MANAGING THE ECOLOGICAL IMPACTS OF GENETICALLY-MODIFIED AGRICULTURAL CROPS: CHANGING PATTERNS OF AGROCHEMICAL USE, NON-TARGET EFFECTS AND GENE FLOW

Dr E.J.P. Marshall

Marshall Agroecology Limited

2 Nut Tree Cottages, Barton, Winscombe, Somerset, BS25 1DU, United Kingdom

E-mail: jon.marshall@agroecol.co.uk

ABSTRACT

GM crops are now grown extensively around the world. Nevertheless, their use is contentious, reflecting concerns over food safety, health and environmental effects. The initial GM traits to be commercialised were for pest resistance and for herbicide tolerance. GMHT crops are widely grown, but attract concerns in regard to the movement of the trait to the same and related species, gene stacking and the impacts of broad spectrum herbicides on weed assemblages. Introductions of GMHT crops impact on herbicide use, potentially increasing the efficiency of weed control, the likelihood of drift to non-target areas and of trophic effects. The use of buffer strips at field edges and areas set aside for farmland biodiversity may be one means of mitigating such effects. Whilst biodiversity concerns are not the same across the globe, where agriculture has evolved its own associated fauna and flora, GMHT crops provide another means of intensification. Gene flow seems to be inevitable for many traits. Therefore the key questions must address the acceptability of the impact of that gene following movement and impacts within the agroecosystem. The challenge is to develop relevant farming systems for different environments, different climates and different societies. Each requires an appreciation of the multi-functionality of the farmed environment. GM technology should be able to contribute to those systems, so long as it is used wisely and is not regarded as the single answer to all agricultural advances.

Keywords: GMHT crops, non-target effects, biodiversity, field margins, weeds

INTRODUCTION

Genetically-modified (GM) crops offer a range of potential benefits to farmers and growers and ultimately to consumers. Genetic modifications to crops can range from incorporating pest and disease resistance genes, herbicide tolerance genes, to genes that engineer the production of specialist oils, pharmaceuticals and other products. In addition, there is considerable interest in developing crops that are capable of withstanding environmental limits to growth. For example, salt tolerance, frost tolerance and nitrogen fixation are all targets of genetic engineering to enhance crop performance. In addition, there is considerable interest in the food industry in more cosmetic changes to crops, e.g. fruit colour, as well as perhaps more altruistic modifications, such as enhancing levels of vitamins (Dunwell, 1998).

Whilst not many of these benefits have been realised, there have been a number of modified crops released for field use in different countries round the globe. GM crops are grown extensively in the USA, Canada, Australia and China. In Europe, there has been a moratorium on the general release of GM crops, reflecting the high levels of environmental and health awareness amongst the general public. The growth of GM crops remains a highly contentious issue in Europe, reflecting worries over commercial globalisation, environmental risks, food safety and perhaps more philosophical issues. The private ownership of genetic information is regarded by some as unethical. The profit motive is seen by such people as incompatible with the exploitation of genetic information. Aspects of the ungoverned exploitation of genetic capital were addressed at the Rio Summit that resulted in the International Convention on Biodiversity in 1992. Specifically, discoveries in individual countries made by foreign companies could not be exploited without the agreement of those countries. Nevertheless, the Convention indirectly promoted the commercial ownership of genetic information. This has resulted in the development of commercial GM crops and the expenditure of large amounts of money on research and development.

The commercial growth of GM crops has increased markedly over recent years. More than 50million ha of GM crops were grown in 2001 (Fig. 1). GM crops are grown in a number of countries (Table 1), with the majority in the USA.

Fig. 1. Area of transgenic crops grown in the world, 1995-2001.

James, C. (2001) Global Review of Commercialised Transgenic Crops: 2001. ISAAA Briefs no. 24: Preview. Ithaca, NY: International Service for the Acquisition of Agri-Biotech Applications. (http://www.isaaa.org/publications/briefs/Brief_24.htm).

Table 1. Area of GM crops by country 1999-2001 (data from James, 2001)

	Total area (million ha)
Country	1999	2000	2001
USA	28.7	30.3	35.7
Argentine	6.7	10.0	11.8
Canada	4.0	3.0	3.2
China	0.3	0.5	1.5
South Africa	0.1	0.2	0.2
Australia	0.1	0.2	0.2
Romania	<0.1	<0.1	<0.1
Mexico	<0.1	<0.1	<0.1
Bulgaria	-	<0.1	<0.1
Spain	<0.1	<0.1	<0.1
Germany	-	<0.1	<0.1
France	<0.1	<0.1	-
Uruguay	-	<0.1	<0.1
Indonesia	-	-	<0.1
Total	39.9	44.2	52.6

Note: in addition to these countries, GM carnations have also been grown in the Netherlands, Japan, Ecuador and Columbia

Whilst the first generation of GM crops have provided production and environmental benefits, they have provided little direct benefit to consumers (Robinson et al., 2000). Second generation GM crops are likely to provide more direct benefits to food processors and consumers. It is generally accepted that the GM technologies have associated risks and some commentators have been critical of the overall approach, e.g. (Steinbrecher, 1996). However, the standard of debate has poor and others have attempted to introduce more reasoned argument (Pretty, 2001; Tester, 2001). Pretty (2001) notes seven types of risk that apply to agricultural systems in regard to GM crops (Table 2).

Table 2. Risks of genetic modifications within agricultural systems (Pretty, 2001).

	Risk
1	horizontal gene flow
2	new forms of resistance and pest problems
3	recombination to produce new pathogens
4	direct and indirect effects of novel toxins
5	loss of biodiversity from changes to farm practices
6	allergenic and immune system reactions
7	antibiotic resistance marker genes

The new European Union (EU) Directive on the release and commercialization of genetically modified (GM) crops (2001-18-EC) includes a requirement for an assessment of indirect effects on the environment of farming practices associated with the introduction of a GM crop. There is also a requirement for post-commercialization monitoring to address impacts of scale and time.

The first GM crops that have been released for commercial use contained a range of genes. These were crops containing Bt genes for pest resistance, notably cotton, glyphosate and glufosinate resistant crops to improve weed control and modifications in tomatoes for delayed ripening and frost resistance. Genetically-modified herbicide-tolerant (GMHT) soybeans, cereals and oilseed rape (canola) were the first major crops to be modified and used on a large scale. As such crops have been grown commercially for some time, this paper will focus on their impacts, so far as is known, but will also examine the many attempts to model the impacts and spread of transgenes, notably in Europe. GMHT soybeans have proved to be extremely popular in the USA, providing good weed control and facilitating crop management (Reddy, 2001). Nevertheless, continuous GMHT cropping may bring problems with it, in terms of weed community shifts and increasing resistance. Conventional weed management, using tillage, crop rotations and herbicides, have affected weed communities. Recently, it has been suggested that weeds within crops are important to other trophic levels within agricultural ecosystems (Marshall et al., 2003). The effects of direct weed removal and indirect impacts of herbicide drift may adversely affect plants and their associated insect, mammal and bird assemblages.

This paper will concentrate on GMHT crops, reflecting the importance of these crops in commercial agriculture, but will briefly consider other traits. The objectives of this paper are 1) to examine the impacts of the introduction of GM crops, 2) to examine the likely patterns of non-target effects associated with GMHT crops, 3) to review the available information on gene flow and transgene movement, in order to 4) propose the best means of managing GMHT crops and 5) to comment on the likely impacts and management of other GM traits that are being introduced into crops.

THE IMPACTS OF INTRODUCED GM CROPS

The likely adverse impacts of GM crops have been reviewed by others, who note that gene flow may be the main risk, associated with both within and between species movement (Kwon & Kim, 2001). Nevertheless, where crops have been introduced, there are have been undoubted benefits to crop management, though there are also a few problems that have been identified (Lutman & Berry, 2003). In general, there is little published scientific data on the field experience of growing GM crops, but a selection of reports are reviewed below.

The introduction of GMHT canola in the USA and Canada was welcomed by farmers, as it facilitated crop management, simplifying weed control operations. One result was an increase in the use of the appropriate herbicides. For example, Agrow No. 307 (p.14) 26 June 1998 reported a 72% increase in the use of glyphosate in the USA, coincident with the introduction of Roundup Ready soybeans. In glufosinate-resistant maize, studies indicate that the effects of weed competition early in the life of the crop are such that pre-emergent herbicide application as part of a spray programme still gives the best yield and economic return (Bradley et al., 2000). Thus, in some modified crops reduced herbicide programmes may not necessarily result from their introduction. Nevertheless, the economic indications are that for some crops in Europe, the introduction of GM technology would enhance profitability, for example sugar beet (May, 2003). A recent review of beet production in Germany noted that the continuing production increases have relied on crop breeding, with inputs, particularly fertilisers, declining. There is particular potential for GM crops to contribute to the industry (Marlander et al., 2003).

GM pest resistant cotton was released for commercial production in 1996 and has been shown to have important economic and environmental advantages (Perlak et al., 2001). Cultivars incorporating Bt genes are now grown on one third of the USA’s cotton area, demonstrating a safe and successful approach.

More than half of the world acreage of GM crops is sown with modified soybeans (Reddy, 2001). This has been facilitated by farmers using a single more effective herbicide that is cheaper than the range of alternatives for unmodified cultivars. However, the use of a single herbicide creates a high selection pressure that may encourage the evolution of resistance and may modify weed assemblages (Reddy, 2001).

Buckelew et al. (2000) have shown that herbicide-resistant soybean crops tend to have lower insect population densities than similar conventional cultivars. The effect is mediated through the impact of weed management, rather than direct effects of herbicide. Cleaner crops support fewer insects. Whilst the indirect effects of weed removal may impact on insects within crops, a study on two GMHT and two conventional oilseed rape cultivars showed no differences in insect pollinator numbers or behaviour associated with crop flowers (Pierre et al., 2003). Nevertheless, the authors suggested that a case-by-case approach to testing such effects might be required.

There has been one development with the commercial growing of oilseed rape that has provided reports of some negative effects of the GMHT approach. In Canada, oilseed rape volunteers have been found to contain three different GM herbicide tolerance genes (Orson, 2001). This example of gene stacking demonstrates that gene flow occurs. It is also of concern that potential weeds are developing resistance to a range of herbicides, adding to the popular myth of the creation of “super weeds”.

Concerns in regard to food safety have been a major factor affecting the acceptance of GM crops and products. These concerns are not static and change over time. For example, there may have been a change in the public attitude against GM food in Japan (Nishiura et al., 2002). Some of these concerns are unlikely to be based on scientific fact. Nevertheless, the possibility of introducing allergens into products is a legitimate concern that should be built into testing regulations (Kuiper et al., 2001). It is also generally accepted that food products should be labelled accurately in regard to GM content. Nevertheless, there is debate on tolerance levels of GM material and methods of assessing that material. Some food sectors, notably the organic (biological) food producers, are demanding zero levels of GM in their products. This may prove difficult to achieve and may require agreed minimum levels (Dale, 2002). In practice, field kits to determine levels of herbicide tolerance in soybean samples have been shown to be unreliable at low levels of contamination (Fagan et al., 2001).

Despite the enormous acreage of GM crops grown around the globe, it is an important concern to environmentalists that there has been little ecological investigation of the impacts of such crops. The work that has been published indicates that there are few if any direct impacts, but indirect impacts may be important (see on). Nevertheless, it is clear that following the introduction of GMHT crops, patterns of herbicide use change. With many pest resistance GM crops, the use of pesticides is significantly reduced, with potential benefits to non-target fauna. The ecological impacts of four GMHT crops have been the focus of large-scale field experimentation in the UK (Firbank et al., 2003). The results of this field study, which has employed split fields of GM and non-GM cultivars, will be released in late summer 2003.

INDIRECT AND NON-TARGET EFFECTS ASSOCIATED WITH GMHT CROPS

The introduction of GM technology is unlikely to have direct adverse effects on products or the environment, though this remains a possibility via the production of toxins, e.g. Bt proteins, allergens. There is a small possibility that genetic material from GM plants may become incorporated into soil bacteria and move between species. However, indirect effects, mediated by changes in crop management and gene flow, may be much more important. The objective of GM pest and disease resistance is the prevention of crop damage. Where this may be achieved by direct toxicity to pests, diseases or disease vectors, there is some risk of the proteins coded by the gene, affecting non-target species. This may be overcome by specificity of action or by limiting exposure. In the case of GMHT crops, the risks are associated with changes in crop management, the crops themselves becoming weeds and the movement of the gene to target weeds.

First, it may be useful to review the approaches to crop management, farm support and the integration of biodiversity concerns into agriculture in different countries. In the heavily populated north western Europe, 50 % of the land surface is in agriculture. In the UK over 95% of the land is managed, mostly as farmland. Under such conditions, land has many functions; as well as providing food, it is used for recreation, nature conservation, biodiversity, water catchment and has historical value and heritage. In many countries, biodiversity and nature conservation are catered for in wilderness areas and are of no concern in agricultural production areas. Under such conditions, the environmental concerns expressed over GM technology seem of little consequence. However, in Europe and increasingly elsewhere, there are attempts to conserve biodiversity in agroecosystems and there are support mechanisms for farmers to achieve this. Financial payments can be made to farmers under agri-environments schemes (Hart & Wilson, 1998; Kleijn & Sutherland, In press; Tahvanainen et al., 2002). This in part is based on the exploitation of functions provided by biodiversity, such as nutrient cycling and natural pest control. In addition, there is a realisation that agroecosystems have evolved over the millennia alongside man and have developed their own assemblages of fauna and flora, some of which are threatened by the rapid advances in agricultural technologies of the past 50 years. In the UK and elsewhere in Europe, it has become apparent that some species commonly found in farmland have shown highly significant declines in population size and geographical range over the past 30 years. For example, many common birds have declined in the UK (Fuller et al., 1995), some by over 90% (Table 3).

Table 3. Percentage changes in farmland bird populations between 1974 and 1999 recorded in the British Trust for Ornithology Common Bird Census plots in the UK. Taken from: http://www.bto.org/birdtrends/appendix71b.htm#cbcfarm25

Species	Plots (n)	Change (%)	Lower limit	Upper limit	Comment
Linnet	73	-46	-58	-30
Lapwing	38	-45	-69	-31	Unrepresentative
Moorhen	55	-43	-52	-29
Treecreeper	29	-42	-68	-11
Yellowhammer	73	-42	-53	-32
Dunnock	93	-40	-51	-27
Goldcrest	27	-37	-54	-6
Blackbird	96	-34	-41	-27
Cuckoo	50	-26	-45	-2
Tree Sparrow	34	-93	-97	-86
Corn Bunting	17	-90	-95	-80	Small sample
Grey Partridge	40	-83	-88	-77
Turtle Dove	25	-81	-90	-67
Spotted Flycatcher	31	-75	-86	-60
Bullfinch	47	-71	-79	-61
Snipe	7	-70	-96	-53	Small sample
Song Thrush	82	-66	-73	-57
Redshank	9	-60	-82	-19	Small sample
Reed Bunting	49	-58	-71	-44
Starling	65	-55	-68	-38
Skylark	83	-54	-60	-44
Mistle Thrush	59	-51	-61	-43

Agricultural intensification, in both arable and grassland areas of the UK, has been shown to play an important part in those declines (Baillie et al., 2001; Chamberlain & Fuller, 2000; Chamberlain et al., 2000; Siriwardena et al., 2000, 2001). Potential mechanisms are reviewed by (Fuller, 2000), and include pesticides, though only for one species,