Sei sulla pagina 1di 10

Pest Management Science Pest Manag Sci 56:299±308 (2000)

Why glyphosate is a global herbicide: strengths,


weaknesses and prospects†
Alan D Baylis*
Zeneca Agrochemicals, Jealott’s Hill Research Station, Bracknell, Berkshire, RG42 6EY, UK

Abstract: After 25 years of sales, glyphosate (N-(phosphonomethyl)glycine) is the world's biggest


selling and fastest growing agrochemical. Glyphosate has many strengths, particularly its systemicity
and ability to control perennial weeds, which have facilitated its role as an essential tool in modern
agriculture. Surveys of recent scienti®c and patent literature point to several key areas of current and
future interest in glyphosate and its use. Biotechnology topics have been the most popular, but there is
still much interest in the molecule per se, particularly in manufacturing processes. The mode of action
and physiology of glyphosate are common subjects. More applied aspects concern the development of
formulations with new adjuvants, or mixtures with other herbicides, to enhance performance. An
increasing number of papers are reporting on its environmental fate and on agronomic techniques
designed to improve the ef®ciency of application and minimise its environmental impact. Lack of weed
resistance is a key feature of glyphosate despite its widespread use. However, differential levels of
tolerance between species have led to changes in weed succession. There is a need to develop integrated
weed management systems in which glyphosate still plays a key role. Glyphosate is an exceedingly
effective herbicide and presents a considerable challenge to the development of new herbicides for
sustainable cropping systems and situations in which cost-effective and environmentally considerate
weed control is required.
# 2000 Society of Chemical Industry

Keywords: glyphosate; weed control; formulations; literature; patents, review

1 A GLOBAL HERBICIDE weeds, and its particular mode of action has meant
Glyphosate (N-(phosphonomethyl)glycine) is a global that any evolved resistance is exceedingly rare, unlike
herbicide because of its versatility in controlling that now facing many other widely used classes of
economically a very broad spectrum of weeds under herbicide. It is of low toxicity and poses no risk of
varied agricultural, industrial, amenity and domestic leaching to ground-water. The more usually noted
situations. weaknesses are: the need for higher rates to control the
Since the introduction of the ®rst glyphosate generally more tolerant broad-leaved weeds; antagon-
herbicide Roundup2 by Monsanto in the early ism by hard water and by many mixture partners; slow
1970s, this chemical, in its several salts and many speed of action and poor rainfastness compared to
formulations, has transformed global agriculture. paraquat. Glyphosate's lack of soil activity is a strength
Together with other non-selective broad-spectrum considering crop safety, but a weakness concerning its
herbicides, particularly paraquat, it has freed millions inability to control weeds emerging after application.
of people from the drudgery of manual weed control. It Overall it allows simple, cheap, ¯exible and effective
has vastly diminished the constraints on productivity weed control while possessing excellent environmental
imposed by perennial weeds. By any measure it has properties. These characteristics and bene®ts are
become the world's leading agrochemical, and is still covered in earlier comprehensive reviews.2,3
the fastest growing,1 enabling the development of The topics covered in this review re¯ect those of
novel and sustainable agricultural systems and facil- current interest and are illustrated by reference to the
itating the production of cheap and wholesome food. more recently published literature; some new data are
Glyphosate has many strengths, and, as shown by its also included. The strengths and weakness of gly-
outstanding record of success, few serious weaknesses. phosate will be considered, not only from the
It controls a very broad spectrum of weeds. Its perspective of the molecule as a herbicide, but also
systemicity means excellent control of perennial from that of the global farming industry as a whole, ie

* Correspondence to: Alan D Baylis, Zeneca Agrochemicals, Jealott’s Hill Research Station, Bracknell, Berks RG42 6EY, UK

Based on a Paper presented at the meeting ‘Twenty-five years of increasing glyphosate use: the opportunities ahead’, organised by
J Caseley and LG Copping on behalf of the Crop Protection Group of the SCI and held on 23 February 1999 at the Royal Aeronautical
Society, London
(Received 30 April 1999; revised version received 15 October 1999; accepted 29 November 1999)

# 2000 Society of Chemical Industry. Pest Manag Sci 1526±498X/2000/$17.50 299


AD Baylis

its role as an essential tool in modern agriculture. In use of `precision agriculture' to improve the ef®ciency
this paper, all application rates are stated on a of glyphosate treatment.
glyphosate acid equivalent basis. The last two categories were concerned with effects
on plants: responses of weeds and weed populations,
and use of glyphosate on crops as a growth regulator.
2 CURRENT RESEARCH INTERESTS Variations in susceptibility between species and
2.1 Scientific literature biotypes to glyphosate and changes in the importance
A survey was conducted on glyphosate literature of particular species in weed ¯ora following repeated
published in scienti®c journals between mid-1995 use of glyphosate have been investigated. The crop
and mid-1998 and abstracted in Chemical Abstracts, itself can be a target for glyphosate, with uses as a
Pestdoc from Derwent, or CABI (Centre for Agriculture desiccant prior to harvest, as a defoliation aid in
and Biosciences International) databases. (Table 1). cotton, as a means of enhancing the sugar content of
There were 630 papers in which some aspect of cane and as a growth retardant all being reported.
glyphosate was a main topic. The following analysis
gives a contemporary view on the scope of research 2.2 Patent literature
into glyphosate and its use. A recent search of Chemical Abstracts revealed 696
Biotechnology topics covering glyphosate-tolerant patent publications pertaining to glyphosate. Twenty-
(GT) crops and the use of glyphosate in GT crops, eight years after the publication of the ®rst patents
were the most popular. There were 37 publications on claiming the herbicidal properties of glyphosate, there
aspects of biotechnology per se, and 70 on weed control is clearly sustained, indeed increasing, commercial
in GT soybeans (Glycine max (L) Merr). Other crops interest in the utility of the molecule. One hundred
included cotton (Gossypium hirsutum L), oilseed rape patent applications were published over the three years
(Brassica napus L), maize (Zea mays L) and sugar beet prior to October 1998 (Table 2). The majority of
(Beta vulgaris L). patents from this period cover claims to inventions in
Papers concerned with conventional weed control the ®eld of glyphosate chemistry. These include
(ie non-transgenic crops) covered land preparation mainly manufacturing topics, but also esters and other
(`burndown') before planting arable crops, control of derivatives of glyphosate acid. A similarly large
perennial weeds in crops prior to harvest (`pre- number cover the genetic transformation of plants to
harvest'), control of annual and perennial weeds in confer glyphosate tolerance. These patents cover not
temperate and tropical orchards, plantations, and non- only the biotechnology involved, but more often the
agricultural situations. A large number of publications elite lines of crop plants created, and even mixtures of
were concerned with glyphosate per se: its mode of glyphosate with other herbicides speci®cally for weed
action, absorption, translocation and metabolism in control in transgenic crops.
plants (especially with respect to environmental Formulation patents include claims to novel types of
in¯uences); its toxicology; analytical techniques; and surfactant and other adjuvants, means of achieving
properties of various analogues. more concentrated formulations, and various presen-
The next largest category comprised the presenta- tations, for instance as a granular product or even as a
tion of glyphosate in conjunction with particular seed dressing to control parasitic weeds. Effects
adjuvants or other herbicides, and the performance claimed usually involve better levels of weed control,
of these formulations and mixtures. This was closely faster action or improved rainfastness, but also include
followed by papers addressing ecological subjects, eg non-agronomic aspects such as lower irritancy.
the impact of glyphosate in the environment, its fate, There have been a number of mixture patents,
and agronomic techniques designed to minimise its usually claiming synergy with other herbicides. Some
environmental impact. Application techniques cov- patents claim novel application methods and one
ered conventional methods, and a particular issue is claims the use of glyphosate as a pharmaceutical
that of spray drift. There were several papers on the product.

Table 1. Analysis of main topics covered in a survey of 630 Table 2. Analysis of main topics covered in a
papers on glyphosate published 1995–1998 survey of 100 patent publications on gly-
phosate published 1995–1998
Main topic No papers % Total
Main topic No patents
Glyphosate-tolerant crops 160 25.4
Conventional weed control 143 22.7 Chemistry 34
Properties of glyphosate 120 19.0 Glyphosate-tolerant crops 29
Formulations 76 12.1 Formulations 13
Ecology 69 11.0 Mixtures 10
Application technology 29 4.6 Formulation effects 9
Responses of weeds 17 2.7 Application techniques 4
Crop uses 16 2.5 Pharmaceutical uses 1

300 Pest Manag Sci 56:299±308 (2000)


Why glyphosate is a global herbicide

These literature analyses illustrate current areas of polyoxyethylene chain surfactants (A11 or A20), but
academic and industrial interest in glyphosate. Overall the salt was up to six times more active than the acid.
they point to several key areas which will be discussed Formulation allows an active ingredient to be
in the remainder of this paper. presented to the target organism in an appropriate
way for the desired effect. By using appropriate cations
and adjuvants formulated at particular concentrations,
it has been possible to tailor glyphosate herbicides to
3 USE IN CROPPING SYSTEMS meet the needs of farmers, applicators and distributors
3.1 Formulation within a wide variety of agricultural and non-agricul-
In recent years a vast array of different formulations of tural systems. Some formulations have further im-
glyphosate have been registered and sold. Being an proved the environmental pro®le of products by using
acid, the active ingredient in glyphosate products can surfactants with lower potential for irritancy and
be formulated as salts or esters. This gives glyphosate reduced toxicity to aquatic organisms.
tremendous ¯exibility. Liquid formulations are more common and suitable
Glyphosate acid is generally less active than the for large-scale operations where ef®cient bulk-hand-
commercially available salts. This is believed to be a ling is required. In such situations, high concentrations
result of poorer absorption into plants.4 At high rates, of glyphosate are preferred. On the other hand, low-
particularly at low spray volumes, glyphosate acid can concentration liquid formulations have been devel-
precipitate out of solution. In the presence of cations oped which are appropriate for the needs of small-
the solubility and physical form of glyphosate change. holders who may wish to purchase small amounts of
The cations used in glyphosate herbicides greatly glyphosate in a conveniently sized pack. A low
increase solubility and make glyphosate more avail- glyphosate concentration also facilitates the inclusion
able. Absorption into leaves from dried spray deposits of high concentrations of adjuvants which may
is much higher and activity on plants is usually improve the degree and robustness of weed control.
substantially greater. Granular formulations have been introduced in some
Results of a glasshouse experiment to compare the markets. These enable higher concentrations of gly-
activity of glyphosate acid to the trimethylsulfonium phosate and offer some advantages in operator
salt (glyphosate-trimesium) are shown in Table 3. handling and easier pack disposal.
Weeds (grasses tillering, broad-leaves with four to six The choice of salt is an important consideration
true leaves) were sprayed with a range of rates in when attempting to develop a particular type of
200 litre haÿ1 water (replicated three times). Applica- formulation. Although dimethylamine and mono-
tions were made with several primary C13/15 alcohol sodium salts were mentioned in the ®rst publication
surfactants (Synperonic series, ICI) varying in poly- on the herbicidal properties of glyphosate,6 for many
oxyethylene chain length5 and AL2042, an alkylpo- years only the isopropylamine salt was sold. Currently,
lyglucoside used in the Touchdown2 formulation of in addition to the isopropylamine salt, trimesium,
glyphosate-trimesium (Zeneca Agrochemicals). A ammonium and sodium salts are the basis of many
visual assessment of herbicidal activity (0±100% different formulations. These salts have been selected
control) was made 25 days after application and the for various reasons, eg compatibility with particular
estimated rates for 90% control (ED90) were calcu- adjuvants, herbicidal ef®cacy, cost and ease of
lated from the rate response. Both acid and salt were manufacture. Suitability for formulating as a liquid
generally more effective with AL2042 or the longer or granular product is a good example of the

Table 3. Comparative activities of glyphosate acid and glyphosate-trimesium (G-TMS) applied with various non-ionic surfactants on annual grasses (G) and
broad-leaved weeds (B)

ED90(g haÿ1) a
Elymus repens Sorghum halepense Chenopodium album Abutilon theophrasti
Surfactant Glyphosate (G) (G) (B) (B)
Synperonic A2 Acid 953 f 1316 c 5197 c 3088 c
G-TMS 537 de 889 b 1091 b 1691 a
Synperonic A7 Acid 632 e 1334 c 3632 c 2869 c
G-TMS 500 d 753 b 1137 b 1961 b
Synperonic A11 Acid 476 cd 897 b 4000 c 2868 c
G-TMS 324 a 443 a 653 a 1535 ab
Synperonic A20 Acid 402 bc 734 b 3797 c 2385 c
G-TMS 366 ab 449 a 609 a 1553 a
AL2042 Acid 362 ab 761 b 3017 c 2706 c
G-TMS 383 b 445 a 510 a 1411 a
a
Means within each column followed by the same letters are not signi®cantly different at P = 0.05.

Pest Manag Sci 56:299±308 (2000) 301


AD Baylis

made. Equilibrium moisture contents decreased at


lower relative humidity, but some salts retained more
water (Fig 1). The least hygroscopic monosodium salt
has been used to make granular products. The ability
of spray deposits of glyphosate to retain and absorb
water after droplets dry down on leaf surfaces may be
important to uptake and consequent biological activ-
ity.4,7
Hygroscopicity and physical form of dried deposits
of glyphosate salts can be demonstrated by observing
how readily they re-hydrate. Figure 2 illustrates this by
showing the path of a mounted needle drawn through
deposits formed from drops of glyphosate salt solu-
tions, applied to glass slides, desiccated and then
returned to ambient conditions (21 °C, 40% relative
Figure 1. Equilibrium moisture contents of glyphosate mono-salts at 90% humidity). The highly water-soluble trimesium salt
(white bars) and 40% (black bars) relative humidity. (solubility 1050 g litreÿ1) will spontaneously re-hy-
drate, but less-soluble salts, eg calcium which has a
solubility of 0.3 g litreÿ1, form a glass-like deposit. In
importance of the glyphosate salt. Salts of glyphosate fact, sub-optimal performance of glyphosate may be
differ in hygroscopicity. Salts with high aqueous associated with the presence of calcium and other
solubility are generally too hygroscopic to allow the polyvalent ions on leaf surfaces.8 Very poor absorption
manufacture of granular formulations. of glyphosate in the presence of calcium has been
A quantitative determination of hygroscopicity was recorded.9
made by a gravimetric measurement of equilibrium
moisture contents of deposits of glyphosate salts. 3.2 Mode of action and symptomology
Three drops of 10% acid equivalent solutions of Glyphosate inhibits 5-enolpyruvylshikimate 3-phos-
several mono-salts of glyphosate were applied to glass phate synthase (EPSPS; E.C. 2.5.1.19) and the
slides and placed in a chamber where they could be literature on its mode of action has been thoroughly
weighed having been allowed to equilibrate at 90% or reviewed.3 EPSPS is located primarily in plastids,
40% relative humidity. Duplicate measurements were although a cytoplasmic form exists also. Inhibition of

Figure 2. Deposits of glyphosate-trimesium and glyphosate-calcium. A mounted needle was drawn through the deposits from bottom to top in each case.

302 Pest Manag Sci 56:299±308 (2000)


Why glyphosate is a global herbicide

EPSPS results in an accumulation of shikimate-3- beans and oilseed rape, especially in situations of late
phosphate, ultimately leading to a block in the harvest or lodging. The symptoms of glyphosate action
production of aromatic amino acids with resultant in plants are not necessarily always herbicidal. Gly-
effects on protein synthesis. phosate and its analogue glyphosine have been used to
Recent work has surmised that glyphosate is not a enhance sucrose concentrations in sugar cane at
transition state inhibitor competing with phosphoe- harvest for many years.17 More recently, it was
nolpyruvate (PEP), as previously believed, but that its reported that in Australia, 135 g haÿ1 glyphosate
binding behaviour with EPSPS is better described as helped to maintain productive pastures of bent grass
an adventitious allosteric interaction, ie a signi®cant (Agrostis castellana Boiss & Reut) by supressing seed-
part of the glyphosate molecule binds outside the head development and reducing apical dominance
EPSPS active site, and that a conformational change such that new tillers dominated the sward.18 This
upon glyphosate binding makes the active site un- increased digestible dry matter and protein content
available to PEP.10 compared with the unsprayed control which was
Although EPSPS is the only known enzyme target of dominated by seedheads. Applications of glyphosate
glyphosate,11 it affects many physicochemical and before spraying cotton defoliants inhibited the re-
physiological processes.12 Among these are reductions growth which is a common problem of cotton in wet
in photosynthesis and degradation of chlorophyll; seasons when harvesting is delayed more than two or
inhibition of auxin transport and enhancement of three weeks after defoliation.19
auxin oxidation. These aspects of the mode of action
of glyphosate are either a direct consequence of the 3.3 Weed control
blockage of the shikimate pathway (through which In any weed control situation the necessary rate is
some 30% of assimilated carbon is estimated to pass)13 determined by the most tolerant species. A range of
or a result of feedback mechanisms. Metal chelation glyphosate rates are recommended for weed control
may also be signi®cant in restricting the availability of from a few hundred g haÿ1 to several kg haÿ1. Inherent
polyvalent cations, eg as enzyme co-factors or as key tolerance to glyphosate in plants can be considered
structural or functional elements, eg Mg2‡ in chloro- between and within species.
phyll.12
Glyphosate is regarded as a slow-acting herbicide, 3.3.1 Between species: tolerance
and this has been seen as a weakness. However, Higher rates are usually required for perennial weed
physiological effects in treated plants can be detected control. These species are usually large plants with a
very much earlier than the ®rst appearance of great potential for regeneration. However, amongst
herbicidal symptoms. Large increases in concentra- annual weeds, relative potency can vary several-fold
tions of shikimic acid occurred within 24 h of gly- (Table 3). Grasses are usually more susceptible to
phosate application to Betula papyrifera Marsh (white glyphosate than broad-leaved weeds. However, there
birch) even from rates which showed no visible can be surprisingly wide differences between closely
symptoms.14 A diagnostic test for determining related species in their degree of tolerance. Clear
whether any crop phytotoxicity related to spray drift differences in the susceptibility of various species of
can be attributed to glyphosate involves an assay for Amaranthus were recorded in a glasshouse experi-
determining the concentration of shikimate in affected ment. Weeds (six to eight true leaves) were sprayed
plants.15 In white birch, sub-lethal doses of glyphosate with a range of rates in 200 litre haÿ1 water (replicated
reduced photosynthesis, chlorophylls a and b declined, three times). A visual assessment of herbicidal activity
and carotenoids increased.14 All these modes of action (0±100% control) was made 22 days after application.
result in the familiar symptoms of glyphosate phyto- Results illustrating differences in control from 225
toxicity: chlorosis, pigmentation, stunting, and reduc- g haÿ1 are shown in Table 4. While A retro¯exus, as
tion in apical dominance.
A further symptom of glyphosate phytotoxicity
which occurs very rapidly in some species is an Table 4. Control of Amaranthus
epinastic response, which is typical of the effects of spp by glyphosate-trimesium
ethylene. Amaranthus retro¯exus L is unusually suscep- (225 g haÿ1) 22 days after appli-
tible for a broad-leaved weed, and the effects of cation.

glyphosate can be observed only a few hours after Control a


spraying. Ethylene production in B papyrifera in- (%)
creased following glyphosate application,14 but little
A retro¯exus 94 a
change in ethylene production was observed in gly- A albus 89 ab
phosate-treated sugar beet (B vulgaris).16 A viridis 77 b
Although primarily a herbicide, of course, the A tamariscinus 58 c
broader utility in the effect of glyphosate on plants is A hybridus 52 c
better illustrated by its uses as a growth regulator. a
Means followed by the same
Glyphosate can be used as a desiccant to facilitate the letters are not signi®cantly differ-
harvesting of high-moisture-content crops, eg soy- ent at P = 0.05.

Pest Manag Sci 56:299±308 (2000) 303


AD Baylis

previously noted, is a broad-leaved weed very (L) Gaertn in Malaysia has been reported in the papers
sensitive to glyphosate, other species are more from this symposium.28
tolerant. This may be signi®cant, in that common The lack of weed resistance to glyphosate has proved
waterhemp (A tamariscinus Nutt) has been noted as a to be unusual considering its widespread and intensive
potentially increasingly serious problem in GT soya.20 use. Several factors may contribute to this:29 the poor
However, this species tends to emerge later in the ®tness of any genotypes evolving with an EPSPS
season, after application of glyphosate, and a residual enzyme resistant to glyphosate, ie a mutated enzyme
herbicide may be needed for good season-long which may not be so affected by glyphosate but which
control. is also less ef®cient at processing PEP; the absence of
The reasons for tolerance of a particular species may ef®cient metabolic degradation or deactivation of
be due to many factors, possibly in combination. glyphosate by plants; the lack of soil activity, with the
Studies on the reasons for tolerance are rare. Calcium result that only the part of the gene pool present in
is a well-known antagonistic ion when present in water plants actually sprayed is selected upon; and the
used for spraying.21 However, calcium and magne- unique mode of action and chemistry of glyphosate
sium, which is similarly antagonistic, are also present among commercial herbicides (ie other herbicides
on the leaf surfaces of many plants.8 More information used for weed control in the same ®eld will not
on tolerance is available from studies on different contribute to the selection pressure).
biotypes of the same species. In future, resistance problems may be posed by GT
crop volunteers and any new biotypes of weeds arising
3.3.2 Within species: tolerance and resistance through out-crossing and transfer of glyphosate
A single population of ®eld bindweed (Convolvulus tolerance genes.29
arvensis L) was found to contain ®ve biotypes which
differed in their susceptibility to glyphosate.22 These 3.4 Integrated weed management
were clearly distinguishable on the basis of their Although resistance to glyphosate is extremely rare,
morphology. The two biotypes with extremes of clearly species and even biotypes differ in their
tolerance were selected for study. The isopropylamine response. This varying tolerance between plants gives
salt of glyphosate at 2.24 kg haÿ1 easily controlled the rise to the very practical problem of weed succession.
more susceptible biotype, but the more tolerant With the advent of GT crops, there has been much
biotype was much less affected, increasingly so when speculation on the extent to which the importance of
older plants were compared. No differences in gross particular species in the weed ¯ora in crops will change
absorption and translocation of glyphosate in the two and the speed at which this will occur. In plantations in
biotypes were recorded.23 In suspension culture, cells SE Asia where glyphosate has been used very
of the more susceptible biotype had higher initial rates intensively, there has been a marked rise in the
of glyphosate uptake.24 The EPSPS from both dominance of annual broad-leaved weeds, eg Asystasia
biotypes was equally sensitive to glyphosate. However, spp, Heydyotis spp, Borreria spp. Several spray rounds
although both biotypes initially had similar levels of each year are required to achieve acceptable weed
enzyme activity at the time of treatment, it then control. Programmes have been introduced in which
increased in tissues of the tolerant biotype. It was integrated spray rounds of paraquat and glyphosate-
postulated that this was a consequence, not a cause of trimesium will give total vegetation control without
better survival by this biotype. There was greater ¯ow further encouraging weed succession.30 Increasing
of carbon through the shikimate pathway in the consideration is also given to the impact of particular
tolerant biotype, a greater pool of phenylalanine and weed species. For example, the presence of some
overall greater growth rate. In conclusion, it seems that grasses is tolerated, as these are less competitive and
multiple mechanisms are involved in the greater their root systems help to prevent soil erosion.
tolerance of this biotype. Indeed, genetic analysis has Euphorbia heterophylla L is a major weed of soybeans
indicated the involvement of multiple genes.25 in Brazil which has become resistant to formerly widely
Many weed species have evolved resistance to used ALS inhibitors, such as imazethapyr. Weed
herbicides of many classes. These situations have pressure from this species can be particularly high
arisen following high selection pressures being exerted because of its emergence over a prolonged period
by repeated use of a herbicide. This has encouraged before and after drilling and re-growth from adventi-
naturally occurring mutations which do not bear a tious buds after application of contact herbicides. A
®tness penalty, or previously small pre-existing popu- weed management system has been devised which
lations of more tolerant biotypes, to become more involves controlling weeds 20 days before drilling with
signi®cant in the weed ¯ora. The International Survey glyphosate-trimesium, followed by paraquat immedi-
of Herbicide Resistant Weeds lists 212 biotypes of ately before drilling, and then application of the
weeds which are resistant to 15 classes of herbicide.26 selective herbicide fomesafen three to ®ve weeks after
Of these, there are 50 biotypes resistant to ALS drilling.31
inhibitors, for example. However, only one, a Lolium An alternative approach integrates weed control
rigidum Gaud biotype from Australia, is resistant to using cultivation and glyphosate.32 Weed germination
glyphosate.27 A further case involving Eleusine indica prior to drilling winter crops was encouraged by tillage

304 Pest Manag Sci 56:299±308 (2000)


Why glyphosate is a global herbicide

and weeds were subsequently controlled by gly- speed of 8 km hÿ1 these correspond to application
phosate-isopropylammonium. This reduced autumn volumes of 100, 200 and 400 litres haÿ1. Droplet sizes
weed populations facilitating effective and economic were measured using a Malvern 2600 Particle Sizer
weed control by selective herbicides. Overall herbicide ®tted with a 1000 mm objective lens. The volume
use over ®ve years was reduced by more than 50%. median droplet diameter (50% of spray volume is
Although many mixtures with selective herbicides found in droplets smaller or larger than this diameter)
were antagonistic to glyphosate, overall they improved decreased and the proportion of spray volume in
post-emergence broad-leaved weed control in GT potentially driftable droplets increased under the
soya.33 Another bene®t of mixtures with soil residual conditions simulating lower volume applications and
activity is to provide protection against weeds germi- higher rates of formulation, which would give a higher
nating after spraying. Mixtures of herbicides with concentration of surfactant.
different modes of action may help forestall the
development of weed resistance. 3.5.2 Precision agriculture
Low rates of glyphosate were preferred to higher Over recent years, the availability of many new and
rates in order to maintain a balanced weed population rapidly advancing technologies ± information technol-
in almond orchards.34 This improved water in®ltra- ogy, satellites, imagery, etc ± have combined with ever
tion, provided a habitat for bene®cial insects and more sophisticated agricultural machinery to offer
helped to preclude invasion by aggressive weeds more farmers the means to tailor and apply crop inputs with
tolerant of glyphosate, such as sedges. far greater accuracy and ¯exibility. `Precision agricul-
ture' is becoming more and more widely adopted,
3.5 Application especially in the corn/soya belt of the US Mid-West.
3.5.1 Spray drift Typical applications at present are variable fertiliser
The use of hydraulic nozzles results in the formation of and residual herbicide rates according to soil maps (of
spray droplets with a wide size distribution. Smaller nutrient and organic matter status) using satellite
droplets are more likely to be retained on leaf surfaces, navigation. Precision application of post-emergence
but those with diameters of < 100 mm are generally herbicides is not widely practised at present, but would
considered to be susceptible to drift. Usually, any be very attractive if its development had economic
glyphosate spray drift reaching crops will be at bene®ts. At present, glyphosate is often used inef®-
relatively low rates. This, and the slow development ciently because weeds occur in patches, the suscept-
of symptoms, may mean that phytotoxicity is not ibility of individual weed species in a mixed ¯ora can
necessarily obvious. However, care has to be taken as vary several-fold, and environmental conditions at the
deleterious effects can occur. Herbicides were applied time of spraying can affect the degree of weed control
to vines in the autumn at low rates simulating drift. In obtained from the rate applied.
the following spring glyphosate, especially when mixed Spraying patches demands that weeds can be
with 2,4-D, affected growth rate of the vines.35 located in some way.41 While in some situations it
Glyphosate caused fruit drop of citrus, especially if seems possible to map weed patches by infestations in
the fruit itself was contacted by spray.36 Fully shielded the previous season, it would obviously be preferable
booms caused less fruit drop than open-boomed to detect weeds shortly before spraying. The presence
sprayers. or absence of a crop will in¯uence how easily this can
The advent of GT soybeans might potentially be achieved. Sprayers with weed-sensing systems have
increase the risk of glyphosate drift causing phytotoxi- been available in N America for a number of years.42
city to young maize.37 Maize injury from glyphosate The presence of weeds may be detected by determin-
spray drift was greater when GT soya was sprayed at ing the differential re¯ectance of red and near-infra-
lower water carrier volumes.38 red wavelengths between green tissue and soil.
Some drift-management adjuvants are available. Forward-mounted sensors signalled the operation of
Reductions in the number of drift-prone droplets each nozzle independently as weeds were detected in
resulted in less drift in a wind-tunnel experiment.39 the spray path. Using such a sensor system allowed the
Surfactants themselves vary in their effects on droplet rate of application of a mixture of glyphosate and
size distribution. At the usual use concentrations, a dicamba to be reduced by up to 60% over three years,
nonylphenyl surfactant with eight ethylene oxide units and gave comparable results to conventional spraying
produced a lower proportion of driftable droplets than on 80% of occasions.43 Best results were obtained
a tallow amine surfactant with 15 ethylene oxide when weed occurrence was patchy, but it was observed
units.40 However, mechanical factors such as nozzle that small weeds, particularly grasses, at low popula-
type and water volume produce the greatest variation tion densities were dif®cult to detect successfully.44
in droplet sizes. Data on spray characteristics of a When spraying weeds in stubble, detection was more
formulation of glyphosate-trimesium are shown in dif®cult in taller stubble. Detection was sub-optimal at
Table 5. The formulation was sprayed at several low sun angles. Besides inaccurate detection, any
concentrations from a pressurised container using a requirements to drive spraying equipment more slowly
single nozzle. Different nozzles and pressures were may limit the advantages of this sort of system.
used to achieve three different ¯ow rates. At a forward The ability to distinguish between species differing

Pest Manag Sci 56:299±308 (2000) 305


AD Baylis

Touchdown
0% 1.25% 2.50% 5.00%
Flow rate
Nozzle (litre minÿ1) VMD a
DD a
VMD DD VMD DD VMD DD
b
Fine 0.60 146 31 132 36 116 41 112 43
Mediumc 1.31 284 9 247 13 222 17 223 19
Coarsed 2.61 423 5 356 6 312 19 264 14
a
VMD = volume median diameter (mm), DD = %spray volume in droplets < 100 mm diameter.
b
Table 5. Effect of spray concentration of product Lurmark 015-F110 with spray at 3 bar pressure.
c
and nozzle aperture on spray characteristicsa of Lurmark 04-F110 with spray at 2 bar pressure.
d
Touchdown2 Lurmark 08-F110 with spray at 2 bar pressure.

in susceptibility would be a further improvement glyphosate has on the environment. The mode of
reducing overall application rates. Various imaging action of glyphosate suggests that toxicity would be
techniques involving plant morphology, leaf colour con®ned to life forms possessing the shikimate path-
and shape are being researched.45,46 `Smart' spraying way, ie plants and micro-organisms. Extensive re-
systems for glyphosate would obviously have to offer search has demonstrated that glyphosate has, indeed,
economic advantages, but there may be environmental very low toxicity to mammals, birds and ®sh; it does
pressures to adopt precision application. More accu- not bioaccumulate; glyphosate and its metabolites are
rate application could also bene®t farmers directly, of not harmful to soil fauna and micro¯ora or honey-
course, by reducing phytotoxicity from spray drift. bees.3
In the future, decision support software could ®ne- Glyphosate is rapidly adsorbed in soil, making it
tune glyphosate rates according to prevailing and essentially immobile. Soil binding is reversible and on
forecast environmental conditions. Environmental desorption glyphosate is degraded by various bacteria
factors are very signi®cant to glyphosate activity.47 to aminomethylphosphonic acid (AMPA) or sarco-
These can have effects both on the plant, conditioning sine, and ultimately to inorganic phosphate, ammonia
its susceptibility, and on the herbicide directly. For and carbon dioxide.3
example, although 820 g haÿ1 glyphosate-isopropyla- Environmental issues surrounding the introduction
mine gave good control of Chenopodium album L and of crops genetically modi®ed (GM) to be tolerant to
Abutilon theophrasti Medik in GT soya in 1994, control glyphosate or other herbicides are of much current
in 1995 was poor.33 In 1995, weeds were at the same interest and the impact of these has already been
growth stage, but markedly smaller. 1995 had been reviewed.51 Although glyphosate itself is particularly
drier and relative humidity was lower at spraying, benign, it indirectly in¯uences the environment by the
although air temperatures were similar. Glyphosate is consequences of its use in agriculture. For example, by
less effective in controlling drought-stressed weeds.48 chemically controlling weeds it enables `conservation
It was also less effective at lower temperatures on tillage' systems to be adopted by farmers direct drilling
plants grown where soil water holding was maintained into uncultivated ground, saving fuel, preventing soil
at ®eld capacity. Glyphosate absorption and transloca- erosion and improving soil structure.52 In GM tolerant
tion is often impaired in droughted plants and those crops, controlling weeds post-emergence with gly-
grown at low temperatures.49 Under drought stress, phosate not only avoids the prophylactic use of pre-
however, reduced glyphosate ef®cacy was associated emergence soil residual herbicides, but allows the
with higher temperatures.48 Increased evapo-tran- development of a weed ¯ora to support birds and
spiration may have exacerbated the effects of soil bene®cial insects. Unsprayed cereal headlands in the
moisture de®cit. Glyphosate was more effective at UK had a greater number of arthropod species
higher relative humidities, and this may be because of important in the diet of young birds of many species,53
greater absorption.50 Leaves of C arvensis had less and a similar situation might be expected in GM
cuticular material when grown at higher relative tolerant sugar beet early in the season before control-
humidity, possibly facilitating better absorption.23 ling weeds with glyphosate. Weeds do not always need
to be removed completely and, if suppressed, the cover
remaining will preclude invasion by more noxious
4 PROSPECTS species. Such systems involving glyphosate pro-
Over the past 25 years many new herbicides have been grammes have also provided a habitat for bene®cial
developed and sold. Many have been far more potent, insects.34
most have been faster acting or had other advantages, This paper has covered the key aspects of gly-
but none have had such a potential to contribute to phosate's role in modern weed control. What are the
global weed control in such a thorough, versatile and prospects of superior herbicides being discovered and
sustainable way as glyphosate. commercialised?
As the world's most widely used and fastest growing Clearly, it is possible to note individual weak points
pesticide there is understandable interest in the impact in the technical pro®le of glyphosate, eg it is less

306 Pest Manag Sci 56:299±308 (2000)


Why glyphosate is a global herbicide

effective on broad-leaved weeds and lacks soil residual 14 Stasiak MA, Hofstra G and Fletcher RA, Physiological changes
activity. Yet, there are always counterpoints, eg the induced in birch seedlings by sub-lethal applications of
glyphosate. Can J Forestry Res 22:812±817 (1992).
extreme rarity of weed resistance and absence of carry-
15 Singh BK and Shaner DL, Rapid determination of glyphosate
over on to crops. And there are always solutions if injury to plants and identi®cation of glyphosate-resistant
needed, eg mix with other herbicides. plants. Weed Technol 12:527±530 (1998).
Glyphosate has not only been a means of weed 16 Cole DJ, Caseley JC and Dodge AD, In¯uence of glyphosate on
control, but it has shaped new agricultural systems selected plant processes. Weed Res 23:173±183 (1983).
17 Boehm RA, Increasing pro®tability with ripeners and specialist
around it, most notably glyphosate-tolerant crops.
adjuvants. Proc West Indian Sugar Technologists 26th Conference,
This now places a new hurdle in the path of developing pp 278±279 (1997).
new and better herbicides. In addition to being 18 Hill RD, Missen DJ and Taylor RJ, Use of glyphosate to prevent
superior commercial weedkillers, they must be su- development of reproductive tillers and extend vegetative
perbly selective to many crops. If not inherently growth of bent grass (Agrostis castellana). Aust J Exper Agric
selective, then not only must they be capable of being 36:661±664 (1996).
19 Martin-Duval T, Vargas R, Wright S and Roberts B, Pre-
used selectively by biotechnology, but then must also conditioning cotton with glyphosate for defoliation. Proc
rely on varieties expressing the resistance trait being Beltwide Cotton Conf 2:1479±1480 (1997).
successfully commercialised. Other developing tech- 20 Owen MDK, North American developments in herbicide-
nologies in precision application may mean that the tolerant crops. Proc Brighton Crop Prot Conf ± Weeds, BCPC,
overall ef®ciency and economy with which glyphosate Farnham, Surrey, UK. pp 955±963 (1997).
21 Nalewaja JD and Matysiak R, Salt antagonism of glyphosate.
can be used will present considerable challenges to
Weed Sci 39:622±628 (1991).
even the most potent of potential new contenders. 22 DeGennaro FP and Weller SC, Differential susceptibility of ®eld
bindweed (Convolvulus arvensis) biotypes to glyphosate. Weed
Sci 32:472±476 (1984).
ACKNOWLEDGEMENTS 23 Westwood JH, Yerkes CN, DeGennaro FP and Weller SC,
I thank Cliff Hart, Dr Anne Waller, John Tobbut and Absorption and translocation of glyphosate in tolerant and
susceptible biotypes of ®eld bindweed (Convolvulus arvensis).
the Zeneca Agrochemicals Non-Selective Herbicides Weed Sci 45:659±663 (1997).
team for data and Dr Ceris Jones and Sarah Harvey 24 Westwood JH and Weller SC, Cellular mechanisms in¯uence
Speck for assisting with the compilation of literature. differential glyphosate sensitivity in ®eld bindweed (Convolvu-
lus arvensis) biotypes. Weed Sci 45:2±11 (1997).
25 Duncan CN and Weller SC, Heritability of glyphosate suscept-
ibility among biotypes of ®eld bindweed. J Heredity 78:257±
REFERENCES
260 (1987).
1 Woodburn A, Glyphosate: Production, pricing and use world-
26 Heap I, International Survey of Herbicide Resistant Weeds [On-line]
wide. Pest Manag Sci 56:309±312 (2000).
Internet. Available: www.weedscience.com. [1998 December
2 Grossbard E and Atkinson D (eds), The Herbicide Glyphosate,
3].
Butterworths, London. 490 pp (1985).
27 Powles SB, Lorraine-Colwill DF, Dellow JJ and Preston C,
3 Franz JE, Mao MK and Sikorski JA, Glyphosate: A unique global
Evolved resistance to glyphosate in rigid ryegrass (Lolium
herbicide. American Chemical Society Monograph 189, Wa-
rigidum) in Australia. Weed Sci 46:604±607 (1998).
shington DC.653 pp (1997).
4 MacIsaac SA, Paul RN and Devine MD, A scanning electron 28 Lee LJ and Ngim J, A ®rst report of glyphosate-resistant
microscope study of glyphosate deposits in relation to foliar goosegrass (Eleusine indica) in Malaysia. Pest Manag Sci
uptake. Pestic Sci 31:53±64 (1991). 56:336±339 (2000).
5 Hollis GL (ed), Surfactants Europa, 3rd edn, The Royal Society of 29 Bradshaw LD, Padgette SR, Kimball SL and Wells BH,
Chemistry, Cambridge. 16 pp (1995). Perspectives on glyphosate resistance. Weed Technol 11:189±
6 Baird DD, Upchurch RP, Homesley WB and Franz JE, 198 (1997).
Introduction of a new broad-spectrum post-emergence herbi- 30 Hee LC, Kim LJ and Jantan B, Comparative studies of a
cide class with utility for herbaceous perennial weed control. paraquat mixture and glyphosate and/or its mixtures on weed
Proc North Central Weed Control Conf 26: 64±68 (1971). succession in plantation crops. The Planter, Kuala Lumpur
7 Nalewaja JD, Matysiak R and Freeman TP, Spray deposit 69:525±535 (1993).
residuals of glyphosate in various carriers. Weed Sci 40:576± 31 Vidal RA, Fleck NG, Oliveira NA, Strello R, Guimaraes FB and
589 (1992). Silva NG, Increasing the number of mechanisms of action of
8 Hall GJ, Hart CA and Jones CA, Plants as sources of cations herbicides for management of weed resistance. Proc Brighton
antagonistic to glyphosate activity. Pest Manag Sci 56:351±358 Crop Prot Conf ± Weeds, BCPC, Farnham, Surrey, UK. pp 363±
(2000). 368 (1997).
9 Nalewaja JD, Devilliers B and Matysiak R, Surfactant and salt 32 Hutcheon JA, Stride CD and Wright KJ, Manipulation of weed
affect glyphosate retention and absorption. Weed Res 36:241± seedbanks in reduced tillage systems for sustainable weed
247 (1996). control. Asp App Biol 51:249±254 (1998).
10 Sikorski JA and Gruys KJ, Understanding glyphosate's molecu- 33 Lich JM, Renner KA and Penner D, Interaction of glyphosate
lar mode of action with EPSP synthase: evidence favoring an with post-emergence soybean (Glycine max) herbicides. Weed
allosteric inhibitor model. Acc Chem Res 30:2±8 (1997). Sci 45:12±21 (1997).
11 Siehl DL, Inhibitors of EPSP synthase, glutamine synthase, and 34 Neu JL and Sieckert EE, Weed suppression with low rates of
histidine synthesis. Rev Toxicol 1:37±67 (1997). Roundup for orchard ¯oor management. Proc California Weed
12 Cole DJ, Mode of action of glyphosate ± a literature analysis, in Sci Soc 47:202 (1997).
The Herbicide Glyphosate, ed by Grossbard E and Atkinson D, 35 Bhatti MA, Al-Khatib K and Parker R, Wine grape (Vitis vinifera)
Butterworths, London. pp 48±74 (1985). response to fall exposure of simulated drift from selected
13 Singh BK, Siehl DL and Connelly JA, Shikimate pathway: why herbicides. Weed Technol 11:532±536 (1997).
does it mean so much to so many? Oxford Surv. Plant Mol Cell 36 Erickson CG, Management of glyphosate-related citrus fruit
Biol 7:143±185 (1991). drop. Proc Florida State Hort Soc., pp 40±42 (1996).

Pest Manag Sci 56:299±308 (2000) 307


AD Baylis

37 Matthews JL and Kapusta G, Corn tolerance to simulated 46 Woebbecke DM, Meyer GE, Bargen K and Mortensen DA,
glyphosate and glufosinate drift. Res Rep North Central Weed Sci Color indices for weed identi®cation under various soil,
Soc 53:505±506 (1996). residue and lighting conditions. Trans Am Soc Agric Eng
38 Franzenburg DD, Owen MDK and Lux JF, Glyphosate drift on 38:259±269 (1995).
corn as affected by carrier volume. Proc North Central Weed Sci 47 Caseley JC and Coupland D, Environmental and plant factors
Soc 50:19±20 (1995). affecting glyphosate uptake, movement and activity, in The
39 Downer R, Wolf TM, Chapple AC, Hall FR and Hazen JL,
Herbicide Glyphosate, ed by Grossbard E and Atkinson D,
Characterizing the impact of drift management adjuvants on
Butterworths, London. pp 92±123 (1985).
the dose transfer process. Proc Fourth Int Symp Adjuvants for
48 Adkins SW, Tanipat S, Swarbrick JT and Boersma M, In¯uence
Agrochemicals. pp 138±143 (1995).
40 Miller PCH, Butler Ellis MC and Tuck C, The in¯uence of of environmental factors on glyphosate ef®cacy when applied
adjuvants on droplet production. Proc Fourth International to Avena fatua or Urochloa panicoides. Weed Res 38:129±138
Symp Adjuvants for Agrochemicals pp 95±102 (1995). (1998).
41 Miller PCH, Stafford JV, Paice MER and Rew LJ, The patch 49 McWhorter CG, Jordan TN and Wills GD, Translocation of 14C
spraying of herbicides in arable crops. Proc Brighton Crop Prot glyphosate in soybeans (Glycine max) and johnson grass
Conf ± Weeds, BCPC, Farnham, Surrey, UK. pp 1077±1086 (Sorghum halepense). Weed Sci 28:113±118 (1980).
(1955). 50 Wills GD, Factors affecting the toxicity and translocation of
42 Felton WL, Commercial progress in spot spraying. Proc Brighton glyphosate in cotton (Gossypium hirsutum). Weed Sci 26:509±
Crop Prot Conf ± Weeds, BCPC, Farnham, Surrey, UK. pp 513 (1978).
1087±1096 (1995). 51 Marshall G, Herbicide-tolerant crops ± real farmer opportunity
43 Blackshaw RE, Molnar LJ, Chevalier DF and Lindwall CW, or potential environmental problem? Pestic Sci 52:394±402
Merits of a weed-sensing sprayer to control weeds in
(1998).
conservation fallow and cropping systems. Weed Sci 46:120±
52 O'Keefe MG and Makepeace RJ, Ef®cacy of glyphosate in arable
126 (1998).
situations, in The Herbicide Glyphosate, ed by Grossbard E and
44 Blackshaw RE, Molnar LJ, Chevalier DF and Lindwall CW,
Factors affecting the operation of the weed-sensing Detect- Atkinson D, Butterworths, London. pp 418±434 (1985).
spray system. Weed Sci 46:127±131 (1998). 53 Moreby SJ, The effects of herbicide use within cereal headlands
45 Bull CR, Zwiggelar R and Stafford JV, Imaging: a technique for on the availability of food for arable birds. Proc Brighton Crop
assessment and control in the ®eld. Asp App Biol 43:197±204 Prot Conf±Weeds, BCPC, Farnham, Surrey, UK. pp 1197±
(1995). 1202 (1997).

308 Pest Manag Sci 56:299±308 (2000)

Potrebbero piacerti anche