ArticlePDF Available

The sustainability of cotton : consequences for man and environment

Authors:
  • Aad Termorshuizen Consultancy

Abstract and Figures

Dit onderzoek van de wetenschapswinkel voor Goede Waar & Co had als doel om de literatuur samen te vatten waarin milieu-effecten vergeleken worden van gangbare, IPM (Integrated Pest Management = geïntegreerde landbouw) en biologische katoenproductie. Tweede doel was de selectie van indicatoren om het milieu-effect van gangbare, IPM en biologische katoenproductie in kaart te brengen. Tot slot wilden de opdrachtgevers een discussie over duurzame aspecten van katoenproductie en ‑handel startten.
Content may be subject to copyright.
The sustainability of cotton
Consequences for man and environment
Karst Kooistra and Aad Termorshuizen
Biological Farming Systems, Wageningen University
Chapter 5 written by Rhiannon Pyburn
Communication and Innovation Studies, Wageningen University
Science Shop Wageningen UR
April 2006 Report 223
- ii -Science Shop Wageningen UR Report 223
The sustainability of cotton
Colofon/Colophon
De duurzaamheid van katoen
Gevolgen voor mens en milieu
The sustainability of cotton
Consequences for man and environment
Karst Kooistra and Aad Termorshuizen
Biological Farming Systems, Wageningen University
Chapter 5 by Rhiannon Pyburn
Communication and Innovation Studies, Wageningen University
Wetenschapswinkel Wageningen UR, rapportnummer 223
Science Shop Wageningen University & Research Centre, report number 223
April 2006
ISBN: 90-6754-90-8585-000-2
Keywords: Cotton, organic farming, IPM, conventional agriculture
Begrippen: Katoen, biologische landbouw, IPM, gangbare landbouw
Omslag/cover: Andrew Zeegers, Domino Design, Groningen
Lay-out: Hildebrand DTP, Wageningen
Druk/print: Drukkerij Grafisch Service Centrum Van Gils, Wageningen
Foto’s/photo’s: Karst Kooistra (2004)
www.wur.nl/wewi
Het binnenwerk van dit rapport is gedrukt op 100 % kringlooppapier.
Inside printed on 100% recycled paper.
Auteursrecht:
De tekst van deze publicatie mag worden overgenomen onder voorwaarde dat de bron als volgt wordt vermeld: Kooistra,
K.J., Pyburn, R., Termorshuizen, A.J. 2006. The sustainability of cotton. Consequences for man and environment,
Wetenschapswinkel Wageningen Universiteit en Reseachcentrum. Rapport No. 223. ISBN: 90-6754-90-8585-000-2.
Copyright:
Text from this publication may be used provided that the source is cited as follows: Kooistra, K.J., Pyburn, R., Termorshuizen,
A.J. 2006. The sustainability of cotton. Consequences for man and environment, Science Shop Wageningen University &
Research Centre. Report 223. ISBN: 90-6754-90-8585-000-2.
- iii -Science Shop Wageningen UR Report 223
The sustainability of cotton
The sustainability of cotton
Consequences for man and environment
Rapportnummer 223/Report no. 223
Karst Kooistra and Aad Termorshuizen, Rhiannon Pyburn (chapter 5)
Wageningen, April 2006
Goede Waar & Co
Mira Vendrig, director
Diederick Sprangers, researcher
Amaliastraat 5
P.O. Box 61236
1005 HE Amsterdam, The Netherlands
phone: +31 (0)20 686 33 38
e-mail: miravendrig@goedewaar.nl
diedericksprangers@goedewaar.nl
www.goedewaar.nl
This project was commissioned by Goede Waar
& Co, association for consumers. Goede Waar
& Co represents the interests of socially and
environmentally conscious consumers. Based
on extensive research into sustainability, the
organisation informs you on how clothing and
food products affect you as a consumer and how
they effect the world around you.
Biological Farming Systems Research
and Education Group
Address: Marijkeweg 22, building 527
6709 PG Wageningen, The Netherlands
phone: +31 (0)317 47 82 01
e-mail: aad.termorshuizen@wur.nl
karst.kooistra@wur.nl
rhiannon.pyburn@wur.nl
www.bfs.wur.nl/uk
Activities at the Biological Farming Systems
Research and Education Group are geared
towards the development and transfer of new
scientific knowledge to analyze, design, and
evaluate sustainable agricultural farming systems
where natural ecological processes are utilized to
the maximum possible. Research is focused on
mixed animal and plant production systems
and specialised plant production systems (if
possible in collaboration with animal production
systems).
Innovation Centre Organic Agriculture,
Wageningen University and Research
centre.
P.O. Box 107
6708 PB Wageningen, The Netherlands
phone: +31 (0)317 47 58 19 or
+31 (0)317 48 58 31
e-mail: info@biologischelandbouw.net
www.biologischelandbouw.net
The Innovation Centre Organic Agriculture
(IBL) co-ordinates, stimulates and initiates the
development and integration of knowledge on
organic agriculture and food.
For the IBL management of this project was done
by Jet Proost Communication Consultancy
Address: Laarweg 18
6721 DE Bennekom, The Netherlands
phone: +31 (0)318 43 14 56
e-mail: jet.proost@planet.nl
Wetenschapswinkel Wageningen UR
Science Shop Wageningen UR
P.O. Box 9101
6700 HB Wageningen, The Netherlands
phone: +31 (0)317 48 39 08
e-mail: wetenschapswinkel@wur.nl
www.wur.nl/wewi
www.wetenschapswinkels.nl
Dutch low-budget citizen groups are welcome at
the Science Shop of Wageningen UR. We help
realising research projects in the fields of nutrition
and health; sustainable agricultural systems;
environmental quality and processes of social
change.
- iv -Science Shop Wageningen UR Report 223
The sustainability of cotton
- v -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table of Content
Preface ..........................................................................................................................vii
Acknowledgements .........................................................................................................vii
Abstract ..........................................................................................................................ix
Korte samenvatting ...........................................................................................................x
Summary .........................................................................................................................xi
Samenvatting ................................................................................................................. xiv
1. Introduction ............................................................................................................. 1
2. Cotton .................................................................................................................... 3
2.1 The cotton plant ....................................................................................................... 3
2.2 Global cotton production overview ............................................................................. 3
2.3 Cotton cultivation ..................................................................................................... 5
2.3.1 Irrigation ........................................................................................................6
2.3.2 Fertilisers ....................................................................................................... 9
2.3.3 Pests and diseases ...................................................................................... 10
2.3.4 Pesticides ....................................................................................................10
2.3.5 Defoliation .................................................................................................... 16
2.3.6 Yields ........................................................................................................... 16
2.3.7 Genetically Modified cotton ............................................................................ 16
2.3.8 Clearing of forest/natural areas ......................................................................17
3. Types of cotton production .....................................................................................19
3.1 Conventional systems............................................................................................. 19
3.2 Organic systems .................................................................................................... 28
3.3 IPM systems .......................................................................................................... 31
3.4 Other promising techniques .................................................................................... 32
3.5 Summary ..............................................................................................................32
4. Environmental impact indicators .............................................................................. 33
4.1 Water use .............................................................................................................. 34
4.2 Human and environmental toxicity ...........................................................................34
4.3 Global warming ...................................................................................................... 34
4.4 Eutrophication .......................................................................................................35
4.5 Acidification ........................................................................................................... 35
4.6 Erosion ................................................................................................................ 35
4.7 Land use ............................................................................................................... 36
4.8 Biodiversity............................................................................................................ 36
4.9 Salinisation ............................................................................................................ 36
4.10 Conclusions ........................................................................................................... 37
5. Social and trade aspects related to sustainable cotton production and processing ...... 39
5.1 Introduction ........................................................................................................... 39
5.2 Social issues addressed in organic cotton certification ..............................................39
5.3 Issues addressed via more comprehensive social certifications ................................. 41
5.4 Potential socially and environmentally sustainable cotton supply chain ........................ 42
6. Discussion ............................................................................................................. 45
6.1 Environmental impact ............................................................................................. 45
6.2 Indicators .............................................................................................................. 45
6.3 Choice of farming system ....................................................................................... 46
6.4 Social and trade issues ..........................................................................................47
6.5 Conclusions ........................................................................................................... 48
6.6 Recommendations .................................................................................................49
References .................................................................................................................... 51
Annex 1: Indicators used in cotton studies........................................................................ 59
- vi -Science Shop Wageningen UR Report 223
The sustainability of cotton
- vii -Science Shop Wageningen UR Report 223
The sustainability of cotton
Preface
It’s hard to find two product groups closer to our skin than food and clothing. This is why Goede
Waar & Co, as an association for socially responsible consumers, has been concentrating
on these themes for the last years. More and more consumers want to know under which
circumstances their food products and garments have been produced. What impact does the
production have on people, animals and the environment?
This demand for information led, amongst others, to the 2004 ‘Cotton Checker’ campaign:
consumers were able to look up their favourite clothing brand on the internet and see how
it performed on a number of social and environmental indicators, concentrating on cotton
production. The consumers could also ask ‘their’ brands for an increased use of organic
cotton.
Certainly this last aspect caused a lot of discussion in newspapers and scientific journals.
Clothing brands were claiming the organic cotton production was too low to satisfy their demand;
scientists were arguing that organic cotton shared various sustainability problems with ‘regular’
cotton. As Goede Waar & Co concentrates on providing dependable information, this called for
more research. For this, we sought co-operation with the Science Shop (Wetenschapswinkel)
and the IBL, Innovation centre for Organic Agriculture, both part of Wageningen University and
Research Centre.
I am pleased to announce that the report you are about to read is the well-defined scientific
base for this research. For example, thanks to this report we now have indicators to compare
the environmental impact of conventional, ‘Integrated Pest Management’ (IPM) and organic
production systems. We have starting points to discuss the social aspects relevant for trade in
sustainable cotton. And we know that even organic production can cause unsustainable use of
water for irrigation.
In other words, we have completed a promising first step. I am confident we will be able to give
the consumer the full ‘sustainability picture’ at the end of the road. Although one thing is already
clear: as far as the sustainability of cotton production goes, there is still a world to be won.
Mira A. Vendrig
Director of Goede Waar & Co
Acknowledgements
In order to produce a well-balanced report external reviewers have commented on a draft version.
The reviewers were: S. Ferrigno (Organic Exchange, USA), R. Holland (WWF - World Wide Fund for
Nature), Prof. Dr. R. Rabbinge (Dean, Wageningen University, the Netherlands), N. Reintjes (PAN-
Germany, Germany), M. Scheffer (Noeton, the Netherlands), and P. Ton (Consultant at AgroEco,
the Netherlands). We are very grateful for their detailed and constructive comments. In addition,
the following persons contributed to this study through valuable discussions and comments:
Dr. W. Blok (Biological Farming Systems, Wageningen University), Prof. Dr. A. van Bruggen
(Biological Farming Systems, Wageningen University), F. Mancini MSc (Biological Farming
Systems, Wageningen University), J. Proost MSc (Independent consultant), and G. Straver MSc
(Science Shop, Wageningen University).
Karst Kooistra, Rhiannon Pyburn and Aad Termorshuizen
- viii -Science Shop Wageningen UR Report 223
The sustainability of cotton
- ix -Science Shop Wageningen UR Report 223
The sustainability of cotton
Abstract
The sustainability of cotton – Consequences for man and environment
The environmental impacts associated with cotton production are increasingly in the spotlight
however, data on the sustainability of cotton production are scanty and not widely available. As
such, the goals of this current project were as follows:
1. To summarise literature that compares conventional, IPM, and organic cotton production
systems in terms of environmental impact;
2. To select indicators, based on literature resources, which can be used to formally compare
the environmental impact of conventional, IPM, and organic cotton production systems and
to carry out comparisons based on these indicators;
3. To suggest starting points for discussion on the social aspects relevant for trade in sustain-
able cotton.
Cotton is produced in more than 100 countries, the most important countries being: China
(24% of global cotton production), the USA (19%), India (16%), Pakistan (10%), Brazil (5%) and
Uzbekistan (4%). Three types of cotton cultivation are distinguished in this study: conventional
systems (representing about 80% of cotton production internationally), Integrated Pest Manage-
ment (IPM) systems (approximately 20%); and organic systems (0.04%). This study provides an
extensive description of the cultivation practices of each of system in the major cotton produc-
ing countries and the related environmental effects.
Despite the general lack of accurate data, it is concluded that water and pesticide use cause the
most significant environmental problems in cotton systems. The use of pesticides in several de-
veloping countries is of particular concern. While organic systems are quite well-defined through
standards and regulations and are third-party verified, the lack of definition and monitoring as
to optimal (i.e. most sustainable) conventional and IPM cotton for both rain-fed and irrigated
systems hampers interpretation as to their sustainability. To address this, certification of con-
ventional and IPM systems and appropriate product labelling is recommended. In addition, water
and land use indicators for all cotton systems need further development and implementation to
allow more comprehensive estimates of the environmental impacts of cotton production.
If cotton production and processing are to be sustainable in a more holistic sense, then social
and trade aspects need to be addressed in addition to environmental indicators. IFOAM Basic
Standards for organic agriculture address baseline social standards which ensures that all
IOAS-accredited certification bodies include some social elements in their own standards.
Coordinated efforts with other more comprehensive social standard-setting bodies are however
necessary in order to ensure an end product that is verified as socially and environmentally
sustainable. For cotton, certification packages that address the many steps of the supply chain
(production, ginning, spinning, dying, weaving/knitting, cut-make-trim) need formulation.
- x -Science Shop Wageningen UR Report 223
The sustainability of cotton
Korte samenvatting
De duurzaamheid van katoen – gevolgen voor mens en milieu
De invloed van de productie van katoen op het milieu krijgt internationaal steeds meer aandacht.
Aangezien er geen gebundeld overzicht bestaat van de milieu-effecten van de productie van
katoen was het doel van dit onderzoek:
1. Samenvatting van de literatuur waarin milieu-effecten vergeleken worden van gangbare, IPM
(Integrated Pest Management = geïntegreerde landbouw) en biologische katoenproductie;
2. Selectie van indicatoren op basis van literatuurgegevens die gebruikt kunnen worden om het
milieu-effect van gangbare, IPM en biologische katoenproductie in kaart te brengen;
3. Start van een discussie over duurzame aspecten van katoenproductie en -handel.
Katoen wordt geproduceerd in meer dan 100 landen, met China (24% van de wereldproductie),
de VS (19%), India (16%), Pakistan (10%), Brazilië (5%) en Oezbekistan (4%) als belangrijkste
producenten. In het algemeen worden drie productiewijzen onderscheiden: gangbare systemen
(80% van de katoenproductie), IPM (20%) en biologische (0,04%). In dit rapport wordt voor de
belangrijkste katoenproducerende landen een uitgebreide beschrijving gegeven van de cultuur-
maatregelen in deze teeltsystemen en hun milieu-effecten.
In verhouding tot de omvang van de katoenteelt is het aanbod aan betrouwbare gegevens over
de invloed van katoenproductie op het milieu beperkt. Desondanks kan geconcludeerd wor-
den dat water- en bestrijdingsmiddelengebruik in de katoenteelt de belangrijkste milieu-effecten
hebben. Vooral het gebruik van bestrijdingsmiddelen in een aantal ontwikkelingslanden is zeer
verontrustend. Terwijl biologische teelt tamelijk goed is gedefinieerd, is het gebrek aan een
algemeen aanvaarde omschrijving van de randvoorwaarden voor de teelt van gangbare en IPM
katoen lastig voor de interpretatie van de milieu-effecten van deze systemen. Dientengevolge
wordt geconstateerd dat er een grote variatie is in milieu-effecten van individuele systemen. Cer-
tificering en labeling van gangbare en IPM katoenteelt wordt aanbevolen. Ook wordt aanbevolen
de duurzaamheid van water- en landgebruik verder te ontwikkelen voor alle teeltsystemen zodat
deze gebruikt kan worden bij milieu-effectrapportages van katoenproductie.
Duurzaam geproduceerde katoen betekent dat naast een duurzame teelt van de plant ook de
keten in zijn geheel duurzaam en sociaal rechtvaardig dient te zijn. De regels van de IFOAM voor
biologische landbouw garanderen minimumeisen. Deze zouden nog verder uitgebreid moeten
worden in samenwerking met certificeringsorganisaties die specifiek gericht zijn op deze
elementen. Voor katoen dient dit gespecificeerd te worden voor alle elementen uit de keten.
- xi -Science Shop Wageningen UR Report 223
The sustainability of cotton
Summary
The sustainability of cotton – Consequences for man and environment
Cotton cultivation represents over 31 million hectares or 2.4% of global arable land, involving
about 20 million farmers who completely depend on cotton production and another 30 million
farmers who include cotton into their rotation scheme. The environmental impact of cotton
production is increasingly in the spotlight with the main problems being salinisation, desertifica-
tion and poisoning of the environment including human health. To date, there is insufficient insight
as to the degree of (un)sustainability of the systems in place. In addition, Integrated Pest Manage-
ment (IPM) and organic systems are usually assumed without critical analysis, to have lower
environmental impacts than conventional production. Therefore, the goals of this project were:
1. To summarise literature that compares conventional, IPM, and organic cotton production
systems in terms of environmental impact;
2. To select indicators, based on literature resources, which can be used to formally compare
the environmental impact of conventional, IPM, and organic cotton production systems and
to carry out comparisons based on these indicators;
3. To suggest starting points for discussion on the social and trade sustainability of cotton.
Cotton is produced in more than 100 countries with the major producers being: China (6.3 Mt of
raw cotton, representing 24% of global production), the USA (5.1 Mt, 19%), India (4.1 Mt, 16%),
Pakistan (2.5 Mt, 10%), Brazil (1.2 Mt, 5%) and Uzbekistan (1.1 Mt, 4%). Although global cotton
production and yield per unit area have increased steadily over the last three decades, the area
used for the cotton for a given country is quite variable, especially for organically grown cotton.
Cotton requires large amounts of water both for cultivation and processing. Irrigation is used in
areas where normal precipitation quantities do not match the requirements for the crop being
cultivated. 53% of cotton fields worldwide are irrigated and because irrigated cotton generally
has higher yields per unit of area, this translates to 73% of all cotton production. The most com-
monly used irrigation system for cotton cultivation is the flood-or-furrow irrigation system, which
although easiest to install, has the lowest water efficiency, 40% at most. More efficient irrigation
systems are more expensive to install. Soil salinisation is a common outcome of poor irrigation.
One-third of irrigated land is already or will soon be affected by salinisation: the numbers are
likely to be higher for cotton as it tends to be cultivated in more sensitive areas. Other negative
side-effects of irrigation are water shortage and pollution downstream. An estimated 5 Mha of
arable land (4% of the world total) is abandoned due to former intensive cotton cultivation with
soil salinisation being the main reason.
15% of cotton yield loss is due to insect damage. Fungal and bacterial plant pathogens, viruses,
and nematodes are of lesser importance. Cotton cultivation has been estimated to consume 11%
of the world’s pesticides while it is grown on only 2.4% of the world’s arable land. Insecticides
used in cotton cultivation represent 25% of the global consumption. In developing countries it is
estimated that approximately 50% of all pesticides are applied in cotton cultivation. In addition,
pesticide use and storage are often badly managed. The techniques used for pesticide applica-
tion strongly affect environmental impact, especially in terms of human exposure. An estimated
13% of the cotton growing areas are aerial sprayed. Cases of contamination of neighbouring
villages due to aerial spraying for cotton production have been reported. Field spraying by hand
accounts for application in 52% of the world’s cotton fields. While cheap and effective, poisoning
of workers who apply the pesticides is common in developing countries. The remaining 35% of
cotton fields are treated by tractor spraying. Recommendations as to pesticide application need
to be implemented.
A wide array of pesticides are being applied on cotton, among them are those listed by the World
Health Organisation (WHO) as highly hazardous, including: monocrotophos, triazofos, parathion,
parathio-methyl, phosphamidon, methamidophos, demeton-S-methyl. In various countries some
of these pesticides are no longer permitted but without adequate verification systems, they are
still in use. Many older, non-patented, toxic, environmentally persistent and inexpensive chemi-
cals are used extensively in developing nations, creating serious acute health problems as well
as local and global environmental contamination. The quantity of these illegal pesticides is very
- xii -Science Shop Wageningen UR Report 223
The sustainability of cotton
difficult to estimate. In more developed countries, pesticides are regulated and hazardous insec-
ticides are not allowed.
In general, three types of cotton cultivation are recognised: conventional (farming systems that are
neither organic nor IPM; approximately 80% of cotton production), Integrated Pest Management
(IPM; 20%) (systems designed to minimise pesticide and often also fertiliser use), and organic
(a certified farming system that amongst other standards, does not permit synthetics in farming
such as synthetic pesticides, fertilisers and genetically modified organisms; 0.04%). In this
report an extensive description of cultivation practices in the major cotton-producing countries
is provided.
If cotton production and processing are to be sustainable in a more holistic sense that goes
beyond environmental indicators, then social and trade aspects need to be addressed. A robust
approach to social sustainability includes the following aspects: freedom of association and right
to collective bargaining; working hours; seasonal workers, contracts; child labour; health and
safety; wages; discrimination; and disciplinary practices. IFOAM Basic Standards for organic
agriculture address baseline social standards which ensures that all IOAS-accredited certification
bodies include some social elements in their own standards. Coordinated efforts with other
more comprehensive social standard-setting bodies (i.e. Social Accountability International,
Sustainable Agriculture Network of the Rainforest Alliance, and Fair Trade Labelling Organisations
International) are however necessary in order to ensure an end product that is verified as socially
and environmentally sustainable. For cotton, certification packages that address the many steps
of the supply chain (production, ginning, spinning, dying, weaving/knitting, cut-make-trim) need
formulation.
The report arrives at the following conclusions:
• Water use (including the potential negative consequences of salinisation and water resource
depletion) and pesticide use represent the greatest environmental problems in cotton
systems. Water use and type of irrigation are not explicitly included in the standards for
organic production at present. In principle, unsustainable water use may occur in all three
farming types, although the majority of organic cotton systems are not irrigated. In developing
countries most organic cotton is rain-fed while the majority of the IPM and conventional cotton
is irrigated. The environmental impact of water use is considerably lower for rain-fed than for
irrigated cotton, however the opportunities for yield gains by optimising water applications are
also consequently lower.
• The widespread use of extremely and highly toxic pesticides in developing countries is a major
concern both for the natural environment and for people who are exposed, especially workers
applying these pesticides. Pesticide application methods strongly affect the potential impact
on field workers and the environment; thus, the way field workers apply pesticides needs to
be better understood.
• Biodiversity is negatively affected through the use of broad-spectrum pesticides, through
conversion of natural habitats and through loss of land to degradation or shifting farming
practices. For cotton production systems biodiversity is addressed through the land use and
toxicity indicators. The effect on biodiversity of cropping vast areas of monoculture of cotton
needs additional attention.
• Limited data on the environmental impact of cotton systems is available. In addition, the great
heterogeneity of farm types hinders data extrapolation.
• Because of the great diversity in farming practises it is difficult to express the environmental
impact of conventional, IPM and organic cotton farming on a systems level. The use of
pesticides is highly problematic in several important cotton-producing countries and here the
distinction between conventional and organic cotton production is evident.
• In developed countries conventional cotton production may to some extent approach IPM
given that the most persistent and hazardous pesticides are not allowed and because inputs
are increasingly optimised for cost reasons. However, the lack of clear-cut definitions for IPM
and conventional systems also contributes to this conclusion.
• Organic farming is a certified type of production, while conventional is not, and IPM is generally
not. With certified production it is easier for retailers and consumers to understand the history
of the produce and thus make inferences about environmental impact. Although the general
aims of organic farming are to promote an overall sustainable approach, the standards need
improvement as they do not sufficiently address land and water use issues. For optimal,
- xiii -Science Shop Wageningen UR Report 223
The sustainability of cotton
highly-technological conventional cotton production systems, a standards and verification
system is necessary because otherwise there is no way to differentiate clearly unsustainable
conventional farming systems from more sustainable conventional farming systems.
• Reflecting on the desirability of organic cotton production, water and land use issues should
be kept in mind. Expanding the organic cotton area in rain-fed regions by converting low-input
conventional farms to organic could be an optimal approach as yields can be as high as those
in IPM and conventional cotton, water use can be generally sustainable, and no pesticides are
used. This applies especially to countries that apply relatively hazardous pesticides. However,
for conversion to organic cotton in irrigated areas it is necessary to address water and land
use issues. Organic cotton production on eroded, pesticide-polluted and/or salinised soils
would not be a viable option. Conversion to organic cotton production on large tracts of land
that are mainly conventional can be a challenge if pest pressure is high.
Recommendations
• The current study could be extended by an analysis of each of the approaches against
their characteristic visions and ultimate goals. Descriptions and definitions of optimal, most
sustainable conventional and IPM cotton systems for both rain-fed and irrigated cotton are
needed. Certification of such systems and labelling of products should be considered.
• Knowledge of what happens in the field is the basis for recommendations. Clearly for such
an important crop as cotton more field work is needed to collect quantitative field data on
environmental impact and yields in various developing and developed countries. This could
lead to an extended Life Cycle Assessment comparing conventional/IPM and organic farms in
several countries using the following indicators: Water use, Human and environmental toxicity,
Global warming, Eutrophication, Acidification, Erosion, Land use, Biodiversity and Salinisation.
Also the environmental impact of cotton production should be compared with the production
of other fibres such as polyester, viscose, hemp, and linen (but see Cherrett et al., 2005 for
first results on a comparison of cotton, polyester, and hemp).
• Further research is needed on how various stakeholders influence cultivation methods.
• Water and land use indicators need further development and implementation in LCA-cotton
studies. The standard for organic production should also address these elements.
• The social, environmental, and trade-related issues that are particularly relevant for cotton
production in different country contexts and at different points along the supply chain merit
further research in order to build a comprehensive picture of sustainable cotton beyond just
environmental indicators.
• Organic certification done by an IOAS-accredited certification body assures that a baseline
level of social justice is maintained at the production level. To address more than these
minimal social concerns, to address trade-related concerns and to address the other parts of
the supply chain, other means are necessary. For example, Fair Trade Labelling Organisations
(for small producer organisations) or Social Accountability International (for larger operations
and other steps along the supply chain) are possible certification partners.
• Certification packages need to be constructed that address environmental, social, and trade-
related concerns throughout the whole supply chain. Costing and traceability for such a system
clearly needs attention.
• Verification and certification bodies in the three sustainability areas (social, environment, trade)
should study how they can co-operate.
- xiv -Science Shop Wageningen UR Report 223
The sustainability of cotton
Samenvatting
De duurzaamheid van katoen – gevolgen voor mens en milieu
Katoen is een belangrijk gewas: het wordt geteeld op 31 miljoen hectare; dit is 2,4% van
het oppervlak dat wereldwijd beschikbaar is voor landbouw. Naar schatting zijn 20 miljoen
boeren volledig afhankelijk van katoen en nog eens ongeveer 30 miljoen hebben katoen in hun
vruchtwisseling opgenomen. De invloed van de productie van katoen op het milieu krijgt steeds
meer aandacht, maar precieze gegevens over de milieubelasting zijn in de literatuur lastig te
vinden. Verder wordt een geringe milieubelasting van geïntegreerde en biologische katoenteelt
tamelijk kritiekloos aangenomen, hoewel ook hier een literatuuroverzicht ontbreekt. Het doel van
dit onderzoek was daarom als volgt:
1. Samenvatting van de literatuur waarin milieu-effecten vergeleken worden van gangbare, IPM
(Integrated Pest Management = geïntegreerde landbouw) en biologische katoenproductie;
2. Selectie van indicatoren op basis van literatuurgegevens die gebruikt kunnen worden om het
milieu-effect van gangbare, IPM en biologische katoenproductie in kaart te brengen;
3. Start van een discussie over duurzame aspecten van katoenproductie en -handel.
Katoen wordt geproduceerd in meer dan 100 landen, met China (6,3 Mt ruwe katoen, 24% van
de wereldproductie), de VS (5,1 Mt, 19%), India (4,1 Mt, 16%), Pakistan (2,5 Mt, 10%), Brazilië
(1,2 Mt, 5%) en Oezbekistan (1,1 Mt, 4%) als belangrijkste producenten. Hoewel wereldwijd
de katoenproductie en de productie per oppervlakte-eenheid de laatste dertig jaar sterk is
toegenomen, is de hoeveelheid geteelde katoen per land aan tamelijk sterke veranderingen
onderhevig, met name voor biologisch geteelde katoen.
Katoenproductie vraagt grote hoeveelheden water, zowel tijdens de teelt als tijdens de verwer-
king. Irrigatie is nodig op 53% van het katoenareaal, waarmee 73% van de wereldkatoenpro-
ductie behaald wordt. Het waterverbruik per eenheid katoen wordt sterk bepaald door het type
irrigatie. Het meest gebruikte type is ‘flood-or-furrow’-irrigatie, waarvan de efficiëntie slechts ca.
40% bedraagt. Andere, meer efficiënte irrigatiesystemen zijn echter voor de meeste landen te
duur. Als gevolg van inefficiënte irrigatie treedt verzilting op. Ongeveer éénderde van alle geïr-
rigeerde landbouwgebieden is aangetast door verzilting. Katoen heeft hierin een relatief groot
aandeel omdat dit vaak geteeld wordt in gebieden die gevoelig zijn voor verzilting. Ook de
relatieve zouttolerantie van katoen draagt hier aan bij. Een neveneffect van irrigatie is dat het
benedenstroomse areaal aangetast kan worden door verzilting, waardoor watergebrek kan ont-
staan. Inmiddels is ca. 100 miljoen hectare land (8% van het oppervlak dat beschikbaar is voor
landbouw) braakliggend ten gevolge van verzilting door niet-duurzame landbouwmethoden, en
hiervan is een groot gedeelte veroorzaakt door de teelt van katoen.
Insectenaantastingen veroorzaken in de katoenteelt opbrengstverliezen van gemiddeld 15%. In
het algemeen zijn schimmel-, bacterie-, virus- en nematodenaantastingen van geringer belang.
Katoenteelt vergt met zo’n 11% van het wereldwijde gebruik relatief veel gewasbeschermingsmid-
delen. Insecticiden, gewasbeschermingsmiddelen tegen plaaginsecten, vergen in de katoenteelt
zo’n 25% van het wereldwijde gebruik. In ontwikkelingslanden worden gewasbeschermingsmid-
delen relatief nog het meest in de katoenteelt ingezet: naar schatting 50%. Bovendien worden
in ontwikkelingslanden gewasbeschermingsmiddelen op gevaarlijke wijze toegepast en worden
middelen gebruikt die niet toegestaan zijn in meer ontwikkelde landen. De wijze van toepas-
sing bepaalt in hoge mate de blootstelling van mensen en het milieu, en daarmee de mate
van ongewenste neveneffecten. Op ongeveer 13% van het wereldwijde katoenareaal worden
gewasbeschermingsmiddelen toegepast met vliegtuigen. In 52% van het katoenareaal worden
gewasbeschermingsmiddelen met handsproeiers toegepast. Dit is goedkoop en effectief, maar
vergiftiging van veldwerkers komt frequent voor in ontwikkelingslanden. Bij de rest van het katoe-
nareaal (35%) worden gewasbeschermingsmiddelen met tractoren met daarop spuitinstallaties
geïnstalleerd aangewend. Aanbevelingen over veilige methoden van toepassing van gewasbe-
schermingsmiddelen dienen snel te worden beschreven en geïmplementeerd. Om een inschat-
ting te kunnen maken over de effecten van gewasbeschermingsmiddelen op het milieu dient
nauwkeurig bekend te zijn hoe deze middelen zijn toegepast.
Vele soorten gewasbeschermingsmiddelen worden in de katoenteelt toegepast, waaronder
een aantal die door de Wereldgezondheidsorganisatie WHO geclassificeerd zijn als hoogriskant
- xv -Science Shop Wageningen UR Report 223
The sustainability of cotton
(‘highly’ of ‘extremely hazardous’): monocrotophos, triazofos, parathion, parathio-methyl, phosp-
hamidon, methamidophos en demeton-S-methyl. In verscheidene landen worden niet-toegestane
gewasbeschermingsmiddelen gebruikt. Het illegale gebruik van gewasbeschermingsmiddelen
is moeilijk in te schatten, maar waarschijnlijk zorgen zij voor grote milieu- en gezondheidspro-
blemen in veel landen. In meer ontwikkelde landen is het gewasbeschermingsmiddelengebruik
gereguleerd en zijn hoogriskante middelen veelal niet toegestaan.
In het algemeen worden drie teeltsystemen onderscheiden: gangbare systemen (systemen die
niet gerangschikt kunnen worden onder IPM of biologisch; 80% van de katoenproductie), geïn-
tegreerde landbouw (‘integrated pest management’, IPM; systemen die tot doel hebben het ge-
wasbeschermingsmiddelengebruik en vaak ook de bemesting te minimaliseren; 20%) en biologi-
sche (een gecertificeerd type van productie waarin synthetische gewasbeschermingsmiddelen en
kunstmest niet zijn toegestaan; 0,04%). Een uitgebreide beschrijving van de cultuurmaatregelen
in deze teeltsystemen en hun milieu-effecten wordt gegeven voor de belangrijkste katoenprodu-
cerende landen.
Sociale en handelsaspecten van katoenproductie zijn van belang om de duurzaamheid van de
gehele keten te beschouwen. Aspecten die hierbij van belang zijn omvatten vrijheid van organi-
satie van arbeiders, recht van onderhandelen, seizoensarbeid, arbeidsduur, kinderarbeid, con-
tractplicht, gezondheid en veiligheid, salarissen en discriminatie. De organisatie die de standaard
voor biologische landbouw bepaalt, IFOAM, heeft ook basisvoorwaarden gesteld aan sociale
aspecten. Om werkelijk een sociaal duurzaam katoenproduct te garanderen is samenwerking
met certificeringsorganisaties nodig die gespecialiseerd zijn op dit gebied, zoals ‘Social Accoun-
tability International’, ‘Sustainable Agriculture Network of the Rainforest Alliance’ en ‘Fair Trade
Labelling Organisations International’. Voor katoen dient dit gespecificeerd te worden voor alle
elementen uit de productieketen.
Het rapport komt tot de volgende conclusies en aanbevelingen:
Conclusies
• Watergebruik, samen met de potentiële gevolgen hiervan (verzilting en uitputting in de geïr-
rigeerde gebieden), en gewasbeschermingsmiddelengebruik zorgen voor de grootste proble-
men in katoenteelt. Duurzaamheid van watergebruik is momenteel niet expliciet beschreven
in de regels voor biologische productie. In principe kan niet-duurzaam watergebruik dus op-
treden in alle teeltsystemen, hoewel op dit moment de grote meerderheid van biologische
katoensystemen gesitueerd is in niet-geïrrigeerde gebieden.
• Het brede gebruik van riskante gewasbeschermingsmiddelen in vooral de ontwikkelingslan-
den is zeer zorgelijk, zowel voor het milieu als voor mensen die aan deze middelen worden
blootgesteld. De wijze van toepassing van gewasbeschermingsmiddelen beïnvloedt in hoge
mate de effecten op mens en milieu. Meer inzicht in de praktijk van wijze van toepassing van
gewasbeschermingsmiddelen is daarom sterk gewenst.
• Biodiversiteit wordt negatief beïnvloed door gewasbeschermingsmiddelen met een breed wer-
kingsspectrum en door toename van het benodigde landareaal voor katoenteelt. Meer aan-
dacht dient te worden besteed aan het effect op biodiversiteit van de teelt van enorm grote,
monotone oppervlaktes katoen in ontwikkelde landen.
• Slechts een beperkte hoeveelheid milieugegevens over het effect van katoenteelt op het mi-
lieu is beschikbaar. De aanzienlijke diversiteit in de wijze van katoenteelt bemoeilijkt de extra-
polatie van gegevens.
• Gezien de grote variatie in de katoenteelt kan op systeemniveau geen algemene uitspraak
gedaan worden over de invloed op het milieu van gangbare, IPM en biologische landbouw. Het
gebruik van gewasbeschermingsmiddelen is met name in ontwikkelingslanden zeer problema-
tisch en daarmee onderscheidt biologische katoenteelt zich duidelijk van de gangbare en IPM
katoenteelt.
• In meer ontwikkelde landen kan het milieu-effect van de gangbare katoenteelt tot op zekere
hoogte die van IPM-katoenteelt benaderen, omdat daar ook in de gangbare teelt de meest
gevaarlijke gewasbeschermingsmiddelen niet zijn toegestaan. Deze conclusie kan overigens
worden beïnvloed doordat gangbare en IPM-katoenteelt lang niet altijd scherp gedefinieerd zijn.
- xvi -Science Shop Wageningen UR Report 223
The sustainability of cotton
• Biologische teelt is een gecertificeerd type van productie. Gangbare teelt is dat niet, en IPM
meestal niet. Voor handelaren en consumenten is met gecertificeerde teelt gemakkelijker in te
zien wat de historie is van het product en dus wat de milieu-effecten zijn geweest. Hoewel het
algemene doel van biologische landbouw een zo duurzaam mogelijk productiewijze is, zijn er
op het vlak van land- en watergebruik zeker verbeteringen mogelijk. Een duidelijk omlijnd en
gecertificeerd productiesysteem is ook nodig voor de gangbare en IPM-katoenteelt aangezien
alleen dan de duurzaamheid van zo’n systeem ingeschat kan worden.
• Als de wenselijkheid van biologische katoen overwogen wordt dan moet water- en landgebruik
hierbij betrokken worden. In gebieden waar irrigatie niet nodig is door voldoende regenval is
omzetting van niet-duurzame, gangbare systemen in duurzame, biologische systemen aan
te bevelen. In deze gebieden kan de opbrengst aan katoen in kg/ha even hoog zijn als die in
IPM en gangbaar geteelde katoen. Het voordeel is groter naarmate er op dit moment meer
gevaarlijke gewasbeschermingsmiddelen worden ingezet. Als echter het doel is om gangbare
katoenteelt in geïrrigeerde gebieden om te zetten in biologische teelt dan dient land- en wa-
tergebruik in de discussie expliciet te worden betrokken. Biologische katoenteelt op geëro-
deerde, verzilte en met gewasbeschermingsmiddelen vervuilde gronden dient niet overwogen
te worden. De introductie van biologische katoenteelt in gebieden waar de gangbare teelt
overheerst is niet aan te bevelen vanwege de daar veelal hoge druk aan plaaginsecten.
Aanbevelingen
• Een analyse is nodig van de karakteristieke uitgangspunten en doelen van ieder van de teelt-
systemen en hoe deze in de praktijk gebracht worden. Er is behoefte aan de beschrijving en
definitie van een optimaal, maximaal duurzaam gangbaar en IPM katoenteeltsysteem voor
zowel niet- als wel-geïrrigeerde gebieden. Certificering van zo’n systeem en de ontwikkeling
van een label voor producten is nodig.
• Kennis over de praktijk van katoenproductie is de basis voor aanbevelingen. In relatie tot de
omvang van de katoenteelt is verrassend weinig bekend over de milieu-effecten van verschil-
lende katoensystemen en daarom is meer veldwerk in zowel meer als minder ontwikkelde
landen dringend noodzakelijk. Dit zou kunnen leiden tot een uitgebreide levenscyclusanalyse
van verschillende teeltmethoden van katoen in diverse landen met gebruik van de milieu-indi-
catoren watergebruik, milieu- en humane toxiciteit, ‘global warming’, eutrofiëring, verzuring,
erosie, landgebruik, biodiversiteit en verzilting. Het effect op het milieu van katoenproductie
dient te worden vergeleken met de productie van andere vezels zoals polyester, viscose, lin-
nen en hennep (een deelstudie hierover werd recentelijk verricht door Cherrett et al., 2005).
• Onderzoek is nodig over hoe de diverse belangenbehartigers in de katoenketen de katoen-
teeltmethoden beïnvloeden.
• Indicatoren voor water- en landgebruik dienen te worden ontwikkeld voor gebruik in leven-
scyclusanalyses. Dit dient ook te worden meegenomen in de standaarden voor biologische
katoenteelt.
• De sociale, milieu- en handelsgerelateerde aspecten in de katoenketen dienen integraal meeg-
enomen te worden in overwegingen over de duurzaamheid van katoenproductie.
• Biologische certificering door een IOAS-geaccrediteerde organisatie verzekert een basaal
niveau van sociale rechtvaardigheid in de katoenteelt. Certificering van katoen met inbegrip
van sociale rechtvaardigheid verderop in de katoenketen dient te worden gedaan door ‘Fair
Trade Labelling Organisation International’ (voor kleine producenten) of ‘Social Accountability
International’ (voor grote producenten).
• Naast de katoenproductie dient de gehele katoenketen te worden gecertificeerd. De kosten
hiervan en de traceerbaarheid van gecertificeerde katoen verdienen met name aandacht.
• Controle- en certificeringsorganisaties zouden moet onderzoeken hoe beter samengewerkt
kan worden op het gebied van duurzaamheid van milieu, sociaal en handel.
- 1 -Science Shop Wageningen UR Report 223
The sustainability of cotton
1 Introduction
Cotton is cropped over large areas in many subtropical and tropical countries: approximately
31 million hectares (2.4% of the arable land) (ICAC, 2005) is cultivated with cotton. Cotton
production involves about 20 million farmers who completely depend on cotton production and
another 30 million farmers who include cotton into their rotation scheme. Cotton fibre makes
up roughly 50% of world fibre consumption. Cotton fibre consumption is steadily increasing
and a shortage of 15% has been predicted for 2010 (Acordis in Scheffer, 2005). Increasing
attention is being given to the environmental impact of cotton production. As will be outlined
in the subsequent chapters, many cotton production systems face majors problems related
to salinisation, desertification and poisoning of the environment, including ill effects to human
health. The Aral Sea area is the most impressive example of this (cf. Chapter 2; Reganold et al.,
2001; Opp, 2005). Cotton is primarily a non-food crop, which lessens constraints on pesticide
use as compared to those imposed on food crops. In many countries where cotton is cropped,
pesticide use is poorly regulated or not monitored.
The lack of discussion as to the desirability of more sustainable cotton cultivation systems may
well be a reflection of insufficient research and reflection as to the degree of (un)sustainability
of the systems currently in place. In addition, lower environmental impact of Integrated Pest
Management (IPM) and organic systems is usually assumed. As such the goals of this project
were:
1. to summarise literature that compares conventional, IPM, and organic cotton production
systems in terms of environmental impact;
2. to select indicators, based on literature resources, which can be used to formally compare
the environmental impact of conventional, IPM, and organic cotton production systems and to
carry out comparisons based on these indicators;
3. to suggest starting points for discussion on the social aspects relevant for trade in sustain-
able cotton.
Chapter 2 is a general introduction as to cotton cultivation with an emphasis on factors that may
be predominant in affecting sustainability: irrigation, fertilisers, pests and diseases, pesticides,
and genetic modification (Chapter 2.3). In chapter 3, the major types of cotton cultivation
systems are introduced, conventional, organic, and IPM, and a description of these systems
within the major cotton-producing countries is presented. In chapter 4, environmental problems
related to cotton cultivation are discussed using ten potential indicators. Their potential to
define the sustainability of cotton cultivation is discussed. Chapter 5 deals with social and trade
issues related to the cotton production chain. Finally chapter 6 presents a general discussion,
conclusions and recommendations.
- 2 -Science Shop Wageningen UR Report 223
The sustainability of cotton
- 3 -Science Shop Wageningen UR Report 223
The sustainability of cotton
2 Cotton
2.1 The cotton plant
The cotton plant includes 40 species in the genus Gossypium (family Malvaceae), and is native
to tropical and subtropical regions. The seeds are contained in a boll and the fibres on the seed
constitute cotton lint. Species of cotton used for commercial purposes are G. hirsutum (‘Upland
cotton’, native to Central America, the Caribbean and South Florida), G. arboreum (‘Tree cotton’,
South Asia), G. herbaceum (‘Levant cotton’, South Africa), and G. barbadense (‘Creole cotton’
or ‘Sea island cotton’, South America). The naturally occurring colours of cotton lint are white,
brown, and green. However, many coloured varieties have been banned due to fears of genetic
contamination of the white cotton (UNCTAD, 2005). Genetically modified cotton varieties were
developed to reduce insecticide and herbicide use and are now being introduced widely (ICAC,
2005).
2.2 Global cotton production overview
Cotton is produced in more than 100 countries, with 78% of all cotton produced in 6 countries.
China is the main producer with 6.3 Mt of raw cotton, representing 24% of the world’s total
production, followed by the US, (5.1 Mt, 19%), India (4.1 Mt, 16%), Pakistan (2.5 Mt), Brazil
(1.2 Mt) and Uzbekistan (1.1 Mt) (ICAC, 2005) (Table 1). Cotton production per unit surface area
varies considerably from below 429 kg per hectare for India and West Africa up to 1760 kg per
hectare for Australia (Table 1).
Table 1. Selected statistics on cotton production for the major cotton producing countries (ICAC,
season 2004/05).
Average
production
(kg ha-1)
Hectares
cropped to
cotton (x103)
Cotton
production
(Mtonnes)
% of world
production
China 1100 5650 6.3 24
USA 960 5284 5.1 19
India 429 9500 4.1 16
Pakistan 780 3200 2.5 10
Brazil 1150 1020 1.2 4.5
Uzbekistan11300 846 1.1 4.2
West Africa2 450 2400 1.0 3.8
Turkey 1300 700 0.9 3.4
Greece 1030 375 0.39 1.5
Australia 1760 198 0.35 1.3
Syria 1400 230 0.34 1.3
Egypt 940 300 0.28 1.1
Total 29703 23.6 90
World production 31000 26.22100
1 Source: FAS (2005).
2 Mali, Burkina Faso, Niger, Nigeria, Ivory Coast, Togo, Ghana, Benin, Chad, Cameroon.
World cotton production over the past three decades has been subject to varying trends. In
particular the land area cropped to cotton is fairly variable (Figures 1 and 4). However, world
cotton production (Figure 2) and cotton yield per unit area (Figure 3) are steadily increasing.
This is likely due to the development of high-yielding cotton cultivars, especially cotton hybrids.
Locally, the production per unit area shows significant changes. Since 1960 and compared to
- 4 -Science Shop Wageningen UR Report 223
The sustainability of cotton
2001, the area dropped by 7.4% in the former USSR as well as in Sudan (69%), Brazil (50%) and
Mexico (90%). On the other hand, production in Pakistan increased by 27%, Paraguay by 54%
and Iran by 25%. During this period, yields rose on average by 28%, and in India yields in some
areas rose by 74% (ICAC, 2002). The general trend has been decreased production in Latin
American countries and Africa (except Egypt), and an increase in Asian production, mainly due to
a change in labour costs (ICAC, Cotton statistics 2002).
2
2.4
2.8
1950 1960 1970 1980 1990 2000
year
% of world arable land
Figure 1. World cotton area expressed as percentage of
the world arable land (1950-2004) (ICAC, 2003).
0
5
10
15
20
25
1950 1960 1970 1980 1990 2000
year
cotton production
(Mtonnes)
Figure 2. World cotton production (1950-2004) (ICAC, 2003).
200
300
400
500
600
700
1950 1960 1970 1980 1990 2000
year
cotton yield (kg ha -1)
Figure 3. World cotton yield (1950-2004) (ICAC, 2003).
20
25
30
35
40
1950 1960 1970 1980 1990 2000
year
area (Mha)
Figure 4. The world cotton area (Mha; 1950-2004) (ICAC, 2003).
- 5 -Science Shop Wageningen UR Report 223
The sustainability of cotton
2.3 Cotton cultivation
Cotton is considered to be quite a difficult crop to grow because it is sensitive to drought, low
temperatures and attacks by various insects.
Plant spacing normally varies between 50/100 x 30/50 cm. Germination is optimum at
temperatures between 18-30°C, with minimum and maximum temperature 14°C and 40°C
respectively. For early vegetative growth temperatures must be higher than 20°C, optimally
30°C. Proper bud formation and flowering requires daytime temperatures between 20-40°C
and night time temperatures between 12-27°C. Optimum temperatures for maturation and boll
development are 27-32(-38)°C. Cotton is a short-day plant although a few day-neutral varieties
exist. Flowering is influenced by a combination of day length and temperature. Determined by
climate and variety, 50-85 days from planting to first bud formation are required, 25-30 days for
flower formation and 50-60 days from flower opening to mature boll. Both vegetative growth and
boll formation continues during flowering (Figure 5) (UNCTAD, 2005).
Depending on climate and length of the total growing period, cotton needs 500-1300 mm of
water. During the early vegetative period, water requirements are low, about 10% of the total
needs, and during flowering the water requirements are at maximum (50-60% of the total needs)
(UNCTAD, 2005).
Figure 5. The developmental stages of the cotton plant (taken from Walker in FAO-AGLW, 2005).
Cotton is extensively grown under rain-fed conditions. Heavy rainfall can cause lodging although
the crop is relatively resistant to short periods of water logging. Continuous rain during flowering
and boll opening will impair pollination, reduce fibre quality, and provoke boll rot, which is caused
by plant pathogens. Flower buds and young bolls can fall due to heavy rainfall and cold winds
during flowering and maturation.
Cotton is grown on a wide range of soils but medium and heavy textured, deep soils with good
water holding capacity are preferred. Acid or dense sub-soils limit root penetration. The pH range
is 5.5-8.0 with 7.0-8.0 regarded as optimum. The nutrient requirements of cotton under irrigation
are 100-180 kg per hectare N, 20-60 kg per hectare P and 50-80 kg per hectare K. Two-thirds of
the nutrients are taken up during the first 60 days of the growing period (Peel, 1998).
High-technological cotton production (large fields, use of machinery, sometimes precision farming
etc.) typically occurs in the US and Australia while low-technological farming (small fields, often
less than 1 hectare with nearly all practises done by hand and animal traction) occurs in large
areas of China, Africa, and India. The Aral sea area of Central Asia (Uzbekistan, Turkmenistan,
Tajikistan, Kyrgyzstan) has intensive cotton production, but there is large environmental impact
due to high pesticide use and water use through inefficient irrigation systems.
Often cotton is cultivated in rotation with other crops so as to maintain soil fertility and structure,
reduce erosion and to manage soilborne plant pathogenic fungi and nematodes. However, as
- 6 -Science Shop Wageningen UR Report 223
The sustainability of cotton
cotton crops bring in significantly more income than other crops, it can be tempting for the farmer
to reduce narrow rotations or eliminate them altogether. This may lead to increased pressure
from soilborne pathogens and, consequently result in the intensive use of soil fumigants. Rotation
also serves to spread the workload, reduce losses due to unfavourable weather conditions,
and reduce large income variation. On the other hand, with an increasingly wide rotation, the
complexity of farming increases because the farmer must be able to grow a variety of crops.
Additional planning and management skills are required (Peel, 1998).
2.3.1 Irrigation
Cotton requires large amounts of water both for cultivation and processing. Irrigated cotton
cultivation requires 550-950 l per square meter with an average production of 1600 kg raw
cotton per hectare or 550 kg lint cotton per hectare. Otherwise stated, to produce 1 kg of
cotton lint, 10,000-17,000 litres of water are needed. In areas where the normal precipitation
quantity or pattern does not match the requirements of a cultivated crop, irrigation is applied.
For this, rivers must be diverted, dams constructed, or soil water pumped up. Irrigation systems
differ considerably in terms of efficiency, reliability and price. Irrigation is applied to 53% of the
world’s cotton fields, generating 73% of the world’s cotton production because irrigated cotton
on average results in higher yields per unit of area (Hearn in Soth, 1999). It has been estimated
that cotton cultivation accounts for 1-6% of world’s total freshwater withdrawal (Soth, 1999). The
major irrigation systems in use for cotton cultivation are described below:
Flood-or-furrow irrigation. This is the most common type of irrigation. In this system, water
is directed from a river or from deeper soil layers to the cotton field that is to be flooded.
When the cotton plants are planted on ridges the system is referred to as furrow irrigation. It
is a cheap and simple technique with the main investment being for pumps and the levelling
of the fields. For Australia, fixed costs (investment, levelling) were estimated at 24 US$ per
hectare per year and variable costs (handling, electricity) at 26 US$ per hectare per year
(Bryant et al., 2004). The flood-or-furrow system is applied on approximately 95% of irrigated
cotton fields in India, Uzbekistan, Australia, Mexico, the Yellow River region of China and the
United States both on large and small scale farms. Supply driven water distribution exists, for
example, in Pakistan, India, Uzbekistan and Egypt (Gleick, 1993) amongst others and results
in a loss of between 50 and 80% of the freshwater used (Shiklomanov, 1996, in Soth, 1999).
The water efficiency (water that reaches the plant / water at source x 100%) of the flood-or-fur-
row system is approximately 40%, due to evaporation, seepage losses and mismanagement
(Gleick, 1993), although water efficiencies as low as 20-50% have also been reported (Shik-
lomanov, 1996). Elements of mismanagement include: allowing (much) more water than the
plants require which leads to runoff; unworkable distances between the water source and ap-
plication, and leaking irrigation canals (Gilham et al., 1995; Kirda et al., 1999; Reller, 1997).
An important factor related to the inherent low water efficiency of the flood-or-furrow system
is that it is difficult to fine-tune the irrigation system to the needs of the crop.
Mobile irrigation system. An overhead mobile irrigation system is usually made of aluminium
piping with attached wheels that allow the system to be moved. This option has the ability
to irrigate a very large area. Currently it is applied to only 2% of world’s cotton areas (ICAC,
2005). The mobile irrigation system is mostly used in the US and Australia. Trees in and near
the fields must be cut in order to use this irrigation system. The average water efficiency is
high: 80-90% (TWDB, 2003). It is a capital-intensive irrigation system with fixed and variable
costs of about 122 US$ and 49 US$ per hectare respectively. For countries where labour is
expensive it can be an attractive option.
Drip irrigation. A branched polypropylene pipe system doses the amount of water required
by the plant and places the water exactly at the plant’s root. Drip irrigation is being applied
on less than 1% of the cotton fields (Postel, 1999). The technique is relatively expensive.
For example, in the USA the cost is roughly 200–600 US$ per hectare per year (Thompson
& Enciso-Medina, 2002). Developments are however underway to reduce the costs of this
system considerably (e.g. Polak, 2004). For example, there are currently a range of low
cost drip systems on the market for small-holders e.g. in India (see e.g. www.cseindia.org/
dte-supplement/water-index.htm). Drip irrigation can reduce the amount of water used by at
least 16-30 % compared to the most efficient flood-or-furrow systems (Hodgson, 1990), and
is likely much higher when compared to many poorly functioning flood-or-furrow systems. The
average water efficiency of drip irrigation is 90-98%. (TWDB, 2003).
- 7 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Due to the high number of inefficient flood-or-furrow irrigation systems globally, the average
efficiency of irrigation systems is estimated to be 40%, meaning that 60% of all water used for
irrigation does not reach the plant (Stockle, 2001).
The potential to substantially increase irrigated land area is limited. Gains from new capacity are
expected to be largely offset by land losses due to water logging, salinisation, and exhaustion
of water reserves. Currently the most prospective innovation is the implementation of efficient
irrigation techniques (Stockle, 2001), but investment costs are usually lacking.
In addition to optimization of irrigation techniques, the water needs of the cotton crop should
also be evaluated. According to recent studies, cotton can also be grown under controlled water
stress (deficit irrigation) without severe negative impacts on yield and sometimes with quality
improvements (Kirda et al., 1999). Plant breeding may further reduce the water needs of cotton.
The main rain-fed cotton areas are in India, Brazil and West Africa. Yields are often considerably
lower than those of irrigated cotton. Dependency on rain is the main constraint.
Table 2. Dependence on irrigation facilities of some important cotton producing countries (Gilham,
1995, elaborated).
Country Region Rainfall during growing
season (mm) % of cotton area
irrigated Source of
irrigation1
Egypt South 0 100
Central <10 100 –
North <20 100 –
Uzbekistan 30-70 100 – –
Pakistan 100-200 100 –
Mexico 25-2000 84-95 –2
China NW-Xinjiang 0-200 100 – –
N-Yellow River 500-700 75-95 – –
S-Yangtse River 700-1000 75-95
India North 200-400 100 – –
Central 800-1200 29-40 +
South 400-1000 29-40 2
1 – – = severely threatened, – = threatened, + = largely sustainable.
2 Pers. comm. Richard Holland, WWF.
Furrow irrigation at small holder farm, Maharashtra, India. (Kooistra, 2004)
- 8 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Environmental impact
An estimated area as large as 108 (hundred million) hectares (representing 8% of the world’s
total arable land) is abandoned due to former use for intensive cultivation (especially cotton
cultivation, although the share is not precisely known), soil salinisation being the main reason
(FAO, 2002), despite the cotton plant’s moderate resistance to salinity (Ashraf, 2002). Soil
salinisation occurs when evapo-transpiration exceeds rainfall and is as such a threat for irrigated
areas in particular. Irrigation water dissolves calcium carbonate and soluble salts in the soil.
Since calcium carbonate is relatively insoluble, it accumulates in the topsoil leading to additional
salt deposition (originally from the irrigation water) and water logging. Soil salinisation may also
occur when the groundwater of land that sits near a sea is overused. In such cases seawater
replaces the use of fresh water e.g. as has been reported in Israel (Postel, 1999) and Northern
China (East Hebei Plain, Jin et al., 1999).
In the twelve leading cotton producing countries an estimated 12-36% of the area under cotton
cultivation is affected to some degree by salinisation (Stockle, 2001). In Uzbekistan 44% of the
land is now unproductive due to salinisation (Opp, 2005). Current estimates for salinisation-affected
cotton areas are as follows: for India 27-60% of the irrigated land, for Pakistan 14%, Israel 13%,
Australia 20%, China 15%, Irac 50%, and Egypt 30% (Stockle, 2001). The areas facing major
irrigation problems are listed in Table 2. One-third of irrigated land is affected by salinity or is
expected to become affected by salinity in the near future. This share will quite likely be higher for
cotton since it tends to be cultivated in more sensitive regions (Droogers et al., 2001).
Usually a soil is considered to be saline at EC < 4 dS m-1 (EC = electrical conductance, for soils
usually expressed in deci-Siemens per metre (1 dS = 0.1 S), where 1 Siemens = 1 kg -1 m-2 s3 A2 =
1 Ω-1). At this level, germination of cotton seed is hampered and yields are reduced. At EC 12 dS m-1
yield is reduced by 25% and at EC 16 dS m-1 the yield reduction is 50%. Although cotton is relatively
salt tolerant, an example of a shift from cotton to other (even more salt tolerable) crops like wheat
can be found in the salinised regions of Pakistan and Northern India (CAZS, 2005). Saline-tolerant
cotton plants that could be cultivated on moderately salinised soils either through classical
breeding techniques or by genetic modification are under development (Ashraf, 2002; Ganesan
& Jayabalan, 2004).
Apart from the direct effects of unsustainable water use on soil productivity there are additional
significant side-effects. For large-scale irrigation projects, rivers are diverted and dammed,
resulting in significant downstream effects, namely water shortages with significant effects on
wildlife and water availability for human consumption. Surface water levels may drop as a result
of overuse, a notable example being the Aral Sea area. A consequence of the overuse in the Aral
Sea area was the concentration of pollutants in the water. The Aral Sea area has been the main
cotton provider for the USSR since 1938, a situation that lead to significant ill effects, including:
a 75% drop sea area (200,000 km2 in 1920, 50,000 km2 in 2005), a 92% drop in fish species
(1500 to 120 fish species), and an 88% drop in plant species (1800 to 200). In 1920, 60,000
fishermen were active, whereas today the fishing industry has completely collapsed (Reganold
et al., 2001; Calder and Lee, 1995). Water shortages downstream of dam projects is very
common. This is true for the Nile (Egypt), Ganges (India), Yellow River (China), the Colorado River
(USA), and to a lesser extent the Rio Grande (USA) where very little or no fresh water reaches its
final destination for significant stretches of time (Barlow, 2003; Postel, 1999). In these areas,
an estimated 20% (Egypt), 30% (India), 40% (China), 25% (Colorado River USA) and 10% (Rio
Grande, USA) of irrigated land is used for cotton production (Postel, 1999).
The use of groundwater for irrigation projects can be considered unsustainable if more water is
used than that provided by rainfall. In the case of isolated aquifers, groundwater can be regarded
as a non-renewable resource. For example, the depletion of the High Plains Ogallala aquifer
stretching 1300 km from the Texas to South Dakota (USA), is estimated to be 8 times faster than
nature is replenishing it and in the San Joaquin Valley (California, USA), the groundwater table has
decreased by 10 m within 50 years mainly due to agricultural use (Postel, 1999). In the latter
case, however, a relatively small portion of irrigation water is allocated to cotton cultivation. Other
examples of water table depletion as a result of agricultural water use prevails in the literature e.g.
for other Chinese regions (Bejing), and large parts of Australia (New South Wales).
Factors determining the environmental impact of irrigation are the vulnerability of the source
(i.e. is depletion higher than the replenishment), the quality and quantity of the water being used,
likelihood of salinisation, and efficiency of use.
Water use can be reduced by introducing more sustainable irrigation techniques, by breeding
- 9 -Science Shop Wageningen UR Report 223
The sustainability of cotton
for cultivars with increased water use efficiency (e.g. Stiller, 2005) and, as mentioned above,
increased tolerance to salinity. Shifting from flood-or-furrow irrigation to drip irrigation would
significantly increase water efficiency for use in cotton production. These systems could be
combined with ‘irrigation-on-demand’, using GPS-based techniques (Sun & Peng, 2003). However,
due to the high investment costs required, these high-tech developments are not expected to be
implemented in many cotton producing countries in the near future (Raine et al., 2000) unless
subsidies are provided (Luqueta et al., 2005). However, even furrow irrigation can be optimised
through the use of conservation techniques, for example mulches (Jin et al., 1999).
2.3.2 Fertilisers
Common synthetic fertilisers used for cotton are typically combinations of nitrogen (N - usually
in the form of ammonium or nitrate), Phosphorus (P), and Potassium (K). Airborne N2 is the
source of ammonium and nitrate, and phosphorus and potassium are mined. Much of the world’s
phosphate rock originates in the US, China, and Russia’s Kola Peninsula. Smaller deposits are
found throughout North Africa and in the oceans (e.g. near the coast of South Africa). Phosphate
present in guano plays a minor role in terms of worldwide supply. Nonetheless, the Nauru-island
deposit produces annually 2 Mt, and is important to local users in Australia and New Zealand.
Potassium (K20) is produced in the few countries where the economic ore reserves are located:
Canada and the former Soviet Union alone account for 55-60% (IFA, 2002).
Livestock manure is the key nitrogen (N) fertiliser for organic farming. In addition, legumes
cropped in mixture or in rotation with cotton contribute to soil N-fertility and serve in many
cases as feed for animals. Manure and legumes both are options for maintaining plant nutrition
in subsistence farming for which synthetic fertilisers are too expensive. Over the past several
decades, widespread availability of synthetic nitrogen fertiliser in many nations has resulted in a
major decrease in legume cultivation. Manure is applied to the field in either a raw (fresh or dried)
or composted state. Apart from supplying nutrients to the crop, it contributes to soil organic
matter, which in turn stimulates soil biological processes, soil structure, root penetrability and
water retention. The quantity of nitrogen biologically fixed by legumes can be up to several
hundreds of kilograms per hectare (e.g. soy-bean) (Power, 1987).
Environmental impact
To determine the effectiveness on crop growth and the environmental impact of fertiliser use, the
most important factors to consider are: rainfall patterns, types and levels of fertiliser used, and
timing of fertiliser application.
Organic fertilisers include animal manure, green manure crops (usually legumes) and compost.
Apart from nourishing the crop, the goal of these organic fertilisers is to maintain organic matter
An example of mixed cropping: strip cropping of cotton and pigeon-pea, Andhra Pradesh, India.
(Kooistra, 2004)
- 10 -Science Shop Wageningen UR Report 223
The sustainability of cotton
content in the soil (Hansen et al., 2001). A disadvantage of using animal manure as a primary
source of nitrogen is that organic nitrogen must first decompose microbially to ammonium and
nitrate before it can be taken up by plants. Therefore, losses may happen via volatilisation (in
the case of ammonia) or leaching to deeper soil layers (for nitrates) if the supply of mineralised
nitrogen is not synchronised with crop demands (Kirchmann et al., 1998). By contrast, synthetic
fertiliser can be taken up instantaneously by a plant and if an amount is applied according to the
plant’s needs, virtually no fertiliser will be lost to the environment. Nitrogen losses associated
with the use of organic sources must be balanced against the positive effect of carbon additions
to the soil that maintain or increase organic matter in the soil and against the negative effects of
synthetic nitrogen fertiliser (see below). Additional carbon storage in the soil may also increase
the nitrogen storage (Poudel et al., 2001).
Phipps & Crosson (1986) and Nielsen & Lee (1987) (cited in Uri, 1998) estimate that 50-70% of
nitrogen and phosphorus found in surface or groundwater originate from fertiliser applications.
Nitrate moves with water and leaches easily to surface and groundwater. Phosphorus binds to
sediment and contributes to eutrophication if soil is eroded (McDowell & Sharpley, 2004; Smith
et al., 1987). Eutrophication can have a devastating effect on aquatic ecosystems leading to,
for example, algae growth, depletion of oxygen and a decline in aquatic plants and animals (Uri,
1998; Spalding & Exner, 1993).
For the production of nitrate and ammonium much energy is needed, which contributes to global
warming (see Chapter 4). Mining phosphorus and potassium may cause environmental impacts
including changes in the landscape, water contamination, excessive water consumption and air
pollution. The impact on a single farm is, however, negligible and therefore not generally taken
into account when estimating the environmental impact of farming systems (Higgins, 2001).
Crews & Peoples (2004) concluded that the environmental impact of nutrient loss related to
fertiliser type (organic versus synthetic manure) is marginal especially when best management
practises are compared. The energy base for N2 fixation in legumes versus the energy consumed
for synthetic fertilisers is arguably the greatest factor differentiating the two systems (Crews &
Peoples, 2004), giving organic manure a slight advantage over synthetic.
2.3.3 Pests and diseases
Worldwide 15% of cotton yield loss is due to insect damage. The economically most important
insect pests include: Pectinophora gossypiella (pink bollworm), Anthonomus grandis (boll weevil
= bollworm), Earias insulana (egyptian bollworm), Diparopsis castanea (red bollworm), and
Bemisia gossypiella (white fly), all of which have very wide distribution. Among the nematodes,
Meloidogyne incognita (root-knot nematode) and Rotylenchulus reniformis (reniform nematode)
are considered to be the most important (Watkins, 1981).
In general fungal, viral and bacterial plant pathogens as well as nematodes are of lesser
importance in cotton cultivation than insects. In areas without a cold period between cotton
seasons in particular, fungal and bacterial pathogens can be an important source of production
loss. The most important bacterial pathogen is Xanthomonas malvacearum (= X. campestris pv.
malvacearum), the causal agent of bacterial blight. A wide range of fungal plant pathogens are
described in Watkins (1981). Important pathogens include: Fusarium oxysporum f.sp. gossypii
(syn. f.sp. vasinfectum) and Verticillium dahliae, both causal agents of vascular wilting of the
plant. Phymatotrichum omnivorum, Thielaviopsis basicola, Sclerotium rolfsii, and Macrophomina
phaseolina, are all causal agents of root rot. Leaf spots can be caused by a wide array of fungi
and bacteria. The same is true for boll rot: it can be caused by at least eight true pathogens and
various saprophytic fungi can be associated with the rate of boll rotting (Watkins, 1981).
Before the development of synthetic pesticides, pests were managed by cultural practices such
as crop rotation, mixed cropping systems and manipulating planting densities. Also, smaller fields
tend to exhibit reduced impact of pest organisms. Nonetheless, losses have been considerable.
Since the 1950s, pesticides have been used as an efficient solution for addressing most pest
problems (TMC, 2003).
2.3.4 Pesticides
In conventional and IPM systems (for definitions of these systems, see Chapter 3), pesticides are
applied mainly to control insects and to fumigate soil in order to control soilborne plant pathogens.
By using herbicides, weeds are controlled before sowing (or, in the case of glyphosate-resistant
GM-cotton, after seedling appearance), and when harvesting is done mechanically, cotton plants
- 11 -Science Shop Wageningen UR Report 223
The sustainability of cotton
are defoliated prior to harvest. As different pesticide application techniques strongly influence
environmental impact, human exposure in particular, a short overview of the various techniques
used is presented below.
Aerial spraying. An estimated 13% of cotton growing areas are sprayed via airplanes (ICAC,
2005), resulting in fast but often inaccurate application and considerable risk in terms of
pesticide drift. Cases of contamination of neighbouring villages have been reported for cotton
production in California (Scarborough et al., 1989), worker contamination in Uzbekistan (IATP,
2002) and water, crop, and cattle contamination in Colombia (CIP, 2005).
Tractor spraying. Approximately 35% of the cotton growing areas are sprayed using tractors
(ICAC, 2005). Tanks containing pesticides are attached to the back of the tractor and beams
are used to cover as wide a range as possible.
Field spraying. Hand pumps and power sprayers are applied in 35% and 17% of cotton fields
respectively (ICAC, 2005). It is cheap and effective, but poisoning of the workers who apply
the pesticides is common in developing countries (Friedrich, 2001; Mancini et al., 2005). A
power sprayer has a wider range, so a field is sprayed more quickly with a power sprayer
than with a hand pump. In both cases the field worker sprays the area ahead of him/her and
then walks through the freshly sprayed plants. A gas mask for protection against the pesticide
sprayed is indispensable for worker health, although they are still very often not used in devel-
oping countries.
Bos et al. (2005) made the following recommendations for conducting “Good Practice for the
Application of Pesticides”:
• spraying of pesticides must be carried out by trained personnel. This should be proven by
certificates or be made credible in other ways;
• spraying should be carried out with the appropriate, calibrated and well-maintained equip-
ment;
• only (nationally or regionally) approved and correctly labelled pesticides should be used;
• labels should be retained for use in the case of accidents;
• health of the operators should be monitored by keeping health records and ensuring regular
medical checkups;
• pesticides should be stored and transported in their original transport containers and pack-
ages and out of reach of people, particularly children, and domestic animals;
• spillage incidents must be accurately recorded and appropriate authorities must be notified of
larger spills;
Unprotected field spraying of pesticides inflicts great risks to the farm. (Kooistra, 2004)
- 12 -Science Shop Wageningen UR Report 223
The sustainability of cotton
• an action plan should be in place to combat larger spills, including access to first aid equip-
ment in the case of personal contamination and poisoning;
• during spraying, operators should use sufficient protective clothing. This should at least in-
clude impenetrable boots, impenetrable gloves, overalls, face or mouth masks, hats and eye
and face protection. These garments should not be worn during other occasions or be used
for other purposes;
• similar protective clothing should be worn when handling the pesticides pre-treatment (dilut-
ing, pouring into the spraying equipment) and post-treatment (rinsing, disposal);
• there should be a system with which all relevant information of pesticide treatments is record-
ed and archived. This should at least include the place and date of the treatment, the amount
and type of pesticide used, the location and surface of the sprayed area, the names of the
operators, and the equipment used;
• operators should carefully wash themselves after spraying, particularly exposed body parts,
and certainly before eating, drinking, smoking or going to the toilet;
• empty containers and packages should be rinsed and/or cleaned thoroughly and be made
unsuitable for second-hand use;
• washing of the containers, packages and equipment should take place at special cleaning
stations where contamination of soil, surface and drinking water is avoided;
• contaminated containers, packages and clothing should be disposed of in the appropriate
ways according to (national) legislation;
• spraying equipment should be stored dry, empty and out of reach of people, particularly chil-
dren, and domestic animals.
The uptake of the above-mentioned practices requires the basic conditions of farmers in
developing countries to be improved with regard to literacy, information and access to credit
(other than from seed/chemical companies).
The selection of pesticide type and quantity depends on location and type of cultivation. In
subsistence farming, often no pesticides are used due to lack of capital or because they are
simply not available. Although not officially certified, such farms could be considered as de
facto organic. On small holder cotton farms however, highly toxic pesticides are often still used,
including pesticides classified by the World Health Organisation (WHO) as highly or extremely
hazardous (Table 3). Thus, monocroptophos and methamidophos, both classed by the WHO
as ‘highly hazardous’, still constitute 9% and 6% of the cotton insecticide market (Table 3). In
addition, various pesticides classed as ‘extremely hazardous’ are still in use though information is
not available as to the volumes (Table 3). India is an example of a country (amongst others) where
some of these pesticides, such as parathion, while no longer officially permitted, due to poor
monitoring and verification systems, continue to be in use. In Kyrgyzstan, banned pesticides
are smuggled into the country (Hadjamberdiev & Begaliev, 2004). Many older, non-patented,
toxic, environmentally persistent and inexpensive chemicals are used extensively in developing
nations, creating serious acute health problems and local and global environmental contamination
(Ecobichon, 2001; Wesseling et al., 2005). The quantity of these illegal pesticides is very difficult
to estimate. In more developed countries, pesticides are regulated and hazardous insecticides
are not allowed.
The cultivation of cotton has been estimated to consume 11% of the world’s pesticides while
it is grown on just 2.4% of the world’s arable land. For insecticides alone, this represents 25%
of world consumption (FAOstat, 2000). In developing countries it is estimated that roughly 50%
of all pesticides are applied in cotton cultivation (Caldas, 1997). In addition, pesticide use and
storage are often badly managed. In India two pieces of out-of-date legislation regulate pesticide
use: the Insecticide Act (IA), which was passed in 1968, and Insecticide Rules (IR), passed in 1971.
Toxicological data are required in order to apply for registration of pesticides, however, they are
still registered without taking into full account the acceptable daily exposure and maximum residue
levels. In addition, the fact that cotton is often regarded as a cash crop means that farmers may
be more willing to increase their yields even given the environmental costs (CIBRC, 2005).
- 13 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table 3. Major pesticides used to control insect pests in cotton according to Allan Woodburn
Associates (1995*).
Chemical name (trade name) Chemical group1WHO toxicity
class 2
% Global cotton
insecticide
market3
deltamethrin (Decis) p II 12
lamda cyhalothrin (Karate) p II 9
monocrotophos (Azodrin) o Ib 9
alpha-cypermethrin (Fastac) p II 8
chlorpyriphos (Dursban, Lorsban) o II 7
esfenvalerate (Sumi-alpha) p II 7
methamidophos (Tamaron B) o Ib 6
dimethoate (Rogor; Perfekthion) o II 5
thriazofos o Ib n.k.4
demeton-S-methyl Ib n.k.
parathion p Ia n.k.
parathion-methyl p Ia n.k.
phosphamidon Ia n.k.
* No more recent data known by author.
1 p = pyrethroids, o = organophosphate.
2 Ia: extremely hazardous Ib: highly hazardous; II: moderately hazardous.
3 The remaining 37% are smaller amounts of another 250 insecticides (Allan Woodburn Associates, 1995).
4 n.k. = not known.
Until 1995, the organophosphate group, one of the most hazardous in terms of worker’s risks,
represented the majority of the insecticide market in cotton cultivation in many developing countries.
A shift has taken place towards much less persistent and less toxic pyrethroids in recent years.
Organophosphates include aliphatics (malathion, monocrotophos (Azodrin), oxydemetonmethyl
(Meta Systox), dicrotophos (Bidrin), acephate (Orthene)), phenyl derivatives (parathion, methyl
parathion, profenofos (Curacron), fenthion (Dasanit)), and heterocyclic derivatives (chlorpyrifos
(Dursban, Lorsban, Lock-On), chlorpyrifos-methyl (Reldan)). While all are significantly toxic to
vertebrates, most are not very persistent. The pesticides on this list are substitutes for the more
persistent organochlorines. Due to their high toxicity, organophosphates are being banned in many
countries. They affect the nervous system, which results in rapid twitching of voluntary muscles and
finally paralysis. In the 1990s some organochlorines (e.g. endosulphan) were re-introduced in West-
Africa as they became cheaply available and were seen as highly effective (Ajayi et al., 2002).
The pyrethroids include permethrin, introduced in 1972-1973 and are used because of their
insecticidal activity (0.1 kg active ingredient per hectare) and photostability. The following
trademarks are permethrin-based: Ambush, Astro, Dragnet, Flee, Pounce, Prelude, Talcord,
and Torpedo. The current generation of pyrethroids has even higher effectiveness than their
precedents (ranging from 0.01 to 0.06 kg active ingredient per hectare). These include bifenthrin
(Capture, Talstar), lambda-cyhalothrin (Demand, Karate, Scimitar, Warrior), cypermethrin (Ammo,
Barricade), deltamethrin (Decis), and zeta-cypermethrin (Mustang, Fury) (Ware et al., 2004).
Pyrethroids have an effect resembling that of DDT. Pyrethroids affect both the peripheral and
central nervous system of insects, stimulate nerve cells to produce repetitive discharges and
eventually cause paralysis. Many of these compounds are also extremely toxic to fish. Pyrethrins
and pyrethroids adsorb strongly to soil and volatilise from soil surfaces very slowly. These
compounds are eventually degraded by microorganisms in the soil and water. They can also be
degraded by sunlight at the surface of water, soil or plants. However, some of the more recently
developed pyrethroids persist in the environment for a few months before they are degraded.
Herbicides are used to control weeds in the fields. The most important herbicides worldwide in
cotton are listed in Table 4.
- 14 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table 4. Major herbicides used for cotton production in 1997 (USDA, 1998).
Active ingredient Area applied1 (%) WHO tox.
Class2Applications
(cultivation-1)Dose (kg ha-1 yr-1)Total applied
(Mkg)
triazines
cyanazine 18 Ib 1.3 1.1 1.0
prometryn 19 U 1.2 0.74 0.76
other herbicides
trifluralin 55 U 1.1 0.85 2.5
MSMA 29 n.l. 1.4 1.46 2.2
fluometuron 44 U 1.3 0.94 2.2
pendimethalin 28 III 1.1 0.77 1.1
norflurazon 13 U 1.0 0.71 0.47
diuron 12 U 1.1 0.62 0.40
metolachlor 5 III 1.1 1.31 0.34
glyphosate 14 U 1.3 0.91 0.70
paraquat n.k. II n.k. n.k. n.k.
1 n.k.= not known.
2 Ia: extremely hazardous Ib: highly hazardous; II: moderately hazardous, III = slightly hazardous, U = Unlikely to be
hazardous, n.l. = not listed.
Overall, cyanazine (WHO toxicity class II, moderately hazardous) is the most toxic herbicide
for human beings, and may be the most toxic of all herbicides found in drinking water (Eclipse
Scientific Group, 2004) (Klotz-Ingram, 1999). Paraquat is also moderately toxic. For defoliation
the main herbicides applied are: dimethipin (WHO toxicity class III - slightly hazardous), thidiazuron
(WHO toxicity class U - unlikely to be hazardous), sodium chlorate (WHO toxicity class III), diuron
and tribufos (WHO toxicity class not listed).
Insecticide sales are globally approx. 9.1 billion US$ in 2000 (EPA, 2002) and 23% of this is
applied in cotton cultivation (Dinham, 2005). The agrochemical market is dominated by Bayer
(20% of the world market for pesticides), Syngenta (17%), BASF (15%), DOW (14%), Monsanto
(6%), and DuPont (6%) (EPA, 2002; Dinham, 2005). The use of pesticides will always lead to
the exposure of non-target organisms and unwanted side-effects that occur in some species,
communities or complete ecosystems (Van der Werf, 1996).
Pesticides arrive at non-target sites through various channels: by aerial drift, run-off, evaporation,
seeping, leakage to groundwater, washing of pesticide spray equipment, and the use of pesticide
packages or buckets (used for mixing pesticides) for other purposes. Pesticides are removed
from the environment by chemical degradation or microbial decomposition (Papadopoulou-
Mourkidou et al., 2004).
Usually the negative side-effects of pesticide use are not calculated as costs (Pimentel, 2005).
Obsolete pesticide stocks in developing countries are another pressing problem (Schimpf, 2004).
The most significant factors influencing the environmental impact of pesticides are pesticide
toxicity, spraying method, frequency of use and quantity used per application.
The so-called ‘pesticide treadmill’ emerges when dependency on pesticides increases over
time reaching levels that are both economically and technically unfeasible (Twigg et al., 2002;
Nicholls & Altieri, 1997). The development of such dependency is more likely to occur for cash
crops (where loss of income is significant if there is a poor harvest), non-food crops (since
pesticide residues present on the product are not ingested by people), and when pest or disease
organisms seriously jeopardise the harvest. The pesticide treadmill was recently summarised
for a cotton system in Central America (ISU, 2004). Cotton production increased steadily in
the region from 1945 onwards. The presence of the Boll Weevil made pesticide application
necessary. The effectiveness of the pesticides to control the Boll Weevil decreased over time
due to increased resistance of the insect (e.g. Kranthi et al., 2002) and the inactivation of
natural predators, leading to 40 times the initial levels of pesticide application per season in
- 15 -Science Shop Wageningen UR Report 223
The sustainability of cotton
the 1970s. Finally, cotton cultivation became uneconomical, leading to soils being left fallow
due to pollution and erosion, and very high rates of unemployment and associated problems
(ISU, 2004). Currently over 500 species of insects, 180 weeds and 150 fungi have developed
resistance to the pesticides developed for their eradication (Lotus, 2004). Other examples of
pesticide treadmills have been described in Peru by Smith & van den Bosch (1967) and in Sudan
by Eveleens (1983).
Human Impact
The potential negative consequences of pesticide use for human health are now widely recognised,
however statistics are still rare. It has been estimated that at the global level annually 40,000
lives are lost due to pesticide application (WHO, 2002), representing 10% of all casualties in the
agricultural sector (ILO, 1997). Although higher numbers have been asserted (Gaag, 2000) it
is difficult to make reliable estimates due to a general lack of data and the inability to separate
pesticide exposure from other causes of disease.
In India, Mancini et al. (2005) followed the health of 50 cotton growers over the course of
one year during which she reported 323 events, of which 207 (84%) were associated with
symptoms of mild to severe pesticide poisoning, and 32 (10%) were associated with symptoms
typical of poisoning by organophosphates (used in 47% of the pesticide applications). Although
in 6% of the spray sessions the workers’ neurotoxic effects were serious, none sought medical
care (Mancini et al., 2005). Scarborough (1989) reported 60-100% higher incidence of various
disease symptoms (fatigue, eye irritation, rhinitis, throat irritation, nausea, and diarrhoea) for
humans living or working near sprayed cotton fields relative to the comparison group. Lotus
(2003) reported that 91% of professional pesticide workers experienced some type of health
disorder. In Egypt in the 1990s, more than 50% of cotton workers suffered symptoms of chronic
pesticide poisoning, including neurological and vision disorders. Benin has suffered from severe
Endosulphan poisonings since the 1990s (Tovignan et al., 2001).
In 1985, in Malaysia and Sri Lanka 7-15% of farmers reported experiencing pesticide poisoning at
least once in their lifetimes. In Thailand, a survey of 250 government hospitals and health centres
revealed that approximately 5500 people suffered pesticide poisoning, of whom 384 died. In the
Philippines 50% of rice farmers have suffered from sickness due to pesticide exposure. In Latin
America 10-30% of agricultural workers show inhibition of the blood enzyme, cholinesterase,
which is an indication of organophosphate poisoning. In Venezuela, approximately 10,000 cases
of pesticide poisoning with 576 deaths occurred between 1980 and 1990. In the state of Santa
Catarina (Brazil), 28% of farmers say they have been poisoned at least once. Likewise, in Parana
state approximately 7800 people were poisoned through pesticide application between 1982
and 1992. In the USA, a 1987 National Cancer Institute Study found a nearly 7-fold higher risk of
leukaemia for children whose parents used pesticides in their homes or gardens (Lotus, 2004). In
China, 42,800 new cases of pesticide poisoning were reported in 1994, including 3900 fatalities.
Many were said to be victims of home-made cocktails marketed illegally (Gaag, 2000).
Multiple reports exist on the unwanted side-effects of pesticides on wildlife. Over-spraying,
accidents and aerial spraying are the most significant events affecting the environment (Mkuula,
1993; Carvalho et al., 2003). Pesticides end up in aquatic organisms (Kumar et al., 2003;
Muschal & Warne, 2003; Erdogrul et al., 2005) and have been shown to bioaccumulate (Zhang
et al., 2005) into birds (Eason et al., 2002).
Impact on the environment
Pesticides applied in cotton production have also been documented as adversely affecting river
ecosystems in Australia, leading to lower quantities and lessened diversity of water organisms
(Hose et al., 2003). In 1995, pesticide-contaminated runoff from cotton fields in Alabama killed
240,000 fish (Lotus, 2004). It is estimated that pesticides unintentionally kill 67 million birds
each year (Lotus, 2004). Tariq (2003) reported pesticide contamination of groundwater due to
cotton cultivation in Pakistan and Shukla et al. (2005) and for Hyderabad in India. CSE (2003)
reported pesticides in the main brands of cola and packaged drinking water in India. Broad-
spectrum pesticides very likely have a negative impact on multiple non-target insects and other
micro-organisms, but to precisely estimate the effects is arduous.
- 16 -Science Shop Wageningen UR Report 223
The sustainability of cotton
2.3.5 Defoliation
Defoliation is done in order to clear the cotton plant of leaves before it is mechanically stripped
during the harvest. Chemical defoliation is quick and effective. Low temperatures and forced
water deficit also serve as natural defoliants (Chaudry, 1996). Chemical defoliants are estimated
to be applied on 15% of the world’s cotton production (ICAC, 2005). Most commonly applied
defoliants are carfentrazone (not listed (n.l.) in the WHO toxicity classification), pyraflufen ethyl
(n.l.), dimethipin, sodium chlorate, thidiazuron, MSMA (WHO toxicity class III, see also Table 4)
and diuron (U) (Vargas et al., 2005). The environmental impact of defoliants on the natural and
human environment resembles the effect of pesticides described in more detail above.
2.3.6 Yields
Farms demonstrate clear yield differences per unit of area. It is currently not clear whether
production in conventional systems is different from that of organic systems, one of the reasons
being that there is also a large within-system variability. For that reason, in Chapter 4 on
environmental impact of cotton production systems, we divide conventional, IPM and organic
systems all in ‘high input’ and ‘low input’ systems (cf. Table 10). Recent sources indicate yields
in organic systems equal those of conventional systems (rain-fed, India: personal communication
Eyhorn, Maikaal project; rain-fed, Paraguay: personal communication Ecklin, Arazy project; rain-
fed, India: personal communication Lanting, ETC; Meyer, 2005). Other sources indicate a 25%
drop in yields in the second year after conversion to organic for a highly productive farm in
California (Swezey et al., 2004). Recent findings suggest that low-input cotton cultivation systems
can even have slightly higher production as compared to certain conventional production systems
(India: personal communication Eyhorn, 2005; Arazy project, Paraguay; Ecklin, 2005). In Turkey,
the first few years after conversion to organic production yields were halved but three years later
yields returned to the previous (conventional) yields (Zeegers, 1993).
The interpretation of comparisons between different farmers or farming systems can be
prone to errors. There have been few organic projects and these projects likely have received
considerable external support (educational, financial) while perhaps the majority of conventional
farmers have not.
2.3.7 Genetically modified cotton
One of the first plants to be modified genetically at a commercial scale was the cotton plant, mainly
in an effort to reduce the quantity of pesticides used. A gene conferring resistance to glyphosate
(an active ingredient in herbicides such as Roundup) was transferred into cotton for the first
time in 1987. Monsanto introduced the Bt-cotton variety in 1989. The Bt-toxin-producing gene
of Bacillus thuringiensis (Bt) was introduced into the plant’s genome resulting in plant resistance
against pests, notably the Bollworm. In 1996 Bt-cotton was first planted at a commercial scale in
Australia and the USA. Since then, Bt and glyphosate-resistant cotton varieties have been planted
in more than 20% of the area under cotton cultivation (IIED, 2004). 50% of the cotton cropped in
Mexico and South Africa has been genetically modified, compared to 80% in the USA and 66% in
China. Argentina, Australia, India, and Indonesia have also approved the commercial planting of
genetically engineered cotton in recent years (UNCTAD, 2005).
Since its release, Genetically Modified (GM) cotton has been the topic of considerable discussion.
It is however not the goal of this study to present an extensive overview of the GM debate. Here,
we limit ourselves to a short overview. In general, IPM and conventional systems are adopting
genetically modified cotton (Bt-cotton and glyphosate-resistant cotton) while organic standards
and regulations do not permit genetic modification.
On the positive side genetically modified cotton is said to reduce pesticide use. Significant and
high reduction in pesticide use for Bt-cotton was reported by Edge et al. (2001; review of 12
studies). Bt-cotton has also been widely reported as leading to increased cotton yields (e.g.
Layton et al., 2003; Robinson & McCall, 2001). Higher revenues have also been noted (e.g.
Morse et al. (2005) in South Africa; Pray et al. (2001) in China; Traxler et al. (2005) in Mexico).
However, the explanation of this reduction in pesticide use is unclear. One explanation could
be that the Bt insecticide in the plant itself is not acknowledged: since the genetically modified
plant contains an insecticide, it is obvious that no insecticide needs to be applied to control the
insects concerned. On the negative side, pests that are not sensitive to Bt have been observed,
which was suggested by ICAC (2005) for regions in China where the occurrence of red spider
mites and aphids were reported (see also Shelton, 2005). Other negative features of Bt-cotton
- 17 -Science Shop Wageningen UR Report 223
The sustainability of cotton
include that the farmer may have a higher economic risk due to the genetic homogeneity of the
cotton, seed is more expensive, the farmer is more dependent on the seed producer, there is
a possibility of outcrossing with wild relatives, environmental risks (the spread of GM genes to
related wild species with unknown consequences), and the lack of clarity as to how the costs
associated with expensive genetically modified crops will affect rural farming communities in
developing countries (Altieri & Rosset, 1999).
The most recent data generally agree that effects of Bt-crops are minimal on both airborne and
soilborne organisms. While Losey et al. (1999) reported that Bt-pollen negatively affected the
larvae of the protected monarch butterfly, this work was criticised (e.g. Crawley, 1999) and
the more widely held scientific opinion now is that the monarch butterfly will not be significantly
affected by Bt-corn (Stanley-Horn et al., 2001). Pilcher et al. (2005) reported 29-60% lower
densities of Macrocentrus cingulum, a parasitoid (bio-control agent) of the European corn borer,
in Bt-corn compared to control corn. Sims and Ream (1997) estimated that a potential maximum
of 1.6 µg Bt-toxin (CryIIA protein) per gram
soil would be formed during the decomposition of
residues of Bt-cotton. Saxena et al. (2002) reported that the Bt-toxin is quite persistent in soil.
It was found that root exudates from Bt-maize releasing the Bt-toxin remained insecticidal for up
to 180 days, but no effects were found on protozoa or the microbial community, and only very
minor, but significant, effects on the nematode community (Saxena et al., 2002; Griffiths et al.,
2005; and references in these publications). Given however the huge areas that are now and will
be in the future planted with Bt crops, continued research on potential negative side-effects is
still needed.
Glyphosate-resistant cotton may reduce the use of herbicides since it can be applied more
efficiently and thus less frequently. However, in conventional, non-genetically modified systems
in Canada, applications of herbicides are already low (Blackburn & Boutin, 2003) so no further
reduction in herbicide applications is expected. Weeds that have developed glyphosate-resistance
have been reported (Nandula et al., 2005) and this may increase the use of multiple herbicides.
Lotz et al. (2000) concluded that the available literature does not indicate that cultivating
glyphosate-resistant crops has major effects on the environment compared to conventional
crops, though they also identified some important knowledge gaps.
IFOAM (cited from the website: www.ifoam.org, end of 2005) is opposed to genetic engineering
in agriculture due to the ‘unprecedented danger’ it represents for the entire biosphere and the
particular economic and environmental risks it poses for organic producers. IFOAM believes that
genetic engineering in agriculture causes, or may cause:
• negative and irreversible environmental impacts;
• release of organisms which have never before existed in nature and which cannot be recalled;
• pollution of the gene-pool of cultivated crops, micro-organisms and animals;
• pollution of farm organisms;
• denial of free choice, both for farmers and consumers;
• violation of farmers' fundamental property rights and endangerment of their economic
independence;
• practices which are incompatible with the principles of sustainable agriculture;
• unacceptable threats to human health.
The general argument for not permitting genetic modification in organic agriculture is that
this technique is synthetic and thus not natural. However, the natural, non-synthetic biocontrol
product Bacillus thuringiensis is allowed as a natural pesticide in organic agriculture (e.g. Zarb
et al., 2005).
2.3.8 Clearing of forest/natural area
Land clearing has been done since the beginning of agriculture. If additional land is needed
for agriculture this often results in deforestation, conversion of grasslands, or draining of
wetlands. For some tropical areas slash-and-burn systems are still part of traditional agriculture.
Deforestation is particularly unsustainable if the soil is not suitable for long-term annual farming,
resulting in fast degradation (Roper & Roberts, 1999). The main land clearing regions are situated
in the Amazon, Central America, China and Indonesia (Wright, 2005). The major consequences
of deforestation are loss of biodiversity, soil erosion and contributions to the greenhouse effect
(Costa & Foley, 2000; Gash & Nobre, 1997).
The majority of deforested areas used for agriculture are not directly used for cotton production.
For example, in Brazil, where cotton is being introduced into rotations with soybean, this is
- 18 -Science Shop Wageningen UR Report 223
The sustainability of cotton
becoming a relevant issue as the soybean expansion is converting natural grasslands and
tropical rainforest. In west and east Africa cotton is grown by livestock farmers who move their
herds around and create shifting patterns of cultivation. Land clearing is however still relevant for
cotton when the yield per unit area changes as a result of changing farm management practices.
Examples where deforestation has been carried out specifically for increasing the cultivated
area of cotton have been described by Benjaminsen (2002) and Moseley & Ikubolajeh Logan
(2004) for Mali (Benjaminsen, 2002), by Grauel (1996) for Paraguay, by the FAO (1999) for China
(FAO, 1999), by the USDA (2001) for Brazil, and INSAE (1999) for Benin. For example, Benin
receives 80% of its export revenues from cotton and as a consequence forests are cut for other
crops. In total in Benin 100,000 hectares of forest is cut annually (Insae, 1999) and currently
approximately 2000 hectares is converted to cotton production.
On a local scale agricultural policies, especially subsidy systems, changes in credit systems and
provision of certain guarantees, may affect the total agricultural area and/or the area cropped
to cotton. This is especially of concern where cotton is regarded as an export crop as sudden
significant changes in land use may lead to deforestation as occurred in the pesticide treadmill
example described in Chapter 2.3.4. Another example is provided by Chipika & Kowero (2000)
for Zimbabwe.
- 19 -Science Shop Wageningen UR Report 223
The sustainability of cotton
3 Types of cotton production
Cotton can be grown using a wide range of cultivation practises determined by climate, soil
type, availability of inputs, and possibly most importantly, the knowledge and skills of the farmer.
Globally, cotton production is done by a relatively small number of large, mechanised farmers
(Australia, USA, Brazil) and a very large number of smallholders (China, South Asia, West Africa)
(IIED, 2004). It is estimated that there are approximately 20 million cotton farmers worldwide
(Gala, 2005).
Within all the different management systems a few distinguish themselves by having a robust
certification system (organic production) or being a technique well-described by the FAO and
research institutes (Integrated Pest Management - IPM). We regard ICM (Integrated Crop
Management) synonymous with IPM. All management systems that are not IPM or organic are
considered to be conventional. Fair Trade systems require IPM with a strong social paragraph
(cf. Chapter 5).
These three main approaches to cotton production are described in the subsequent paragraphs and
the specifics of these farming systems for the major cotton growing regions are presented.
3.1 Conventional systems
Conventional farming systems refer to all farming systems that are neither organic nor IPM.
Usually farming practices are set either by legal restrictions or by the availability of capital.
Conventional farming includes a wide range of farms since size, agrochemical inputs/outputs,
energy efficiency, production efficiency and the related environmental impact differ enormously
among farms. Roughly 80% of world’s cotton is produced under conventional management
(ICAC, 2005).
In developing countries, farms are usually small (less than 5 up to 20 hectares, rarely more).
These farms tend to be family-run and are usually part of a mixed cropping system. This multi-
purpose farm produces food for family consumption and animal feed, and generates an income
for the family through the cultivation of cash crops like cotton. Currently food security among
smallholders, especially in West Africa, is being affected by pressure to produce cotton, which
seems to be replacing mixed cropping systems (Williamson, 2003).
Intermediate-sized farms are for the most part situated in developing countries, and are usually
family–run. They grow more cash crops than the smaller farms.
Larger farms (mostly greater than 20 hectares and up to 1200 hectares) are situated mainly
in the USA and Australia. Continuous cropping (i.e., without crop rotation) is the main cropping
system and is designed to generate as high an income as possible for the owner. In Uzbekistan
only a few very large farms can be found. The degree of mechanisation is high since areas are
too large to be picked by hand and the wages are high and costs of mechanisation are relatively
low (ICAC, 2005).
- 20 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table 5. Frequency distribution (%) of farm sizes in the major cotton producing countries.
Country Farm size (ha) Total surface
6
(Mha) % World
production6
< 6 6-10 10-20 20-100 >100
China 100 0 0 0 0 5.6 24
US10 0 0 10 90 5.3 19
India 87 8 3 2 0 9.5 16
Turkey250 19 32 0 0 0.7 3
Pakistan36 90 4 0 0 3.2 10
W-Africa 100 0 0 0 0 2.4 4
Uzbekistan410 0 40 50 0 0.8 4
Australia 0 0 0 0 100 0.2 1
Egypt590 5 5 0 0 0.3 1
Mexico 5 3 65 27 0 0.2 0.4
share of world
production (%) 45 11 4 4 18 31 82.4
1 USDA, 2005.
2 Turkey Case Study: Aslantas Dam / Ceyhan River Basin (1983).
3 Pakistan Census of Agriculture (1990).
4 State Committee on Statistics (2004).
5 ICAC, 2002.
6 ICAC, 2005.
The data per country are mainly derived from the September 2005 ICAC survey of cotton
cultivation practises. This survey was done by national panels mainly consisting of staff from
ministries of agriculture with information centrally collected by the ICAC. Since the sources are
diverse, the information provided is mainly indicative.
China (24% of world production, 5.6 Mha)
China is the largest cotton producing country in the world with about 14 million farmers directly
involved in cotton production (IIED, 2004). From 1978 to 1984, China’s cotton output increased
steadily, reaching 6.3 Mt in 1984; after 1984, cotton output fluctuated between 3.8-5.7 Mt, and
again reached 6.3 Mt in 2004/05. China has three major cotton growing regions, the Xinjiang
Autonomous Region, the Yangtse River Basin Region (which includes principally Jiangsu and
Hubei), and the Huang-Huai Region (principally Hebei, Henan, and Shandong). The farm types in
each production area are quite similar. Since China joined the WTO in 2001, both opportunities
and challenges have arisen in the agricultural sector. Low production costs in China are a key
factor for successful competition in the international cotton market (ICAC, 2002).
Chinese cotton farms are generally smallholders (less than 6 hectares; see Table 5), cultivating
cotton on a small segment of their farm. 90,000 farms cultivate cotton (ICAC, 2005) with
an average production of 1100 kg per hectare, and a range from 150-1800 kg per hectare.
Virtually all cotton farms are irrigated: 90% with flood-irrigation, 5% furrow, 3% drip, and 2%
sprinkler irrigation. The water originates from soil (74%) and rivers (26%). Fertiliser use per farm
averages 225-300 kg N per hectare, 50-150 kg P per hectare, and 100-220 kg K per hectare.
Approximately 60% of the area under cultivation is on sandy loam, 10% sand, and 30% clay.
Weeds and cotton stalks are burned after harvest. A common rotation cycle is done with wheat:
one year wheat and one year cotton. To a lesser degree rotations with other crops such as onion,
garlic and potatoes are applied. Mixed cropping is rare (but see Van der Werf, 1996).
Most commonly used pesticides are monocrotophos (Azodrin, Ib), cyfluthrin (Baythroid, II),
dicofol, (III), trichlorfon (Dipterex, II), imidacloprid (II), chlordimeform (Fundal, n.l.), and piperonyl
butoxide (Pyrethrin, n.l.). Herbicides include: MSMA/prometryn (Caparol, III/U), glyphosate
(Roundup, U), alachlor, (III) and fluometuron (Cotoran, U). On average the fields are sprayed 20
times per season and small-scale tractors are used for 90% of the land labour including pesticide
- 21 -Science Shop Wageningen UR Report 223
The sustainability of cotton
application. Defoliants are used on 5% of the surface, with the most important defoliant being
MSMA (III). All cotton is hand-picked. GM cotton is widely accepted in China and constitutes 66%
of the crop (UNCTAD, 2005).
UNEP (2002) advices to promote new hybrid cotton varieties (higher yield and higher pest
resistance), to produce special purpose cotton like naturally pigmented cottons and special
fibre length and strength, to establish cotton production cooperatives, to carry out integrated
assessments of the environmental impact of cotton, and to decrease chemical fertiliser
application.
United States (19% of world production; 5.3 Mha)
American cotton farms are situated in the warm southern states, mainly in Arizona, Texas,
Florida, Georgia, California, Kansas, New Mexico, and Texas. In total 25000 farmers cultivate 5.3
Mha of cotton producing land. The cotton farms are usually very large (around 1200 hectares)
and intensive monocultures using high-yielding varieties without rotation crops. The average
production is around 960 kg per hectare, varying between 475 and 1726 kg per hectare.
Of the cotton fields 41% are irrigated (flood irrigation 50%, sprinkler irrigation 50%). Quantities
of water applied differ widely between regions and exact data is not available. Fertiliser use
averages 100 kg N per hectare, 90 kg P per hectare and 80 kg K per hectare. Almost all
nutrients are applied as synthetic fertiliser. Cotton stalks tend to be left on the fields. The
most frequently used insecticides are ferbam/parathion (Carbamate, U/Ia), organophosphates
(ranging I-U) and pyrethroids (ranging II-U). An important fungicide is benzoate (n.l.). Important
herbicides are dinitroaniline (soap, n.l.). It is estimated that 25% of all pesticides applied in the
USA in agriculture are used for cotton, while cotton is cultivated on only 4% of agricultural land
(USDA, 2005). Defoliants are applied on 70% of the surface, most commonly carfentrazone
(n.l.), pyraflufen ethyl (n.l.), dimethipin (III), sodium chlorate (III), thidiazuron (O), and diuron (U)
(Vargas, 2005). Genetically modified cotton covers approximately 80% of the cotton growing
area in the US (UNCTAD, 2005).
Since the price of fertilisers, chemicals and recently energy have increased, cotton has be coined
‘poverty weed’ in the US (Meyer and MacDonald, 2001). Virtually all cotton farms in the USA are
classified as poor or marginal (TA&M research, 2003). Still, subsidies of US cotton farmers
(which number only 25,000) play a dominant role in recent WTO negotiations as they affect the
export position of many cotton exporting developing countries. An American cotton farm must
be at least 200 hectares to be economically viable, due to the need for economies of scale in
order to justify high capital costs and machinery investments (TA&M research, 2003). For the
US economy, cotton is marginally important, constituting in 2004 a mere 0.2% of the Gross
Domestic Product (USDA, 2005).
India (16% of world production; 9.5 Mha)
Cotton farms in India are found in the ‘cotton belt’ that starts in the north-west, crosses through
the centre of the country and ends in the south-east. An estimated 4 million farmers cultivate 9
Mha of cotton. In the central and south, cotton farms are usually family-run, mixed, not irrigated,
and small (87% are less than 6 hectares). Larger farms are typically found in the north, where
the cotton is grown more intensively and is under irrigation (IIED, 2004). The national average
production is 429 kg per hectare, being the world’s lowest average. 50% of the fields are
fertilised with farm-yard manure, 30% with synthetic fertiliser and 20% are not fertilised. On
average, synthetic fertiliser applications are 85-100 kg N per hectare, 32-50 kg P per hectare,
and 0-50 kg K per hectare. 30% of the cotton area is irrigated. 50% of the cotton is cultivated on
black soil, 25% on loamy soil and 15% on red soil. Weeds and cotton stalks are generally burned
on the field or used as stove fuel.
Most commonly used pesticides include monocrotophos (Nuvacron,Ib), endosulfan (Thiodan, II),
quinalphos (Ekalux, II), chlorpyriphos (Dursban, II), dimethoate (Rogor, II), methyl-o-dematon, Ib),
fenvalerate (Fenvel, II), cypermethrin (Radar, Ib), Imidacloprid (Confidore, II), spinosad (Tracer, U).
Popular herbicides include Fluchloralin (Basalin, III), diuron, (U), and pendimethalin (Stomp, III).
Per season the fields are sprayed on average 8 times ranging from 0-40 times. No defoliants
are applied.
Bt-cotton was approved by Government of India for commercial cultivation in March 2002. Bt-
cotton was grown on 530800 hectares in 2004 (6% of the cotton-cropped area) (Manjunath,
2004). A nation-wide survey indicated that Bt-cotton growers in India obtained a 29% yield
- 22 -Science Shop Wageningen UR Report 223
The sustainability of cotton
increase due to the effective control of the Bollworm, a reduction in insecticide applications by
60% and an increase in net profit by 78% as compared to their non-Bt counterparts (Manjunath,
2004). The high investment costs of genetically modified cotton have however also resulted in
negative revenues (Qayum et al., 2004) (see also Chapter 2.3.5.).
Pakistan (10% of world production; 3.2 Mha)
The main cotton areas in Pakistan are located in the south of the country and 1.3 million farmers
grow cotton (Punjab: 2.6 Mha; Sidh: 0.6 Mha). The farms are mainly small-sized, averaging
2 hectares. Farms are family-run and quite extensive. In addition, there are 35 state farms
cultivating cotton with an average farm size of 25 hectares (Tayyab, 2004). Average cotton
production is 768 and 803 kg per hectare in Punjab and Sidh respectively. Hybrid varieties are
most widely used. GM-cotton is currently not licensed, as Pakistan is still developing its own GM
cultivars (NIBGE, 2004). Punjab receives 12 mm of rain per year and Sidh 21 mm. All fields are
irrigated, 70% with flood irrigation and 30% furrow. During the growing season the average daily
temperatures range between 25 and 43°C.
As to fertilisation, farm-yard manure is applied on 13% of the fields, and on the remaining 87%
synthetic fertilisers are used (ICAC, 2005). Typical fertiliser application levels are between 75-
150 kg N per hectare, 40-50 P per hectare, and 25-50 kg K per hectare. The major soil type is
silty loam (75%). Cotton stalks and weeds are most commonly burned on the fields or used as
stove fuel. Almost all cotton (more than 80%) is rotated annually with wheat.
The most commonly applied pesticides are imidacloprid (Confidor, II), thiamethoxam (Actara, n.l.),
malathion (Arrivo, III), deltamethrin (Decis, II), bifenthrin (Talstar, II), triazophos/deltamethrin (Deltaphos,
Ib/II), tracer (Spinosad, U), indoxacarb (Steward, n.l.), chlorpyriphos (Larsban, II), profenophos
(Curacran, II). The most common herbicides: are pendimethalin (III), trifluralin (U), metolachlor (Dual
gold, III), and pendimethalin (Stomp, III) (ICAC, 2005). No defoliants are applied (Banuri, 1998).
Cotton and textiles are vitally important for Pakistan’s economy generating over 60% of the
annual export income (Rao, 2005).
Brazil (5% of world production; 1 Mha)
In Brazil the cotton cultivation zone reaches from the south of the country to the northeast. In
the central area (Mato Grosso) in particular, there is a trend towards expanding the area under
soy-cotton rotation. This area currently has about 200 rain fed cotton farmers (averaging 2000
hectares of cotton per farm) and accounts for 50% of Brazil’s cotton production (IIED, 2004), with
yields up to 1290 kg per hectare (ICAC, 2005). The farms in the centre and the north of the country
are on average up to 10 hectares in size and produce much less: 370 kg per hectare. Virtually no
cotton is irrigated. The large scale farms in the west apply synthetic fertilisers averaging 180 kg
N per hectare, 120 kg P per hectare, and 220 kg K per hectare. Only half of the northern farms
apply synthetic fertilisers. Those that apply synthetic fertilisers do so in quantities of 30 kg N per
hectare, 60 kg P per hectare, and 40 kg K per hectare (ICAC, 2005). The major soils are Oxisols
(50%). Cotton stalks are usually left on the fields and weeds are burnt. The large western cotton
farms cultivate soybeans every 4th year, as well as millet rotations every second year. Common
pesticides include acetamiprid (Mospilan, n.l.), carbosulfan (Marshal, II), diafenthiuron (Polo, U),
endosulfan (Thiodan, II), lambda cyhalothrin (Karate, II), cypermethrin (Fury, Ib), deltamethrin
(Decis ultra, II), beta cyfulthrin (Bulldock, II), teflubenzuron (Nomolt, U), imidacloprid (Gaucho, II).
Most commonly used herbicides are trifloxissulfuro (n.l.), clomazone (III), and diuron (U). The main
defoliant applied is MSMA (III) (ICAC, 2005).
In the northern region rotation schemes are wider and more complicated, including corn, cowpeas
and fallow periods. The most widely used pesticides in this region are melathion (III), metyl parathion
(Ia), deltamethrin (II), and endosulfan (II). Usually five sprays per season are applied, and herbicides
are not used (ICAC, 2005). Through mechanisation and support to farms with large economies of
scale, Brazil is planning to expand its cotton growing area, especially in the central region.
Uzbekistan (4% of world production; 0.8 Mha]
The reliability of information obtained for this country may to some extent be questionable. For
example, the high yields per unit of area mentioned in Table 1 is doubtful.
97% of crop production is done on irrigated land. Since the 1980s, cotton production has
been reduced by half. Partial reform of the agricultural sector has resulted in a large increase
in the number of family and small farms. Small farms are expected to become the main
- 23 -Science Shop Wageningen UR Report 223
The sustainability of cotton
agricultural producers in Uzbekistan: in 2002 there were approximately 72,000 small farms,
2000 cooperatives, 3.3 million family farms, and 500 other private sector farms (IIED, 2004).
Defoliants are applied on 75% of the surface with mostly sodium chlorate (n.l.) being used. In
2000, cotton provided 12.9% of the GDP (Fortucci, 2000).
Pesticide use has been reported as high as 54.6 kg per hectare in some cotton growing areas
of Uzbekistan (Libert, 1995). However, more and recent information is not available. Water use
for cotton production has had and is still having a significant impact on the Aral Sea area (see
Chapter 2.3.1, Environmental impact).
West Africa (4% of world production; 2.4 Mha)
In Africa, the western part is the main cotton cropping region, notably Benin, Cameroon and
Mali with around 2 million cotton farmers (IIED, 2004). Benin cultivates 384,000 hectares with
an average production of 504 kg per hectare (ranging from 200 to 1050 kg per hectare),
Cameroon 215,000 hectares with an average production of 578 kg per hectare, and Mali
cultivates 540,000 hectares, producing 435 kg per hectare.
Benin receives on average 1000 mm rainfall and has a temperature in the cotton season of about
30°C. Cameroon receives 1050 mm of rainfall and Mali 784 mm. This means that irrigation is not
a prerequisite for cotton cultivation in the region. In all three countries hybrids are used.
In Benin farms tend to be very small: about 1 hectare. Still, 30% of the farms range from 5 to
10 hectares. In Cameroon all farms are smaller than 3 hectares and they are mainly situated on
sandy loam soils. In Mali the average farm size is 5.3 hectares, and the cultivation takes mainly
place on clay soils and sandy clay grits. In all of these countries crop rotations are done; in Benin
with maize or cowpea, in Cameroon with cereal crops, and in Mali with corn, millet, or sorghum.
Usually cotton cultivation is alternated with one other crop. Mainly manual and animal force is
used for soil preparation. Benin and Cameroon rely for the most part on artificial manure (NPK,
74-46-28 and 42-27-24 kg per hectare respectively). In Mali, about 40% of the farms apply farm-
yard manure as their single source of nutrients and the remaining 60% apply NPK (44-33-18
kg per hectare). In Benin, the most frequently used pesticides are endosulfan (II), cyfluthrine-
chlorpyriphos (II/II), and lambdacyhalothrine (II); in Cameroon cypermethrin (Ib), endosulfan (II) and
cetamipride (n.l.); and in Mali profenofos (Calfos 500, Cuacron 500, Tenor 500, II), endosulfan
(Phaser 500, Phaser ultra caps, Thiofanex 500, II), cypermethrin (Cytofos 286, Cyperfos 336,
Conquest C88, Ib). In all three countries the herbicides glyphosate (Roundup, U), diuron (U) and
paraquat (II) are applied. In general defoliants are not used. In all three countries the crop is
harvested by hand and cotton stalks are burnt on the fields or used as household fuels.
Cotton contributed 8.8% to the GDP of Benin (Fortucci, 2000) and 1.3% and 5% respectively to
the GDP of Cameroon and Mali (Fortucci, 2000).
Turkey (3% of world production; 0.7 MHa)
The major areas of cotton cultivation are West (20%), South-West (60%) and South (20%) Turkey.
Most of the 300,000 cotton farms are small scale and family-run averaging 5 hectares. Average
production levels of 1350, 1600 and 1289 kg per hectare are reached respectively in West,
South-west and South Turkey. With South Turkey having very low annual rainfall (189 mm), it
is very much dependent on irrigation. West and South-west Turkey receive 600 - 1100 mm of
rainfall annually and are consequently less dependent on irrigation. Irrigation is done via furrows
(78%), flooding (9%) and sprinklers (5%). Only 8% of the area is not irrigated. In West and South-
west Turkey cotton is cropped mainly on Entisola fluvent soils (60%) and Vertisols (40%), and
in South Turkey on alluvial soils (70%) and vertisols (30%). Cotton stalks and weeds are usually
burnt after the harvest. For the most part, locally developed hybrid varieties are used. The share
of GM cotton is not known. Generally 1-year rotations are done with cereal crops (wheat, maize,
barley) and sometimes a fallow year is included. Less commonly, multiple-year rotations with
wheat and maize are done. The soil is prepared using machinery.
The most commonly applied pesticides in Turkey are diafenthiuron (Polo 50, II), carbosulfan
(Marshall 25 EC, II), lambda cyhalothrin (Karate, II), chlorfluazuron (Atabron, U), thiodicarb (Larvin,
II), nissorin, (n.l.), acetamiprid (Mosphilan, n.l.), Endosulfan (Thiodon, II), and imidacloprid (Gaucho,
II). Popular herbicides are linuron (Match, U), trifluralin, (U), fluazifob butyl, (n.l.), and prometryne
(U) (ICAC, 2005). In general no defoliants are used.
The share of cotton in the GDP is estimated to be approximately 4% (Gazanfer, 2004).
- 24 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Australia (1% of world production; 0.2 Mha)
In Australia, 1100 farmers cultivate 198,000 hectares of cotton, with an average yield of 1760
kg per hectare (ranging from 300-3000 kg per hectare, depending mostly on water availability).
The average farm size is 800 hectares, with 18% being more than 1000 hectares. In general the
cotton fields receive 500 mm per year rainfall and the average temperature is 25°C. Mainly GM
cotton is cultivated – as of 2005 GM is expected to be 80% of all production (Chandler, 2005).
Most commonly cotton is planted on the same plots every year, with a fallow period of a few
months in between. All farming activities are mechanised. Nutrients are added via synthetic
fertilisers (NPK, 200, 80, 75 kg per hectare). Most commonly applied pesticides are amitraz
(Ovasyn, III), indoxacarb (Steward, n.l.), emamectin (Avermectin, n.l.), endosulfan (Thiodan, II),
dimethoate (Rogor, II), spinosad (Tracer, U), fipronil (Regent, II), chlorpyrifos (Predator, II), and
deltamethrin (II). Commonly used herbicides are glyphosate (Roundup, U), fluometuron (U), and
pendamethalin, III (ICAC, 2005). On all fields defoliants are applied with thidizuron (n.l.) being by
far the most widely applied.
Other regions
Egyptian cotton is famous for its long staple length combined with strong fibre that produces
very fine cloth. In total Egypt cultivates 303,000 hectares, which are all flood irrigated.
Israel has the highest average production per hectare, with 1850 kg per hectare (ranging from
1500-3000 kg per hectare), with drip irrigation applied on all parcels. The 14,000 hectares that
are cultivated are managed by 130 farmers.
Peru cultivates an interesting naturally coloured variety of cotton, with colours ranging from
dark brown/black to green. The coloured cotton is limited to 300 hectares (ICAC, 2005) and is
cultivated in Peru’s dry zone (with 17 mm of rain per year).
- 25 -Science Shop Wageningen UR Report 223
The sustainability of cotton
(Kooistra. 2004)
- 26 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table 6. List of most commonly used pesticides in cotton (ICAC, 2005; PAN, 2005).
Active ingredient WHO
class1 Type2
China
India
USA
Pakistan
Brazil
Benin
Cameroon
Mali
Turkey
Australia
Toxicity class I
parathion Ia ix
cypermethrin Ib ix x x x
methyl-o-dematon Ib ix
methamidophos Ib ix
monocrotophos Ib i x xx
thiofanex Ib ix
triazophos/ Ib ix
Toxicity class II
bifenthrin II i x
carbosulfan II i x x
chlorpyriphos II ix x x x
cyfluthrin II i xxx
deltamethrin II ixxx
dimethoate II ix x
endosulfan II ix x x xx x x
fenvalerate II ix
fipronil II ix
imidacloprid II i x xxxx
lambdacyhalothrine II ixxx
organophosphates in general II ix
paraquat dichloride II h
profenophos II ix x
pyretroids in general II x
quinalphos II ix
thiodicarb II ix
trichlorfon II i x
Toxicity class III
alachlor III h x
amitraz III ix
clomazone III hx
dicofol III i x
dimethipin III d
fluchloralin III hx
malathion III ix
- 27 -Science Shop Wageningen UR Report 223
The sustainability of cotton
metolachlor III hx
MSMA III h/d xx
organophosphates in general III Ix
pendimethalin III hx x x
pyretroids in general III ix
sodium chlorate III d
Toxicity class U
chlorfluazuron Ui x
diafenthiuron Ui x
diuron U h/d x x
ferbam U ix
fluometuron U h xx
glyphosate, isopropylamine salt U h xx
linuron Uhx
piperonyl butoxide U i x
prometryne U h xx
spinosad Ui x x x
teflubenzuron Ui x
trifluralin U hx x
Toxicity class O
chlordimeform O i x
EDTA, tetrasodium salt O ix
thidizuron O dx
profluralin O hx
Not listed by WHO
acetamiprid n.l. i x x x
emamectin, benzoate n.l. ix x
fluazifop-butyl n.l. hx
Heliothis armigera n.l. o.i. x
indoxacarb n.l. ix x
profluralin n.l. hx
sodium carbonate n.l. dx
thiamethoxam n.l. ix
trifloxysulfuron-sodium n.l. hx
1 WHO toxicity classes. Ia = extremely hazardous, IB = highly hazardous, II = moderately hazardous, III = slightly hazard-
ous, U = unlikely to present acute hazard in normal use, FM = fumigant, not classified, O = obsolete as pesticide,
not listed, n.l.= not listed by WHO.
2 i = Insecticide, h = herbicide, d = defoliant.
- 28 -Science Shop Wageningen UR Report 223
The sustainability of cotton
3.2 Organic systems
The European Community Council Regulation EC 2092/91 definition of organic farming states:
‘Organic production systems are designed to produce optimum quantities of products of high
quality by using management practices which aim to avoid the use of agro-chemical inputs and
which minimise damage to the environment and wildlife’ (EEC, 1993).
The International Federation for Organic Agricultural Movements (IFOAM), established in the early
1970s, represents over 600 members and associate institutions in over 100 countries. IFOAM
(1996) defines the “organic” as referring to the particular farming system described in its Basic
Standards. The “Principle Aims of Organic Agriculture and Processing” are based on the following
equally important principles and ideas (Soil Association, 2005):
• to produce food of high nutritional quality in sufficient quantity;
• to interact in a constructive and life enhancing way with all natural systems and cycles;
• to encourage and enhance biological cycles within the farming system, involving micro organ-
isms, soil flora and fauna, plants and animals;
• to maintain and increase long-term fertility of soils;
• to promote the healthy use and proper care of water, water resources and all life therein;
• to help in the conservation of soil and water;
• to use, as far as is possible, renewable resources in locally organised agricultural systems;
• to work, as far as possible, within a closed system with regard to organic matter and nutrient
elements;
• to work, as far as possible, with materials and substances which can be reused or recycled,
either on the farm or elsewhere;
• to give all livestock conditions of life which allow them to perform the basic aspects of their
innate behaviour;
• to minimise all forms of pollution that may result from agricultural practices;
• to maintain the genetic diversity of the agricultural system and its surroundings, including the
protection of plant and wildlife habitats;
• to allow everyone involved in organic production and processing a quality of life conforming to
the UN Human Rights Charter, to cover their basic needs and obtain an adequate return and
satisfaction from their work, including a safe working environment;
• to consider the wider social and ecological impact of the farming system;
• to produce non-food products from renewable resources, which are fully biodegradable;
• to encourage organic agriculture associations to function along democratic lines and the
principle of division of powers;
• to progress towards an entire organic production chain, which is both socially just and ecologi-
cally responsible.
To be able to market products as ‘organic’ the farm must be certified by an organic certification
body. Known organic cotton certifiers are the Swiss-based Bioswiss and IMO control, the UK-
based Soil Association, and the Netherlands-based SKAL (Ministry of Agriculture, 2005),
amongst others. Regulations in place that set the parameters for the organic production of
cotton areRegulation (EEC) 2092/91 (EEC, 1993) or USDA National Organic Program, Standards
and Guidelines (USDA NOP, 2005). These regulations outline a specific set of rules in place for
organic cultivation, describing what must be certified, relevant exemptions and exclusions from
certification, the use of the term “organic”, and allowed and prohibited substances, methods and
ingredients in organic production and processing.
Organic cultivation excludes the use of synthetic pesticides, synthetic fertilisers and genetically
modified organisms. The organic certificate covers agricultural land only. Neither the origin of
irrigation water nor the origin of agricultural soils (e.g. reclaimed land, natural forest etc.) are
included. Recently, an organic crop cotton guide was published (Eyhorn et al., 2005).
Today certified organic cotton is grown in 10-15 countries. Production is concentrated in Turkey
(1750 tonnes; 29% of total organically produced cotton, 2005) and the USA (1625 tonnes; 27%,
2005). Other important sites are found in India (1000 tonnes; 17%), Peru (550 tonnes; 9%),
Uganda (275 tonnes; 5%), and in Egypt, Senegal, and Tanzania (200 tonnes each; 3%) (Table 7).
- 29 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Table 7. Certified and traded organic cotton production worldwide in tonnes of fibre (Ton, 2003).1
Country 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01
Argentina 75 - - - - -
Australia 400 300 300 - - -
Brazil 1 1 1 5 10 20
Benin - 1 5 20 20 30
Egypt 650 625 500 350 200 200
Greece 150 125 100 75 50 50
India 925 850 1000 825 1150 1000
Israel 50 50 20 - - -
Kenya - - 5 5 5 -
Mozambique 100 75 50 - - -
Nicaragua 20 20 20 - - -
Paraguay 50 50 50 - - -
Peru 900 600 650 650 500 550
Senegal 1 10 10 50 125 200
Tanzania 10 100 100 100 200 250
Turkey 725 850 1,000 1,200 2,000 1,750
Uganda 75 300 450 250 200 275
USA 3350 1550 1300 1900 2900 1625
Zimbabwe - - 1 5 5 -
Total 7482 5507 5562 5435 7365 5950
Change222 -16 1 -2 36 -29
1 Recently organic production has sharply increased due to increased demand, but reliable figures were not yet
available.
2 In %, compared to previous year.
Organic cotton farms cover a wide range of farm sizes and their organisation depends on the
country where they are located. A rough division can be made between the two types of farms:
large–scale, mechanised farms and small-scale, animal traction driven farms. The first type is
mostly found in the US and the second type is mostly found in Turkey, India and Peru.
Turkey (29% of world’s organic cotton, 8000 hectares)
Presently there are around 350 organic cotton growers in Turkey with in total 7883 hectares
under organic cotton cultivation (OE, 2005). The average farm size is 20 hectares, with a wide
range from 2-1000 hectares. The average production per hectare is not known precisely, but is
estimated to be 350 kg per hectare (OE, 2005). These data are, however, not certain. Ferrigno
(personal communication 2005) estimates an average yield of 1500 kg per hectare. 40% of the
farms are irrigated. More detailed information about organic cropping systems in Turkey is not
yet available.
United States (27% of world’s organic cotton, 3200 hectares)
Organic cotton farms in the US are situated in the cotton belt of the country and make up about
3200 hectares in total. In 1995 the area cropped to organic cotton was still 36,000 hectares,
but due to low prices and unfavourable weather conditions many farmers have concluded organic
cotton cultivation and returned to conventional. Cultivation was limited to Arizona, California,
Missouri, New Mexico and Texas (OTA, 2004). Organic cotton was cultivated by 14 farmers and
the average organic farm size was 500 hectares producing in total about 1200 kg cotton per
- 30 -Science Shop Wageningen UR Report 223
The sustainability of cotton
hectare. Data about irrigation practises or fertiliser use are not available, but it is likely that US
organic farms are irrigated in a similar way to conventional or IPM farms in the US. All use tractors
for field labour and cotton is harvested with machines. Currently no information is available about
rotations and use of bio-control methods including pesticides of natural origin.
The organic farms are assisted by research institutes (California University) with site-specific
information. All farms meet the requirements of the National Organic Program (USDA NOP,
2005).
India (18% of world’s organic cotton, 8000 hectares)
In this country 4000 farmers cultivate organic cotton on 7500 hectares. Most of the fibre is
grown in three major organic projects named Agrocel (Mandvi, Gujrat), Maikaal bioRe Ltd. (Central
India), and Pratibha Syntex Limited (Vasudha) (North India). In general the organic cotton farmers
belong to the middle and poor classes. Farm size is less than four hectares. Literacy levels
are low, but are slowly on the rise. The level of knowledge concerning organic cultivation is still
low. Rotations and strip cropping with millet, peppers, maize and cowpea are applied (personal
communication Lanting, 2005 and personal communication Eyhorn, 2005). Organic manure is
applied in quantities of 400-2000 kg per hectare and more than 20% of the organic cotton area
is irrigated. All cotton is handpicked (Organic exchange, 2005). More information about organic
cropping systems in India is not available.
Peru (8% of world’s organic cotton, 300 hectares)
In 2004, 255 farmers were contracted to grow organic cotton over an area of 280 hectares.
Peruvian cotton farmers have long-term experience with cotton production and some indigenous
and traditional peasant farming technologies are still in use. Most organic cotton is grown in the
Cañete valley, located on the desert coast 150 km south of Lima on private farms. Due to the
lack of rainfall agriculture depends completely on (flood) irrigation coming from the Cañete River
originating in the Andes Mountains. Cotton is often intercropped with maize and rotated with a
limited number of winter crops like green beans or fodder maize. The only information about
pest control is that to a limited degree pheromones are applied to control the Pink Bollworm and
cotton plants are burned after picking to avoid pest transmission (Organic exchange, 2005).
More information about organic cropping systems in Peru is not available.
Mali
In Mali, Helvetas (a Swiss NGO) started an organic cotton project in 2002; it is located in the
Province of Bougouni, in the counties of Kolondieba and Yanfolila. The project started with 174
farmers, and has rapidly increased to 385 farmers. In 2003/2004 170 hectares were planted
with organic cotton, averaging a yield of 475 kg of seed cotton per hectare. The average farm size
is 5 hectares, with less than 10% cropped to cotton. The project plan foresees production of 50
tonnes of fibre per hectare in 2004/2005. Crops included in the rotation are groundnut, sesame
and millet with shea butter trees in the fields. Some extension and micro-credit schemes for farm
equipment are available. It is stated that there is a huge potential for further expansion.
Benin
In Benin, two organic cotton projects started in 1996/1997, one in Northern Benin and one in
Central Benin. The Central Benin project (the OBEPAB; a Beninese NGO) works with 800 farmers
(data 2003). The cotton is grown on small farms with an average of 2-8 hectares each. The project
works with a guaranteed purchase price, which is set in advance of the season. Yields range from
270 - 562 kg per hectare. (OBEPAB a and b, 2002, in Ferrigno et al., 2005) (OE, 2005)
Senegal
The main production region is in the Central-East (project Koussanar). In 2005 there were
approximately 1750 producers with an average farm size of 5 hectares in more than 50 villages.
Another project in the Southeast project (Velingara) has not produced organic cotton due to
management problems since 2001 (OE, 2005).
Egypt
In total 80 hectares was cultivated with organic cotton in the 2003/2004 season by Sekem,
the largest grower of organic cotton in Egypt. Sekem manages an integrated supply chain from
- 31 -Science Shop Wageningen UR Report 223
The sustainability of cotton
the farm to the finished product. Besides this the company supports local community economic
development and health and education projects in areas where they source their products (OE,
2005).
China
In China, organic cotton is only planted in Xinjiang due to the local low disease and insect
pressure. Of the total of 1.0 Mha of cotton grown in Xinjiang, only 0.12% or 1300 hectares
(2004) are grown organically and those farmers are organised into three producer groups: Of
The Earth (OTE), Esquel and a third project whose name is unknown.
OTE’s farms are located near Urumqi City. The Xingjian region has very good natural conditions
for the development of organic cotton, and the local government can play an important role
in promoting organic production by extension and the promotion of international trade (UNEP,
2002). Farms are about 45 hectares. Instead of synthetic fertiliser, the following sources of
fertiliser are used: cotton seed residues, green manure crops, cotton stalks, leaves, and organic
wastes. Methods that have been implemented to prevent pests and weeds include crop rotation,
winter irrigation, planting of corn around the cotton field to distract the pest, manually pest and
worm entrapment on the field, aspen traps to block and kill pests, and black light lamp traps
(OE, 2003).
3.3 IPM systems
Integrated Pest Management (IPM) is characterised by the fact that it tries to keep a balance
between pests and their natural enemies and to reduce the use of pesticides to a minimum
level. It is defined by the Food and Agriculture Organisation as: ‘A site-specific strategy for
managing insect, weed, disease and other pests in the most cost effective, environmentally
sound and socially acceptable way’. IPM is not a rigidly defined form of crop protection, but a
dynamic system that adapts and makes sensible use of local resources and the latest research,
technology, advice and experience’ (FAO, 2004). The USA-based IPM-program of UC Davis/
California (Vargas, 2005) defines IPM an ecosystem-based strategy that ‘focuses on the long-
term prevention of pests or their damage through a combination of techniques such as biological
control, habitat manipulation, modification of cultural practices, and use of resistant varieties’.
Pesticides are used only after monitoring indicates they are needed according to established
guidelines, and treatments are made with the goal of removing only the target organism. Pest
control materials are selected and applied in a manner that minimises risks to human health,
beneficial and non-target organisms, and the environment (Vargas, 2005). The Australian AAS
(1999). IPM for cotton is defined as ‘a system that integrates all available, practical means
of managing pest populations with the aim of reducing insecticide use while maintaining or
improving profitability, yield, fibre quality and crop maturity.’ In practice this means that (cited
from Bos et al., 2005):
• control measures are only considered when pests occur (no preventive pesticide spraying);
• the presence of pests does not automatically require (chemical) control because the infesta-
tion may be insignificant;
• when control measures are necessary, non-chemical control should be considered first (such
as mechanical weeding);
• suitable control measures should be used in an integrated manner;
• pesticides should be used appropriately (no overdosing, avoid sensitive areas, wear protec-
tive clothing, etc.;
• pesticides and formulations used should be easily degradable and less toxic to man and the
environment (aquatic organisms, key soil organisms, pollinators);
• pesticide use should be selective in order to minimise mortality of natural enemies of pests
such as predators and parasites.
The wider term Integrated Crop Management (ICM) includes the reduction of synthetic fertilisers
and energy to minimum levels. IPM and ICM are in practice often interchangeable terms, and
here we use the term IPM when referring both systems. It is estimated that globally 20% of all
cotton is grown on IPM-managed farms. Main contributors include the FAO-project ‘IPM on cotton
in Asia’, including IPM projects in India, Vietnam, China, The Philippines, Pakistan and Bangladesh
(FAO, 2004) and the IPM program of the United States Department of Agriculture (USDA) in
- 32 -Science Shop Wageningen UR Report 223
The sustainability of cotton
collaboration with state universities like UC Davis (2000) (Curtis, 1998). In the FAO-project,
90,000 small cotton producers have been trained in integrated pest management in China, India,
Pakistan, Bangladesh, the Philippines, and Vietnam using the Farmer Field School technology
(FAO, 2004).
Farmer access to IPM information and training relates to issues of who has the power to design
a course or to select trainers and participants, and who holds the expert knowledge. Farmer
education is also related to social structures, power relations and processes like the privatisation
of farm services and literacy (Fliert and Proost, 1999).
In the USA, IPM is a widely accepted practise, applied on approximately 70% of the total cotton
area. It should be stressed that the coverage however does not necessarily reflect the adoption
of all IPM techniques (Zehnder, 2004). In China, pilot projects were financed by the Asian
Development Bank and demonstrated that cotton farmers had reduced their use of pesticides
and increased yields simultaneously (FAO-ADB, 1999). In Syria, the introduction of IPM has
reduced pesticide application six-fold (FAO, 2004). In Egypt, IPM is being introduced slowly, but
precise data are lacking (FAO, 2004). The UN (2000) estimates a 90% reduction in pesticide
use in Egypt in the late 1990s due to the introduction of IPM (UN, 2000).
3.4 Other promising techniques
Conservation tillage is currently a method with potential to minimise soil erosion. Ploughing
is usually reduced or even eliminated. Additional benefits are a reduction in weed populations
and reduced tillage costs. According to a report by the Cotton foundation in the USA, 78%
of US cotton farmers who have adopted conservation tillage practices since 1997 have done
so specifically because herbicide-resistant cotton varieties had made it more feasible (Dinham,
2004).
3.5 Summary
The above paragraphs illustrate that there are important differences and similarities between
conventional, IPM and organic systems. Two activities having major influence on the environment,
irrigation water use and land clearing are not fundamentally different for the three farming systems
(Table 8). In addition, production per unit of area very strongly differ for the different countries.
Table 8. Characteristics of organic, IPM, and conventional cotton farming systems.
Organic IPM Conventional
synthetic/organic fertiliser use org1syn/org1syn/org
synthetic/natural pesticide use nat2syn/nat2syn/nat
irrigation water use yes yes yes
average yields rel. low4rel. high variable
monocultivation /mixed cropping systems mo/mi5mo/mi mo/mi
continuous cultivation no6no/yes no/yes
land clearing allowed yes yes yes
burning organic material (weeds, plants, etc.) no yes yes
mechanised labour yes yes Yes
share of world production (%) 0.04 20.0 80
1 org = organic fertiliser; syn/org = synthetic and/or organic fertiliser.
2 nat. = natural pesticides allowed; only few natural pesticides are on the market and their use in cotton systems is very
likely far less than the use of synthetic pesticides in IPM and conventional systems.
3 syn/nat = synthetic and natural pesticides allowed; mainly synthetic pesticides are used.
4 rel. = relatively.
5 mo/mi = monoculture and/or mixed cropping systems. Mixed cropping systems are usually limited to the developing
countries, more specifically to subsistence farming.
6 continuous cultivation of cotton is not prohibited in organic systems, but most usually this is not done.
- 33 -Science Shop Wageningen UR Report 223
The sustainability of cotton
4 Environmental impact indicators
Naturally each agricultural activity has a certain impact on the environment. This impact is
generally considered unsustainable if the environment is damaged to an extent that future
generations cannot cultivate the same crop. The following sustainability problems have been
reported in relation to cotton cultivation (IEEP, 2005):
• degraded land as a result of salinisation and erosion;
• water depletion by excessive use of soil and surface water;
• natural habitat conversion due to cutting of forests and dam constructions;
• eutrophication of surface water;
• wildlife contamination by pesticides (insects, fish, mammals, birds);
• human health due to direct pesticide intake primarily by farm workers but also by regional
inhabitants through contamination of drinking water and food contamination.
The Department for Environment, Food and Rural Affairs (EU) has recently predicted two major
environmental problems related to cotton cultivation (IEEP, 2005):
• intensification of pesticide use. Countries where this is expected to be the most severe are
Australia, Brazil, Uzbekistan, Turkey, Pakistan, Mexico, India, China, and West and Central Af-
rica (Benin, Burkina Faso, Chad and Mali), thus including the major cotton producing countries.
The reason for intensification of pesticide use according to the IEEP (2005) is an increase in
economic development leading to more opportunities to buy pesticides. If this does not goes
hand-in-hand with the introduction of IPM-related techniques, it will lead to unwanted side-ef-
fects of pesticides on human health. Currently it is estimated that 15% of cotton farmers are
not using any pesticides due to an inability to buy them (ICAC, 2005);
• increase in irrigation problems, leading to salinisation and unsustainable use of fresh water
resources.
To be able to quantify the problems as mentioned above, the environmental impact can be
divided into indicators, each representing an aspect of environmental impact. The relationship
between agricultural practises and recognised environmental problems can be found in Table 9.
The ‘indicators’ are derived from Life Cycle Assessment (LCA) methodology. The selection was
done based on 17 different studies on the environmental impacts of cotton cultivation (Annex
1). In the following paragraphs the indicators and the differences between the three farming
systems will be discussed.
Table 9. Recognised problems in cotton cultivation related to various agricultural practises and in
relation to environmental impact indicators. Environmental impact indicators: 1: water use,
2: human toxicity, 3: environmental toxicity, 4: global warming, 5: eutrophication,
6: acidification, 7: erosion, 8: land-use, 9: biodiversity, 10: salinisation, 11: energy use.
Organic
fertilisers Synthetic
fertilisers Pesticides Land
clearing Irrigation
recognized problems
in cotton cultivation produc-
tion use produc-
tion use dam con-
struction water with-
drawal
degraded land 1 3 7 10
water depletion 1 1
human health effects 11 4, 11 11 2 11
natural habitat
conversion 6 6 4, 5 3, 6 3 9 8 1
wildlife contamination
by pesticides 3 1
eutrophication of
surface water 4 5 5 7 1 1
- 34 -Science Shop Wageningen UR Report 223
The sustainability of cotton
4.1 Water use
As seen in Chapter 2.3.1, cotton consumes large amounts of fresh water. According to the World
Resources Report 2005 (UNEP, 2005), ‘The world’s thirst for water is likely to become one of the
most pressing resource issues of the 21st century’. Water use (expressed in litres per hectare
of cotton or litre per kilogram of cotton yield) as an indicator for estimating the sustainability of
cotton production has been recommended repeatedly (e.g. Tobler & Schaerer, 2001; Banuri,
1998; IERE, 2005; Hanssen, 2001; Clay, 2004) but currently it has not been applied in Life Cycle
Analysis (LCA) studies on cotton cultivation, due to a lack of proper information. Bos et al. (2005)
developed water use indicators for cotton cultivation, mainly based on water use efficiency. To
date, methodologies have not taken sustainability of the source of irrigation water into account
(groundwater versus surface water). However, without extensive field measurements, both on
the farm and at the water source, water use is not easy to estimate at a farming systems level.
The potential water use indicator does not indicate significant differences between the three
farming systems, as irrigation water application is allowed within all three farming systems, and
mainly defined by local circumstances rather than by management characteristics. It is however
important to note that irrigation water becomes relevant if conversion from one system to another
would lead to a reduction in production, thus leading to more surface requirements for the same
amount of product. Organic cotton is often regarded as a low yielding management system.
The actual water use indicator however, as shown in Chapter 3.2, does show a clear difference
between farming systems as the main organic cotton producing areas (around 60%) are rain-fed.
Expanding the organic cotton area in rain-fed regions by converting low-input conventional farms
would therefore be unlikely to lead to increases in water consumption.
4.2 Human and environmental toxicity
The main impact of pesticides is on the environment, most notably on human beings. The most
widely used pesticide risk indicator is the Environmental Index Quotient (EIQ), developed by
Kovach et al. (1992) and used by the FAO as a monitoring tool for their IPM programs. The EIQ
takes into account fish, bird, and bee toxicity, as well as toxicity to beneficial arthropods. Other
pesticide risk indicators include PRIMET, which combines pesticide exposure assessment, effect
assessment and risk assessment (Brink in Bos et al., 2005).
In Chapter 3.1 it was shown that in conventional cotton a large range of the most toxic pesticides
are used. Since IPM aims to reduce pesticide use and to use necessary pesticides more safely,
and organic abandons pesticide use altogether, this indicator clearly distinguishes the three
management systems. Attention should also be paid to the use of natural pesticides in organic
farming systems. Although currently there are no indications that the natural pesticides allowed
in organic systems are widely used, it should be kept in mind that they are not necessarily less
hazardous than synthetically produced pesticides. Pesticides were used as an LCA indicator by
Clay (2004), Banuri (1998), and Slater (2003). A target value can not be set for this indicator
because generally lower application levels are better (personal communication Kovach, 2005).
Indirect costs (see Chapter 2.3.4) are currently not applied for any of the indicators. The way a
pesticide is applied is crucial to the toxicity indicator (Chapter 2.3.4).
4.3 Global warming
The so-called ‘greenhouse effect’, the rise in the earth’s temperature due to human activity, has
become an important topic of debate and research over the past several years. Emissions of
CH4 and N2O are also considered to be important contributors to climate change.
Cotton cultivation contributes to global warming primarily through energy use (including the costs
of fertiliser transport and production of synthetic fertiliser-N), the burning of organic materials
and to some extent the CH4 produced from the use of animal labour.
The direct effects of global warming on cotton production are two-sided. On the one hand,
increased concentrations of CO2 may lead to increases in cotton yields and global warming may
lengthen the growing season (Perkins, 2002). At the same time, global warming may lead to
desertification e.g. in Central Asia.
The Kyoto Protocol provides for carbon sequestration in terrestrial sinks and carbon trading
through ‘Clean Development Mechanisms’. Terrestrial carbon sequestration can be realised
through increasing soil organic matter content (Lal, 2004) by applying materials that have a
- 35 -Science Shop Wageningen UR Report 223
The sustainability of cotton
relatively large persistent carbon pool such as compost, and by cropping practices such as
choice of rotation, tillage and cover cropping (Sainju et al., 2005).
In agricultural LCA studies, contribution to global warming is an important indicator (Tobler &
Schaerer, 2001; Holme, 1993; Dettenkofer et al., 2001; Leffland, 1997; Weidema & Meeusen,
1992; Banuri, 1998; Dahllöf, 2004; Olson et al., 2001; Fryendal, 2001; Kooistra et al., 2005).
Global warming is expressed in CO2 equivalents. For example, the greenhouse gases CH4 and
N2O contribute 21 and 243 times that of CO2 (IPCC, 1999).
Factors influencing energy use are degree of mechanisation, irrigation techniques, amount of
fertilisers used, and cultivation methods. According to Canakci et al. (2004), most energy needed
for conventional cotton cultivation in Turkey is required for seedbed preparation and irrigation
with shares of 14–65% and 26–40% respectively. Without irrigation the most significant energy
inputs are fertiliser and diesel, contributing 40–54% and 17–43% respectively (Canakci et al.,
2004; Gül et al., 2004).
In relation to the potential impact of Global Warming the degree of fertilisation with synthetic
fertilisers, mechanisation and soil organic matter content are the most important factors.
Therefore, systems with low or no synthetic fertiliser use, low or no mechanisation and high
organic matter content are favoured, which refers to low input conventional, IPM and organic
farms. Relative to fertiliser use, pesticide production and use contribute very little to the global
warming indicator.
4.4 Eutrophication
Eutrophication causes algal bloom and subsequent oxygen shortages in surface water and
affects natural vegetations. Phipps & Crosson (1986) and Nielsen & Lee (1987) estimated that
50-70% of all nutrients reaching surface waters, principally N and P, originate from fertilisers.
Nitrates move with water and leach easily to surface and groundwater. P binds to sediments
and contributes to eutrophication if soil is eroded (Smith et al., 1987). Eutrophication can have
a devastating effect on aquatic ecosystems leading to, for example, algal growth, depletion of
oxygen and a decline in aquatic plants and animals (Uri, 1998). This indicator is usually subdivided
into terrestrial and aquatic eutrophication. To estimate the Eutrophication Potential, the emissions
of NO3, P, PO4, NH3 and NO2 are taken into account. The emitted quantities are derived from the
amounts of mineral fertiliser, diesel, farm yard manure and burnt organic material used as well as
the estimated quantity of leaching. This indicator is considered to be relevant by multiple authors
(Tobler & Schaerer, 2001; Clay, 2004; Kalliala, 1999; Dettenkofer et al., 2001; Leffland, 1997;
Dahllöf, 2004; Fryendal, 2001; Home, 1993; Olson et al., 2001). Since limited research has
been published on this subject it is not possible to give a clear conclusion as to the differences
between the three systems based on this indicator.
4.5 Acidification
The effects of acidification mainly depend on the magnitude of deposition (wet or dry) and the
sensitivity of soil and water. Main effects are found in soil (affecting soil communities), forests
(vegetation damage) and surface and groundwater (shifts in aquatic species). Some health risks
can be involved since acidification can lead to increased mobility of metals such as aluminium,
copper and lead. This indicator represents the effects of cotton production on the acidification
of natural ecosystems (Audsley et al., 1997). For an estimate of the acidification potential of SO2
emissions, NH3 and NO2 are taken into account, and the emitted quantities are derived from the
use of mineral fertiliser, diesel, pesticides, farm yard manure and the burning of organic material.
Several studies consider acidification to be important (Tobler & Schaerer, 2001; Dettenkofer,
2000; Leffland 1997; Weidema & Meeusen, 1999; Banuri, 1998; Dahllöff 2004; Olson et al.,
2001; Kooistra et al., 2005; Holme, 1993; Fryendal, 2001). As with eutrophication, almost no
research is published on the indicator and it is therefore not possible to conclusively distinguish
the three systems.
4.6 Erosion
The most important factors determining erosion are the percentage of soil covered with a crop,
the angle of the field surface, type of irrigation and rainfall pattern, soil type, and measures taken
- 36 -Science Shop Wageningen UR Report 223
The sustainability of cotton
to prevent erosion. Since erosion is caused by factors that are not necessarily restricted to one
farm management type, it is not possible to pin erosion problems on one type of management.
However, conventional methods do not usually include erosion prevention, which can lead to
major problems. Since organic management aims to maintain a high level of organic matter in the
soil, cover crops and green manures are grown and erosion prevention measures are prescribed.
Thus, organic farms will likely have less erosion compared to IPM and conventional farms within
the same area (USDA-NOP, 2004). A common method for estimating soil erosion is done using
the Revised Universal Soil Loss Equation (RUSLE, Nyakatawa, 2001), which expresses the
amount of soil loss per surface unit.
Although the erosion may, as already stated, be an indicator that separates farms doing
comprehensive soil organic matter management from those that do not, it seems impractical
to include this indicator for estimating the sustainability of cotton production. Potential erosion
depends to a large extent on topographical factors, and to simply state that cotton must be
produced only on flat lands, for example, would be highly unrealistic. The measures a farmer
takes to reduce erosion depend on the local situation, making a clear-cut erosion indicator for
sustainable cotton production difficult. Clearly more work is needed on this important topic.
4.7 Land use
For agricultural production land is always used. The current distribution of crops over the
earth’s surface is mainly determined by climate, soil type, labour availability, price of product,
distance to consumers, characteristics of varieties (quality, quantity, economic return), cultural
aspects and the availability of planting material. Since land use for a cash crop like cotton
competes with food and other cash crops, land use must be calculated in order to be able to
compare the different crops. Also inputs into the system such as synthetic fertilisers (open
mines, Chapter 2.3.2) and irrigation water (dams, watersheds, Chapter 2.3.1) have an impact
on natural areas. Important aspects influencing land use are the productivity (yield) per hectare,
and synthetic fertiliser and irrigation water use (and origin) and the value of additional land used.
Several organic cotton projects report the incorporation of reafforestation programs (Senegal,
Paraguay). Reafforestation is also included in some IPM programs. Further development of this
LCA indicator is recommended.
4.8 Biodiversity
In addition to land use, agricultural practises affect biodiversity. The biodiversity lost from an
area where original vegetation has been cleared, such as a forest (Chapter 2.3.4) as well as
the biodiversity found in the cotton crop would need to be estimated. This indicator is now
under development for the LCA methodology (Tobler & Schaerer, 2001). This indicator relates to
differences in yield between farming systems and the associated land required in order to reach
a certain yield. If more land is needed, which may be the case for many current unsustainable
systems, biodiversity is negatively affected (see also Chapter 5, Discussion). Biodiversity is
also negatively affected through the use of broad-spectrum pesticides. We conclude that for
cotton production systems biodiversity is addressed via land use (including conversion of new
habitats) and toxicity indicators. The only factor not addressed by these indicators is the effect
on biodiversity of vast areas of monoculture cotton (e.g. in the USA). Although such monotonous
land use clearly affects biodiversity (e.g. Panek, 1997; Liu et al., 2005), no recommendations
are available on how landscape design can optimise biodiversity.
4.9 Salinisation
The effect of salinisation was introduced in Chapter 2.3.1. It mostly addressed the loss of
agricultural land due to an accumulation of mineral salts. Compared to water use it is relatively
independent of management type, because in a specific area water quality and the need to irrigate
tends to be equal for all farms. It remains however an important effect of cotton cultivation. To
date no LCA indicator has been developed for salinisation, which could be part of a water use
indicator (Chapter 4.1).
- 37 -Science Shop Wageningen UR Report 223
The sustainability of cotton
4.10 Conclusions
Conventional and organic farming systems are not as different as expected (Table 9). For example,
organic production does not take water use explicitly into account. When organic growers apply
improper amounts of organic/natural fertilisers or apply them when the crop does not need
fertiliser, they contribute to the same unsustainable elements as high-input conventional systems
(Table 9). In other words, transforming a conventional system that has unsustainable water use
to an organic system with the same degree of unsustainable water use has limited advantages.
However, the majority of organic farming systems are rain-fed and these, according to Table 9
(see foot-notes), are considered to be ‘low-input organic’.
Table 9. Relationship between farming systems and their corresponding environmental impact potential.
Recognised problems in cotton
cultivation Expressed in LCA indicators
farm management system2
degraded land
water depletion
human health effects
natural habitat conversion
wildlife contamination by pesticides
eutrophication of surface water
water use
human toxicity
environmental toxicity
global warming
eutrophication
acidification
erosion
land use
biodiversity
salinisation
energy use
conv-h + ++ ++ + ++ + ++ ++ ++ + + + + ++ ++ ++ ++
conv-l - - - - - + - - + + + - - + + + + ++
IPM-h + ++ - + +/- - ++ - + - - - +/- + ++ ++ ++
IPM-l - - - - - +/- - - - + - - - +/- + + + -
org-h + ++ + + - + ++ - - + + + - ++ ++ ++ +
org-l - - - - - - - - - - - - - - - - - - - - + + + -
1 + = moderately important impact potential,
++ = highly important impact potential,
- = slight impact potential,
- - = negligible impact potential.
2 conv-h = high-input conventional, irrigation from non-sustainable source, high pesticide (quantity and toxicity) and
fertiliser input;
conv-l = low input conventional, no irrigation, no excessive pesticide input and high fertiliser input;
IPM-h = low-input Integrated Pest Management (IPM) but with unsustainable water use. Inputs of pesticides and
fertilisers are by definition low-input;
IPM-I = low input Integrated Pest Management (IPM) with sustainable water use. Inputs of pesticides and fertilisers
are by definition low-input;
org-h = high-input organic management system, but using non-sustainable water sources, with relative low yields
and/or unsophisticated fertiliser use;
org-l = low-input organic management system and using sustainable water sources, with optimum yields and
sophisticated fertiliser use; in general, this type includes the majority of rain-fed organic cotton.
- 38 -Science Shop Wageningen UR Report 223
The sustainability of cotton
(Kooistra. 2004)
- 39 -Science Shop Wageningen UR Report 223
The sustainability of cotton
5 Social and trade related aspects of sustainable cotton
production and processing
This chapter introduces key aspects to social sustainability labour conditions, occupational
health and safety, human rights and to a certain degree, trade issues. Social issues are an
important aspect of sustainability for ideological reasons, but also for purely commercial or
public relations ends. For example, consumers would likely be appalled if they were to learn
that the organic or “sustainable” cotton t-shirt or jeans for which they had paid a premium, were
produced in sweat shop conditions or using child labour. As such, there are pragmatic as well
as ethical reasons for addressing social justice along the supply chain. It can be convincingly
argued that in order for the cotton supply chain to be truly sustainable it must address social and
economic issues, as well as environmental.
5.1 Introduction
Generally, three types of issues are distinguished in sustainability questions: social, environmental
and trade-related issues (usually indicated as ‘people, planet, profit’). It is important to note the
difference between social and trade issues. The social issues are the labour conditions of the
labourers in the production chain, whereas the trade issues are the economic conditions under
which the companies in the production chain deal with each other. Social issues are about
people (child labour, freedom of association, etc.); trade issues are about companies (premiums,
guarantees, etc.). Some confusion may arise from the fact that the word social is also used in
the expression ‘social accountability’, which is often used as a synonym for sustainability. In the
expression ‘social accountability’, ‘social’ refers to the people one is accountable to (i.e. society),
whereas ‘social’ in ‘the social issues of sustainability’ refers to issues one is accountable for.
Hence, social accountability covers more than just social issues.
This chapter seeks to present key social issues to consider in the cotton chain, outline issues
for which there are applicable standards, and present ideas as to how to coordinate along the
supply chain so that consumers can buy an end product that is both socially and environmentally
sustainable. IFOAM Basic Standards are described as a starting point and their relevance for the
organic cotton sector is discussed. Then social issues addressed by stronger social standard-
setting bodies are explored (i.e. Social Accountability International, Sustainable Agriculture
Network of the Rainforest Alliance, and Fair Trade Labelling Organisations International). Finally the
sustainable cotton supply chain is considered in terms of social and environmental sustainability
and certification based primarily on a Ugandan organic fair trade cotton cooperative case. Areas
demanding more research, coordination and consideration are presented in the conclusions.
5.2 Social issues addressed in organic cotton certification
Cotton and textiles is a new area for organic certification. The International Federation of
Organic Agriculture Movements (IFOAM) is a membership-based organisation with 771 member
organizations in 108 countries representing the organic sector. IFOAM is in a process of
developing a set of Basic Standard for textiles and cotton. Some individual certification bodies
(e.g. the Italian certification body ICEA (Instituto per la Certificazione Etica ed Ambiental) or the
UK-based Soil Association) have developed and are already using organic cotton and textile
standards. The standards generally cover organic cotton production. The dying process is also
under discussion in terms of standard development.
As of January 2003 all International Organic Accreditation System (IOAS)-accredited organic
certification bodies are compelled to include the social justice component in their own standards,
and therefore all certified operations are obliged to comply with the social standards in addition
to production and processing standards. As of October 31 20051 there were 32 IOAS Accredited
Certification Bodies (ACBs) of more than 364 certification bodies internationally. (While this number
appears quite small, it should be noted that many of the larger organic certification bodies are
1 See IOAS website – http://www.ioas.org/WEBSITE/pdfs/051031_accred_List.pdf
- 40 -Science Shop Wageningen UR Report 223
The sustainability of cotton
IOAS-accredited and applications by others are pending.) Organic cotton operations also need to
comply with these social standards in order to meet the requirements for certification if they are
certified by an IOAS-accredited certification body (e.g. Soil Association Certification Ltd, ICEA).
In this way organic cotton certification does address social issues.
In the past, all IFOAM Basic Standards were comprised of a general principle, recommendations
and requirements. The general principle of the social justice chapter (8) stated: “Social justice
and social rights are an integral part of organic agriculture and processing”. Over the past two
years, IFOAM Basic Standards have been re-structured so that the principles are in a separate
document from the recommendations and requirements. The re-structuring was voted on and
accepted at the September 28 2005 General Assembly in Australia. The articulation of principles
related to social justice now fall primarily under the recently developed ‘Principle of Fairness,’2
which states:
• organic agriculture should build on relationships that ensure fairness with regard to the com-
mon environment and life opportunities;
• fairness is characterized by equity, respect, justice and stewardship of the shared world, both
among people and in their relations to other living beings;
• this principle emphasizes that those involved in organic agriculture should conduct human
relationships in a manner that ensures fairness at all levels and to all parties – farmers,
workers, processors, distributors, traders and consumers. Organic agriculture should provide
everyone involved with a good quality of life, and contribute to food sovereignty and reduction
of poverty. It aims to supply a sufficient amount of good quality food and other products.
Some aspects of the so-called ‘Principle of Care’ in IFOAM’s new statement of principles can
in part also be linked to social justice. Box 1 contains the requirements of the IFOAM Basic
Standards pertaining to social justice. In addition, the recommendations state:
• operations should comply with all conventions of the International Labour Organization (ILO)
relating to labour welfare and the UN Charter of Rights for Children;
• all employees and their families should have access to potable water, food, housing, educa-
tion, transportation and health services;
• operators should provide for the basic social security needs of the employees, including ben-
efits such as maternity, sickness and retirement benefits;
• all employees should have equal opportunity and adequate wages when performing the same
level of work regardless of color, creed, and gender;
• workers should have adequate protection from noise, dust, light and exposure to chemicals
that should be within acceptable limits in all production and processing operations;
• operators should respect the rights of indigenous peoples, and should not exploit land whose
inhabitants or farmers have been or are being impoverished, dispossessed, colonized, ex-
pelled, exiled or killed, or which is currently in dispute regarding legal or customary local rights
to its ownership;
• contracts should be fair, open to negotiation, and honoured in good faith.3
IFOAM social standards are non-comprehensive as compared to Fair Trade Labelling Organisation
(FLO) or other social standards (SA8000, for example). However, they are based on International
Labour Organisation (ILO) requirements and do represent a baseline of social concerns and a
commitment to address social justice.
To summarize, for the purposes of exploring social sustainability, organic certification bodies
can be divided into two sub-groups: IOAS-accredited or not. The 32 IOAS-accredited certification
bodies must have social standards as part of their certification requirements. This ensures
that some basic social justice concerns are addressed with organic production and processing
certification. It is possible that organic certification bodies that are not IOAS-accredited may also
have social standards, but there is no third-party verification system in place to check this.
2 See IFOAM website: http://www.ifoam.org/organic_facts/principles/pdfs/Principles_Organic_Agriculture.pdf for
details and for the elements of the “Principle of Care.”
3 IFOAM (2002). IFOAM Basic Standards for Organic Production and Processing. Approved by the IFOAM General
Assembly Victoria Canada August 2002.
- 41 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Box 1 IFOAM Basic Standards (Chapter 8 on Social Justice)
Standards shall require that:
8.1 Operators shall have a policy on justice. Operators who hire fewer than ten (10)
persons for labour and those who operate under a state system that enforces social
laws may not be required to have such a policy.
8.2 In cases where production is based on violation of basic human rights and clear cases
of social injustice, that product cannot be declared as organic.
8.3 Standards shall require that operators not use forced or involuntary labour.
8.4 Employees and contractors of organic operations have the freedom to associate, the
right to organise and to bargain collectively.
8.5 Operators shall provide their employees equal opportunity and treatment, and shall
not act in a discriminatory way.
8.6 Children employed by organic operators shall be provided with educational
opportunities.
5.3 Issues addressed via more comprehensive social certification
The SASA (Social Accountability in Sustainable Agriculture www.isealalliance.org/sasa)
Project was a two year joint initiative of IFOAM, FLO, RA and SAI under the umbrella of the
International Social and Environmental Accreditation and Labeling (ISEAL) Alliance that finished
in April 2004. The Project was an innovative learning exercise that addressed key social issues
in sustainable agriculture in terms of certification methodologies, compatibility of the social
standards amongst the organisations involved, and key content issues such as the facilitation
of small holder access to ‘just’ markets. The social auditing issues that were addressed in the
SASA Project provide a clear indication of the kinds of concerns that need to be addressed in
sustainable agriculture production and processing. The SASA Project covered the following social
audit issues: (1) freedom of association and right to collective bargaining, (2) working hours, (3)
seasonal workers, contracts and undocumented workers, (4) child labour, (5) health and safety,
(6) wages/compensation, (7) discrimination, (8) basic treatment and disciplinary practices, and
(9) forced labour. These represent the main social concerns that are currently part of the major
social certification schemes. Although each organisation has different approaches and different
requirements or priorities within this list, it provides a sound starting point for addressing social
justice beyond the very basic social standards encapsulated within the organic requirements.
FLO’s Fair trade criteria have a slightly broader focus than the social audit issues listed above
that includes economic development in addition to labor conditions and human rights. The FLO
Fair trade standards address the following concerns, amongst others:
• fair trade premium: the organisation has the commitment and capacity to administer the fair-
trade premium in a way which is transparent for both the beneficiaries and FLO. Decisions are
taken democratically;
• export ability: producers must have the logistical, administrative and technical means to bring
a quality product to market;
• economic strengthening of the organisation: members will gradually take on more responsibil-
ity over the whole export process;
• price premium: producers receive a premium they can invest in the development of their co-
operatives;
• long-term purchase guarantee: producer and buyer sign contracts that allow the producer and
cooperative to make long-term plans;
• pre-financing/credit: producer is paid in advance at his request;
• producers receive a price for their product that covers the costs of sustainable production
and living.
Fair trade standards are important to bear in mind in considering sustainable cotton as they often
go hand-in-hand with organic certification (i.e. an operation may have both fair trade and organic
certification), in developing country contexts However, Fair trade certification is not limited to
organic production. The relevant FLO fair trade standards are those for cotton lint, and for cotton
textiles and garments. Fair trade standards pertain particularly to small farmers cooperatives in
‘developing countries’ and, in some cases, to hired labor. For example, Max Havelaar Switzerland
- 42 -Science Shop Wageningen UR Report 223
The sustainability of cotton
(a FLO member) certifies cotton from India, Pakistan, Mali and Senegal and Max Havelaar France
certifies cotton grown in Mali, Senegal and Cameroon and in 2006 will also certify cotton from
Burkina Faso. For 2005/2006, Max Havelaar France expects to certify approximately 20,000
cotton growers(Ratter 2005)4.
Social concerns are also discussed more widely within the sustainable agriculture sector.
For example, participants in a Social Justice Workshop held parallel to the IFOAM Organic Trade
meeting in Bangkok (December 2003) compiled a list of social issues needing attention in
sustainable agriculture5 some of which are addressed through certification schemes, others not.
They include the following:
• mechanisms needed for implementing social justice standards;
• no means for farmer price-setting outside market forces;
• no fair trade mechanisms in the North;
• limited capacity amongst farmers to set standards;
• no bottom-up strategies in place for evolving standards;
• no definition of child labor within the reality of the farmers;
• lack of food security mechanisms for small farmers;
• lack of standards for buyers – community/rights and responsibilities/pricing;
• lack of recognition for good practices among family/traditional farmers;
• assessing whether standards undercut traditional market models;
• no discussion of wild harvest products;
• no development element;
• undocumented workers must be more comprehensively addressed;
• better training for social justice needed (e.g. auditor training for social skills);
• chapter 8 in the IFOAM Basic Standards has no reference to price setting and subcontrac-
tor/operator relationships;
• land tenure of indigenous people needs to be addressed;
• lack of technical assistance for compliance to producers;
• IBS do not address farmers contracts with buyers;
• gender – standards address discrimination, but what about cultural values?
• uncertified organic or ‘de facto’ organic is not addressed in standards;
• regulations needed for the negotiation of fair contracts;
• farmers need access to government programs.
A significant step in assuring social sustainability for cotton would be to address the list of
nine social issues addressed by the SASA Project (see above), while bearing in mind the larger
context as reflected in some of the points raised at the Bangkok workshop. It must be noted
that specific social issues needing attention in cotton production will vary by nation, region, and
operation. That said, it is well-documented that occupational health and safety is a major concern
in conventional cotton production (see also Mancini et al., 2005).
5.4 Potential socially and environmentally sustainable cotton
supply chains
Cotton is a complex commodity in many ways. It combines social and environmental concerns
at many points along the supply chain and will demand creative coordinated efforts in order to
truly ensure and end product that is sustainable in both social and environmental realms. One
of nine pilot audits for the SASA Project was in Uganda, looking at a large organic fair trade
cotton cooperative and the parts of the supply chain that were within the country. The pilot audit
experience covered the following steps in the chain: production (e.g. farmer fields and cooperative
management); ginnery; spinning; dying; and cut-make-trim. The audit sought to consider how
the social and environmental certification systems could coordinate in order to provide a fully
certified end product for consumers. Gaps began to be understood and concerns about the
4 Article entitled, “Textiles made of fair-traded cotton launched by Max Havelaar in several countries” 08.2005 by , Saro
G. Ratter. Found on the PAN-Germany site on November 11 2005 http://www.pan-germany.org/info_db/gbr/news.
html?id=56
5 RAFI (2004). Social Justice Workshop Bangkok Thailand November 3-5 2003. Proceedings. Published, Pittsboro,
NC.
- 43 -Science Shop Wageningen UR Report 223
The sustainability of cotton
cost distribution throughout the supply chain were raised as well as issues of traceability and
cooperation amongst different certification schemes6. These issues remain pertinent and need
further exploration.
Below is an illustration (figure 6) of some of the different steps in the processing/supply chain
and certification options are provided as examples of how each level might be certified so as
to ensure that social and environmental criteria are met. This is by no means a definitive guide
but rather a tool to stimulate further thought on coordination for a more sustainable cotton end
product.
Production or farm level
Ginnery
Spinning
Dying
Cut-Make-Trim
Figure 6: An example of a cotton processing/supply chain.
Production or farm level:
• organic certification of production and basic social standards for any size of operation;
• fair trade certification social and economic standards with basic environmental, if it is a
small producer cooperative.
Ginnery:
• fair trade certification as part of a small cooperative certification for social and economic
standards;
• SA8000 certification of any size operation (though generally larger) of social criteria.
Spinning:
• SA8000 facility certification of social standards.
Dying:
• organic certification as to the dye recipes and toxicity levels (environmental sustainability;
• SA8000 certification of social criteria for the facility.
Cut-Make-Trim:
• fair trade for social criteria related to labor conditions (where linked to small producer organi-
zations);
• SA8000 certification as to social criteria in the facility.
6 From unpublished internal documents of the SASA Project and participation notes of the author of this section who
was doing doctoral research through the SASA Project from February 2002-April 2004.
- 44 -Science Shop Wageningen UR Report 223
The sustainability of cotton
Pesticide shop in Amrawati, India. (Kooistra, 2004)
- 45 -Science Shop Wageningen UR Report 223
The sustainability of cotton
6 Discussion
The goal of this study was to compare the three major types of farming systems, conventional,
IPM, and organic systems, in terms of environmental impact and which indicators would be useful
for quantifying environmental impact. How these farming systems influence the subsequent
cotton chain will also be discussed in this chapter.
A limited amount of data was available for analysis of the environmental impact of three types
of cotton production systems. The reliability of sources is not always clear, or, alternatively, a
reliable source is sometimes seriously criticised. An example of the latter is the publication in
Science by Mäder et al. (2002), who concluded that energy use per unit of agricultural product
was lower in organic than in conventional systems. This was extensively criticised by Trewavas
(2004) and others. For the current study, time did not allow all potential uncertainties from the
data provided to be traced and therefore, we concentrate on the clear-cut differences between
farming systems. Clearly, more field-work must be done in order to evaluate the environmental
effects of cotton systems.
It is interesting to speculate as to why so little has been published about the environmental
impact of cotton systems, when cotton has likely the most impact on a global scale. Much cotton
cultivation takes place in remote areas or in developing countries where it is difficult to obtain
reliable information. Furthermore, agricultural LCA studies are quite rare. Another reason could
be that environmental impact is not interesting for the processing and retail industry since it does
not directly affect the production chain. In addition, the great heterogeneity among farm types
hinders data extrapolation.
6.1 Environmental impact
Of the total cotton produced in 2004, an estimated 80% was produced under conventional farm
management, 20% under IPM and 0.04% under organic management. Although in developed
countries like the USA and Australia, regulations (including verification and monitoring) are such
that most conventional farming can be regarded as IPM, pesticide and fertiliser use may still be
high. Moreover, very large cotton fields existing in the USA and Australia may have significant
impact on biodiversity (Chapter 4.8). In developing countries much conventional farming is far
from sustainable, with demonstrated heavy pesticide impacts on the environment and on people.
Pesticide use in cotton cultivation in developing countries is high: about 50% of pesticides are
used in cotton cultivation and the types of pesticides used tend to be highly toxic (WHO classes
I and II) and, although several developed countries have banned the most hazardous pesticides,
they may still be in use illegally. With respect to environmental impact the main difference between
the three systems is the amount of pesticides used, since aspects like water use, land clearing,
and labour requirements are independent of the management system.
Average environmental impacts for a farming system are not good indicators of the environmental
impact of a specific farm. Cotton practices especially within conventional systems vary greatly.
For example, there is great diversity in farm size, degree of mechanisation, pesticide levels and
fertiliser use, varying from no till, family-run farms of less than two hectares in Burkina Faso
and India to highly mechanised 25,000 hectare farms in Brazil and the USA. Even within areas
farm size and practices can vary enormously. For example, Kooistra et al. (2005) reported
small conventional farms that did not use pesticides and other conventional farms of a similar
size carrying out 12 pesticide applications per year. IPM projects in Asia and the United States
have succeeded in reducing the amounts of pesticides used (Swezey et al., 2004). Small-scale
organic farming systems in rain-fed areas show that it is possible to cultivate cotton without
irrigation water and pesticides.
6.2 Indicators
To compare conventional, IPM and organic systems, or any other system, it is most relevant
to take the effects of pesticides as a primary indicator as this indicator inflicts the greatest
damage to people and the environment and is the most decisive factor distinguishing the three
systems. To quantify pesticide use for a given farm, amounts and types of pesticides should be
- 46 -Science Shop Wageningen UR Report 223
The sustainability of cotton
known and it is crucial to have insight into the type of pesticide application (Chapter 2.3.4). The
environmental impact could be assessed using the Environmental Impact Quotient (Kovach et
al., 1992).
Pesticide use is however not the only factor expressing negative environmental impact. Notably
water use and wide-spread negative consequences to cotton cultivation (Chapter 2.3.1) is a very
significant factor, but as noted in Chapter 4.1, the sustainability of water use is not necessarily
related to the type of farming system. Nevertheless, if current systems are compared in
developing countries, then it should be noted that most organic cotton systems are rain-fed and
the majority of conventional and IPM systems are irrigated, often unsustainably.
Water use implies more than simply monitoring the amount of irrigation water used per unit of
agricultural surface, as water quality, probability of salinisation and vulnerability of the source to
depletion are also important. To the best of our knowledge a comprehensive water use indicator
has not yet been developed. Bos et al. (2005) provides an interesting proposal on water use
efficiency in cotton fields, but this should be extended with sustainability of the water source.
Because of the important environmental impact of water use related to cotton production systems,
we recommend further research in this area. Since most cotton production is carried out in areas
that are sensitive to salinisation and desertification, cotton production systems should be geared
towards avoiding this, so that the land cropped to cotton can be used sustainably.
Global warming is strongly related to the amount of synthetic fertiliser and fuel used. Relatively
fixed emission values are known for the production and use of both compounds. The same holds
for eutrophication and acidification where also the burning of organic material is important. Burning
of cotton residues is common practice in many conventional and IPM farms to reduce the build-
up of pathogens such as Verticillium dahliae, which forms massive amounts of microsclerotia
on the decaying shoot. Most likely the build-up of these pathogens is of lesser importance in
systems where crop rotations are done.
The selection of the indicators will depend on the aim of the customer. If a specific company aims
to lower its overall impact on the environment by buying more sustainably produced cotton, it
would probably be a good choice to buy cotton from organic, low-input, rain-fed cotton farms in
e.g. West Africa, India and Paraguay. It is, however, realistic to keep in mind that this would have
significant additional cost implications to the company and hence the price of the end-product.
If however the demand for this type of cotton would exceed the rain-fed area available (20% of
the world cotton production), it could negatively affect land use. In that case, a balance needs
to be found between an organic system with lower yields and a conventional system with higher
yields. On the other hand, if the goal would be to decrease the environmental impact of the
world’s cotton production significantly, very likely large-scale support of IPM practices combined
with more efficient irrigation systems would be the best option. Whether such an IPM system
would then be environmentally better than an organic system depends of the quantity and types
of pesticides allowed in the IPM system and the differences in yield between the two systems.
Clearly a significant step forwards can be made through farmer education as happens in IPM
Farmer Field Schools and organic projects (often supported by Non-Governmental Organisations),
given that the main practical problems (e.g. insect resistance to pesticides, uneconomic spraying
and irrigation) are based in lack of knowledge vis a vis the agro-ecology of the cotton crop (FAO,
2004; personal communication Tobler, 2005). Further indicator development for water use,
land use in terms of effects on wildlife, biodiversity, and destruction of natural habitats would be
relevant in order to be able to quantify the environmental impact of cotton cultivation systems.
A main research task is to generate information from cotton farms in tropical regions, so that
environmental impacts can be quantified more precisely. Currently, all published LCA studies to
date (Tobler & Schaerer, 2001; Laursen, 1997) and LCA software (Simapro, Ecoindicator, GaBi
4) rely on information from highly productive US cotton farms.
6.3 Choice of farming system
As illustrated above, the type of cotton production, conventional, IPM or organic, does not
necessarily predict the environmental impact of a particular cotton farm. The difference between
the three farming systems is less in developed countries and in places where subsistence
conventional farming is applied without pesticides. In developed countries, conventional farming
often approaches an IPM system because persistent and highly toxic pesticides are not allowed
and the farmer is often aware that a sophisticated minimisation of pesticide use tends to be
- 47 -Science Shop Wageningen UR Report 223
The sustainability of cotton
economically optimal. In terms of environmental impact, optimal IPM and organic systems
converge to some extent in developed countries. In optimal systems, IPM scores well with respect
to fertiliser use efficiency and the efficient use of narrow-range, low-persistence pesticides and
in terms of higher yields. Organic farming scores well in terms of its wide rotations and lack of
pesticide use. This, however, is not to say that the types of pesticides promoted by IPM are
harmless. Still, more data are needed on yields of IPM and organic systems relative to the inputs
required. Mäder et al. (2002) concluded for food rotations in Switzerland that organic production
was more efficient in terms of inputs per unit of harvested products than IPM production. Although
this is a widely different system than (sub)tropical cotton, it may indicate the potential of other
organic systems.
Recently, Cherrett et al. (2005) calculated the ecological foot-print of different cotton systems
on the basis of literature data. The ecological foot-print of cotton cultivation in Punjab was,
irrespective whether it was organic or not, the highest (approx. 3.0 ha), while that of conventional
and organic cotton cultivated in the USA was lower (approx. 2.6 and 1.8 ha respectively), which
is due to the low yields per unit of area in Punjab. However, in this comparison the environmental
effects of pesticide use have not been taken into account.
In less-developed countries, where most cotton farms are situated, differences between organic
and conventional cotton production in terms of sustainability favour organic production of which
the differences primarily lie in the unsustainable application of toxic pesticides in conventional
systems. Thus, in future studies the usually higher yields in current conventional and IPM systems
must be weighed against their levels of sustainability. Comparisons in yields between different
farming types are lacking. Part of high yield levels in conventional systems is certainly due to an
excessive use of pesticides. It should also be noted that conventional systems do not necessarily
provide the highest yields (e.g., USA, Table 1).
The differences mentioned in potential yield between organic systems and conventional or IPM
systems are largely due to the fact that in conventional systems crop growth can be better
managed. Crop growth is managed through the use of artificial fertilisers, which can be directly
taken up by plants, and by the application of synthetic pesticides, which can be used preventatively
or curatively after a disease or pest has developed. By contrast, in organic systems, farmers
have less flexibility as they rely primarily on organic manure that first needs to be mineralised
by micro-organisms (which takes time) for plant nutrition, and on pesticides from natural origins
which generally have a relatively low impact on selected pest and/or disease organisms.
An argument for prioritising organic rather than low impact conventional or IPM cotton production
is that that organic cotton production is certified by independent third-parties. Although a
consumer needs to be able to trust the certification body or label’s claims, certified production
systems offer a monitoring of farmer practices and farm-level evaluation of the system in
place. Currently organic standards do not specifically address the origins of the land under
cultivation (i.e. whether it was a natural area or agricultural land previously) nor do they address
the origin and sustainability of the irrigation practises, despite general aims to promote an
overall sustainable approach to production. For optimal, highly-technological conventional cotton
production systems, standards and a verification system are necessary because without these
elements, there is no way to differentiate clearly unsustainable conventional farming systems
from more sustainable conventional farming systems. Having said this, the term ‘conventional’
(or IPM) may no longer be accurate for optimal systems. In conclusion, a clear description of
a sustainable IPM system is needed as well as some form of system to verify that results are
obtained with respect to the main impacts. Subsequently a label so that the consumer can
identify end products as having come from that system.
6.4 Social and trade issues
Most verification and certification bodies mainly deal with one aspect of sustainability (social,
environmental or economic) organic certificates mainly have environmental requirements,
SA8000, FWF and ETI focus on social issues, and FLO (fair trade) concentrates on trade and
development concerns. Some integration of the three aspects of sustainability is underway
amongst standard-setters and certification bodies in recognition of the relevance of all three. For
instance, IFOAM standards include six social requirements but these standards are only valid for
production and do not address the whole supply chain for cotton.
- 48 -Science Shop Wageningen UR Report 223
The sustainability of cotton
6.5 Conclusions
• Water use (including the potential negative consequences of salinisation and water resource
depletion) and pesticide use represent the greatest environmental problems in cotton
systems. Water use and type of irrigation are not explicitly included in the standards for
organic production at present. In principle, unsustainable water use may occur in all three
farming types, although the majority of organic cotton systems are not irrigated. In developing
countries most organic cotton is rain-fed while the majority of the IPM and conventional cotton
is irrigated. The environmental impact of water use is considerably lower for rain-fed than for
irrigated cotton, however the opportunities for yield gains by optimising water applications are
also consequently lower.
• The widespread use of extremely and highly toxic pesticides in developing countries is a major
concern both for the natural environment and for people who are exposed, especially workers
applying these pesticides. Pesticide application methods strongly affect the potential impact
on field workers and the environment; thus, the way field workers apply pesticides needs to
be better understood.
• Biodiversity is negatively affected through the use of broad-spectrum pesticides, through
conversion of natural habitats and through loss of land to degradation or shifting farming
practices. For cotton production systems biodiversity is addressed through the land use and
toxicity indicators. The effect on biodiversity of cropping vast areas of monoculture of cotton
needs additional attention.
• Limited data on the environmental impact of cotton systems is available. In addition, the great
heterogeneity of farm types hinders data extrapolation.
• Because of the great diversity in farming practises it is difficult to express the environmental
impact of conventional, IPM and organic cotton farming on a systems level. The use of
pesticides is highly problematic in several important cotton-producing countries and here the
distinction between conventional and organic cotton production is evident.
• In developed countries conventional cotton production may to some extent approach IPM
given that the most persistent and hazardous pesticides are not allowed and because inputs
are increasingly optimised for cost reasons. However, the lack of clear-cut definitions for IPM
and conventional systems also contributes to this conclusion.
• Organic farming is a certified type of production, while conventional is not, and IPM is generally
not. With certified production it is easier for retailers and consumers to understand the history
of the produce and thus make inferences about environmental impact. Although the general
aims of organic farming are to promote an overall sustainable approach, the standards need
improvement as they do not sufficiently address land and water use issues. For optimal,
highly-technological conventional cotton production systems, a standards and verification
system is necessary because otherwise there is no way to differentiate clearly unsustainable
conventional farming systems from more sustainable conventional farming systems.
• Reflecting on the desirability of organic cotton production, water and land use issues should
be kept in mind. Expanding the organic cotton area in rain-fed regions by converting low-input
conventional farms to organic could be an optimal approach as yields can be as high as those
in IPM and conventional cotton, water use can be generally sustainable, and no pesticides are
used. This applies especially to countries that apply relatively hazardous pesticides. However,
for conversion to organic cotton in irrigated areas it is necessary to address water and land
use issues. Organic cotton production on eroded, pesticide-polluted and/or salinised soils
would not be a viable option. Conversion to organic cotton production on large tracts of land
that are mainly conventional can be a challenge if pest pressure is high.
- 49 -Science Shop Wageningen UR Report 223
The sustainability of cotton
6.6 Recommendations
• The current study could be extended by an analysis of each of the approaches against
their characteristic visions and ultimate goals. Descriptions and definitions of optimal, most
sustainable conventional and IPM cotton systems for both rain-fed and irrigated cotton are
needed. Certification of such systems and labelling of products should be considered.
• Knowledge of what happens in the field is the basis for recommendations. Clearly for such
an important crop as cotton more field work is needed to collect quantitative field data on
environmental impact and yields in various developing and developed countries. This could
lead to an extended Life Cycle Assessment comparing conventional/IPM and organic farms in
several countries using the following indicators: Water use, Human and environmental toxicity,
Global warming, Eutrophication, Acidification, Erosion, Land use, Biodiversity and Salinisation.
Also the environmental impact of cotton production should be compared with the production
of other fibres such as polyester, viscose, hemp, and linen (but see Cherrett et al., 2005 for
first results on a comparison of cotton, polyester, and hemp).
• Further research is needed on how various stakeholders influence cultivation methods.
• Water and land use indicators need further development and implementation in LCA-cotton
studies. The standard for organic production should also address these elements.
• The social, environmental, and trade-related issues that are particularly relevant for cotton
production in different country contexts and at different points along the supply chain merit
further research in order to build a comprehensive picture of sustainable cotton beyond just
environmental indicators.
• Organic certification done by an IOAS-accredited certification body assures that a baseline
level of social justice is maintained at the production level. To address more than these
minimal social concerns, to address trade-related concerns and to address the other parts of
the supply chain, other means are necessary. For example, Fair Trade Labelling Organisations
(for small producer organisations) or Social Accountability International (for larger operations
and other steps along the supply chain) are possible certification partners.
• Certification packages need to be constructed that address environmental, social, and trade-
related concerns throughout the whole supply chain. Costing and traceability for such a system
clearly needs attention.
• Verification and certification bodies in the three sustainability areas (social, environment, trade)
should study how they can co-operate.
- 50 -Science Shop Wageningen UR Report 223
The sustainability of cotton
- 51 -Science Shop Wageningen UR Report 223
The sustainability of cotton
References
Ajayi, O., Camara, M., Fleischer, G., Haïdara, F., Sow, M., Traoré, A., van der Valk, H. 2002.
Socio-economic assessment of pesticide use in Mali. GTZ/University of Hannover Pesticide
Policy Project. FAO/INSAH. Special Issue Publication Series, No. 6. February 2002.
Allan Woodburn Associates. 1995. Cotton – the crop and its agrochemicals market. Pesticide
News 30.
Altieri, M.A., Rosset, P. 1999. Ten reasons why biotechnology will not ensure food security,
protect the environment and reduce poverty in the developing world. AgBioForum 2: 155-
162.
Ashraf, M. 2002. Salt tolerance of cotton: some new advances. Critical Reviews in Plant Sciences
21: 1-30.
Audsley, E., Alber, S., Clift, R., Cowell, S., Crettaz, P., Gaillard, G., Hausheer, J., Jolliet, O., Kleijn,
R., Mortensen, B., Pearce, D., Roger, E., Teulon, H., Weidema, B. and Van Zeijts, H. 1997.
Harmonisation of environmental Life Cycle Assessment for agriculture. Final Report Concerted
Action AIR3-CT94-2028. Brussels: European Commission.
Aumonier, S., Collins, M. 2005. Life Cycle Assessment of disposable and reusable nappies in the
UK. Environment Agency, Rio House, United Kingdom.
Australian Academy of Science (AAS). 1999. Integrated pest management – the good, the bad
and the genetically modified. No.4. 2002.
Banuri, T. 1998. Environmental impact of cotton production and trade. Published within the UNEP
project on trade and environment. International Institute for Sustainable Development.
Barlow, M., Clarke, T. 2003. Blue gold: The fight to stop the corporate theft of the world’s water.
The New Press, New York.
Benjaminsen, T.A. 2002. Enclosing the land: cotton, population growth and tenure in Mali. Norsk
Geografisk Tidsskrift 56:1-9.
Blackburn, L.G., Boutin, C. 2003. Subtle effects of herbicide use in the context of genetically
modified crops: a case study with glyphosate (Roundup®). Ecotoxicology 12: 271-285.
Bos, M.G., van den Bosch, R., Diemont, H., van Keulen, K., Lahr, J., Meijerink, G., Verhagen,
J. 2005. Cotton-Made in Africa: Quantifying the sustainability of cotton production, Alterra,