 100-150kg seeds/ha which is equivalent to 2.000-3.000 seeds/m2 or 20-30 mio seeds/ha.

Ever since rapeseed is grown as a crop, farmers have to handle this problem. The rapeseed seedlings which emerge from the lost seeds behave like a weed in the following crop. Appropriate cultivation practice as well as herbicide rotation solved this problem in the past as they will do in the future.

Weed Resistance to Glufosinate Ammonium

Due to the mode of action of Glufosinate Ammonium it is very unlikely that weeds become resistant. The reason is that this would require a mutation of the target enzyme glutamine synthetase. However, mutated glutamine synthetase which lost its binding affinity for Glufosinate Ammonium simultaneously lost its binding affinity for glutamate, a structural analogue of Glufosinate Ammonium. A mutated enzyme therefore could not catalyze the amidation of glutamate to glutamine, the essential detoxification step for ammonia. Such a mutation would be lethal. Therefore it is extremely unlikely that weed will develop a spontaneous resistance towards Glufosinate Ammonium.

This hypothesis is well supported:

  1. Glufosinate Ammonium has been used on some areas since over 17 years several time a season. No observations of resistant weeds have been made.
  2. Extensive invitro plant selection programmes for maize and alfalfa failed to yield Glufosinate Ammonium tolerant plants. On the other hand for other herbicides it is easy to select tolerant mutants.

Impacts on weed management - herbicide use pattern

Since the first field tests with Glufosinate Ammonium tolerant tobacco plants in 1989 in France there was proof for the applicability of the new approach for selective weed control with Glufosinate Ammonium. In the following years field trial work was extended tremendously in Europe and particularly in North America. Glufosinate Ammonium tolerant Oilseed Rape, Maize, Soybeans and more recently Sugar Beets have been grown in over 1.100 field trials. Glufosinate Ammonium has been sprayed post emergence with single and sequential applications of 150 to 600 g ai/ha. The timings of treatments were choosen primarily according to the following growth stages of weeds: early post (2-4 leaves) mid post (3-5 leaves), late post (5-8 leaves). Generally broad spectrum weed control of all major weeds was achieved with one to two treatments of up to 600 g ai/ha depending on one hand on weed pressure and on the other hand on duration of weed emergence and row closure ot the crop. Crop safety for Glufosinate Ammonium was excellent. Yield assessments revealed high yields confirming the concept of superior weed control with post emergent application of Glufosinate Ammonium.

The impact on weed management can be summarized as follows:

  1. The farmer acquires an additional option for controlling weeds after they have emerged. Nonetheless, all previous methods or products continue to be available to him.
    Volunteers from a previous Glufosinate Ammonium tolerant crop can therefore be controlled by the same means as before and cause no new problems.
  2. Tolerance by the crop ensures maximum protection of yield. The dependence of treatment on the growth stage of the crop is reduced, thus making application easier to time. Technical management requirements can be taken into consideration to a greater degree. It is easier to keep weeds below damage thresholds.
  3. The new system can also make some treatments completely unnecessary (if weeds are below damage thresholds) or reduce sequential sprayings. Decisions based solely on the level of weed infestation and its development help to reduce the amount of herbicides applied.
  4. Crop rotation benefits from reduced herbicide residues in the soil.
  5. The opportunities for different growing methods, e.g. erosion control using undersowing or similar techniques, can be improved in combination with the new system.


There is a proof now that genetically modified herbicide tolerant crops become reality in agriculture. For the first time Glufosinate Ammonium tolerant oilseed rape varities have been registered and launched in Canada 1995.

This new approach will substantially contribute to further improved cultivation of major crops. It allows the safe use of herbicides with well accepted favourable profiles. This enables farmers to control weeds as competitors of crops with as little side effects as possible.

© Copyright Zentrum BATS: Kontakt Legal Advisor: Advokatur Prudentia-Law Veröffentlichungsdatum: 1995-10-17

Suchen Sie bei
Passende Antikörper aus
über Produkten.