Cotton farmers have been adapting and changing their farming systems for as long as cotton has been grown. New machinery, new technology and streamlined management have all assisted in this process. But what of the future? Can we make further improvements? What factors and new information will drive to a more sustainable cotton farming system?

We have conducted experiments on this problem for five years, and summaries of this work have been reported at earlier cotton conferences (1994 and 1996) and in the CSD premature senescence grower information leaflet. In this paper we would like to focus on the topic of how damaging is premature senescence? Is it a problem that growers need to be concerned about, or is it just a sign of a high yielding crop?

New insecticide technology is needed in the Australian cotton production system to reduce reliance on old chemical groups and to maintain profitability. Conventional insecticides provide us with the backbone of insect control in cotton and will remain as integral and necessary tools for incorporation into integrated pest management (ERM) programs in the future. Several new classes of insecticides such as the oxadiazines, phenylpyrazoles, avennectins, neonicotinoids, spinosyns, pyrroles and diacylhydrazines have new modes of action and offer great pronxise for the future of insecticide resistance management (ERM) and IPM in cotton. Generally these toxicants are more environmentally acceptable than existing insecticide chemistry; because they offer greater specificity to target pests, have greater intrinsic activity, achieving acceptable insect control at low field use rates and because they degrade rapidly in the environment to harmless metabolites

The Australian cotton industry continues to face a major challenge in controlling heliothine caterpillars. As the bollworm, Helicoverpa armigera, has evolved resistance to most major chemical insecticides, interest in biological control agents has increased. Such agents would avoid the environmental dangers and costs associated with chemical insecticides. An important approach to using orally acting biological control agents that confer protection agains the bollworm is to genetically engineer cotton by inserting genes for the agents (Llewellyn et al 1992). The first genes available for use in this approach were those encoding the insecticidal proteins from Bacillus thuringiensis (Bt). INGARD-cotton carrying a Bt gene is now being grown commercially in Australia and will form the basis of heliothis control strategies (Llewellyn et al 1994). However, this current dependence on a single control gene means that an alternative form of resistance to heliothis would be of great interest. This would allow the careful management of the Bt gene now used in INGARD-cotton in order to retard the development of resistance and provide an insurance policy should resistance to Bt become insurmountable.

Scepticism on the applicability of models is not due to their rarity nor lack of exposure. A recent survey of models targeted at farm production and catchment management (Hook, 1997) recorded over 90 models or computerised decision support systems (DSS) developed and/or supported in Australia. The compilation included models relevant to the cotton industry, the OZCOT and CBRCOT cotton models and the now defunct STRATAC DSS. It would be safe to assume that there are probably runny more such products than those listed in this report. Yet, in Australia (Cooke, 1994) and world-wide (Plant, 1997), farmer acceptance of models and decision support systems has been disappointingly low. In this paper, our objective is to relate our experiences of how some farmers and consultants have benefited from the application of systems models in their farming operations. We argue that this recent effort in applying models within industry has important distinguishing features from past efforts on decision support systems and that the current market pull for commercial access to systems models may be sustainable. We conclude with proposed plans to progress towards commercial delivery of systems simulation to farmers in the northern cropping region

Residues of agricultural chemicals in soil and run-off water are a major cause for concern in cotton production. Microorganisms are known to be important in residue degradation but the role of soil fauna in this process has received much less attention. In some cotton-growing soils the Collembolan, Proisotoma minuta, is a predominant species and may contribute to residue reduction. The effect of herbicides used in cotton-producing areas on P. minuta is being examined to determine whether residues are likely to be metabolised by the insect.

Dryland cotton production is an important component of cropping systems in the northern grain belt. It has been a difficult crop for many growers to integrate into their existing cropping systems because of its high soil water usage, narrow planting window, long growth period, and high pesticide inputs, to name just a few factors. A number of cotton industry research projects have commenced in the last 5 years to look at different aspects of dryland cotton cropping systems. This work, in conjunction with the practical experiences of many growers who have been developing conservation farming systems over the last decade is providing some clear guidelines towards the development of sustainable cropping systems and best management practice for this section of the industry.

The commercial release in Australia, in 1996, of transgenic cotton containing the Cry1Ac gene of Bacillus thruingiensis Berliner var. Kurstaki (Bt), was a significant step towards reducing the use of conventional pesticides on cotton crops, for the control of Helicoverpa armigera. It is possible, however, that H. armigera could develop resistance to the Bt toxin, as has occurred with the use of other insecticides. The seasonal variations in the efficacy of Bt transgenic cotton, reported from field trials (Fitt et al 1994), could exacerbate the development of resistance (Daly, 1994), so there is a need for these variations to be better understood.

Cotton leaf roll dwarf virus (the causal agent of Cotton blue disease) or a virus closely related to this species was detected in an asymptomatic Sea Island cotton plant (Gossypium barbadense) in Laivai, Timor Leste.

Precision agriculture (PA) as a crop management philosophy was first hypothesised in the early 1990's as a way of utilising the development of technology such as the Global Positioning System (GPS) and variable-rate crop applicators to produce crops in a more sustainable fashion. Historically, fields have been managed as homogeneous units, receiving equal amounts of crop production inputs. The rational behind precision agriculture, or site-specific crop management, is that by identifying within-field variability in crop and soil attributes (e. g. cotton yield, soil nitrogen levels) and their origin, it then becomes possible to optimise crop production inputs such as pesticides and fertilisers on a point-by-point basis. Implicitly, this lowers the potential for the over and under-application of these crop production inputs, thus increasing profitability for the grower whilst simultaneously reducing the probability of adverse environmental impacts such as groundwater or surface water contamination from the over application of agrochemicals. While progress in developing a filmy integrated PA system has been slow over the past decade, research from a number of industries worldwide has highlighted there will be many benefits to be galled from adopting such a system. Secondly, triller technological advances in airborne imagery collection, on the-go sensor development and computer processing techniques means there is an unprecedented number of tools available to aid in crop management. Furthermore over the last decade it has emerged that each country and crop will its own unique requirements within the larger framework of PA.