Since1988, CRDC has funded a series of projects focussing on the management of problem weeds (nutgrass, polymeria takcall, cowvine, budda pea, lippia and others) and weed management systems for the cotton industry. The research included work on transgenic, herbicide tolerant cotton (Roundup Ready, Oxygene,

Liberty Link and 2,4-D tolerant cotton), noogoora and thornapple competition, herbicides for use with pigeon peas, and monitoring weed management systems.

Nevertheless, weeds continue to cause significant problems for the cotton industry, and the cost of weed control increases year by year. Four of the key weed problems in cotton are nutgrass, cowvine, bellvine, and polymeria takcall. While considerable research has been undertaken on nutgrass, there are still considerable gaps in the knowledge of this weed and its management. Even less is known of the management options for cowvine, bellvine and polymeria takcall in irrigated cotton.

Problems with weed control are being exacerbated by changes in the farming system, with trends towards reduced cultivation, reduced chipping and stubble retention. These changes are placing increasing pressure on the use of herbicidcs, and as a consequence, a group of problem weeds that are largely tolerant of the

commonly used herbicides has been selected out. The increase in residual herbicide use also has had flow-on environmental impacts both on and off the farm.

In order to address this situation, ACGRA has requested that the weeds team put all available information together into a WEEDpak format during the next year. Much of the data collected in DAN 124C and previous projects will form an important part of WEEDpak. This will provide a valuable resource to cotton growers and will also allow future research and extension priorities to be more readily identified.

Abstract and poster format of research relating to project CSP105C was presented at the International Conference held in Colorado, USA, as part of the Keystone Symposium series. The conference was entitled "Plant Foods for Human Health: Manipulating Plant Metabolism to Enhaiice Nutritional Quality". The speaker list included researchers leading the field in plant metabolism and engineering, and also included speakers whose expertise were in the fields of nutrition and assessing solutions for world hunger. The key plant metabolic pathways discussed at the meeting were broad and often overlapping and included vitamin E, folate, vitamin C, iron, zinc, calcium oxalate, lipids and fatty acids, amino acids, terpenoids, sterols, carotei-toids (vitamin A), isoflavonoids and lignans, alkaloids and phytonutrients.

An important issue highlighted at the conference was the enormous potential of genetic engineering to modify plant metabolic pathways for specific downstream applications. Examples of these include nutritional applications such as altering fatty acid profiles for healthier vegetable oils which could reduce cardiac problems, a pharmaceutical application may be tlie diversion of glucosinolate pathways in poppy for controlled production of morphine or codeine, and an indtlstrial application may be the overproduction of rnonoterpenoid oils from mint species for use as fragrances. The research conducted as part of CRDC project CSP105C has multiple applications. Modifying terpenoid biosynthesis in transgenic cotton could result in plants with increased tolerance to diseases and insect feeding, reducing gossypol levels in the seeds would improve the quality of cottonseed for use as stock feed and edible oils, and overproduction of gossypol may provide a market opportunity if gossypol is proven to be a safe and effective male contraceptive. The

conference highlighted just how achievable these goals are since several other pathways have been successfully modified using genetic engineering approaches.

The development of decision support systems to assist cotton growers is a key part of

facilitating sustainable cotton cropping within Integrated Pest Management principles.

One system that has been in use in Australia is the SIRATAC/OZCOT simulation

model which predicts yield based on agronomy and weather input daki. This model is

currently undergoing improvements including the incorporation of cotton response to pest damage. Reliable prediction of yield compensation to pest damage will enable growers to improve confide~icein the principles and maximise natural recovery and minimise pesticide application and its negative impact.'

Over the last several yeal-S,field trials at the CSIRO Narrabri have generated a substantial dataset on the growth and yield effects of damage ranging from early season defoliation and tip damage to single and multiple fruit removal events (by Lewis Wilson, Victor Sadras and Tom Lei). It has become clcar that reliable prediction of compensation requires the modelling of not only fruit losses but canopy development which can bc significantly altered by pl-c-squaringterminal and leaf damage. To fully account for the process of compensation, we need to include in the model space- and time-specific variations in light interception, carbon assimilation, regrowth and fruiting

potential following damage. Specifically, the spatial arrangement of leaves and branching structure should be explicitly accounted for in the mechanistic modelling of pest damage, the response in plant growth, and its consequences in light capture.

T o best achieve rapid initiation of the modelling work, I spent one month collaborating with Dr. AkioTakenaka at the National Institute for Environmental Studies (Japan). Dr. Takenaka has an established expertise in modelling the interaction between canopy architecture, light interception and carbon uptake (see a list of his relevant papers below). W e will develop a compensation submodel capable of simulating the effect of tip, leaf and early square damage on lateral branch and canopy growth and the consequences of architecture on light interception, carbon assimilation, and fruiting dynamics. Calibration of parameters (e.g., leaf area index, photosynthesis, number of fruiting sites) will be derived from published results (see references below) and recent field data (from C S R O Narrabri). Although Dr. Takenaka has not worked with cotton, the physiological principles from his research and modclling should apply to cotton and may provide input of new ideas into OZCOT which could assist with futureenhancements. We will also explore the possibility for Dr. Takenaka to visit the OZCOT modelling group in future to expand this collaboration if necessary. The linking of the compensation submodel to OZCOT will be done with the approval and support of Dr. Stevc Milroy and Dr. Mike Bangc who are currently directing the core model revision.

the 5th International Conference on Precision Agriculture* was held in Bloomington, Minnesota from July 16 - 19, 2000. This conference brought together 698 delegates from around the world including representatives from commercial companies, researchers, farm managers, consultants and growers. The conference highlighted that an expanding number of companies are becoming involved in the development and application of technologies specifically for use in site-specific management systems. Researchers are using this technology to develop more economicalIy and environmentally sound farming systems for an increasing number of crops worldwide. We presented papers on establishing an opportunity index for precision agriculture and the variable-rate application of nitrogen fertiliser to Australian cotton fields, which are included as Appendices I and 2 respectively. Full proceedings from the conference are now available.

Dr Brubaker was invited to present of summary of the research completed under CRDC grant CSP85Cand continuing under CSP120C in the cotton workshop during the IX Plant and Animal genome conference (Jan 17-18 2001) in San Diego California. I also was asked, as one of the two Australian representatives to the steering committee of the International Cotton Genome Initiative (ICGI), to a lead a discussion of this new initiative at the end of the Cotton Workshop at the conference. The International Cotton Genome Initiative arose from a cotton marker workshop organized by Greg Constable, funded by CRDC and CSD, and held at CSlRO Plant lndustry March 2000. Additionally I presented a poster describing the first genetic linkage map of an Australian cotton species, and I was co-author on a second poster arising from a collaboration with US cotton researchers that developed out of the ICGI.

The initiation of the purchase of the high clearance applicator was derived after the research officer (Grant Roberts) conducted experiments at seven different field sites in I997/1998 from Carrol in NSW to Bowenville in Queensland. In conducting these experiments it became obvious that a major limitation to managing these experiments was the ability to quickly apply herbicides to experimental plots in the field. At that time the researcher and assistant prepared herbicides the previous day and applied herbicides to trial plots using a 4m hand boom. Where large scale field trials were necessary it was nearly impossible to apply all the required treatments in the same week, which was necessary to maintain the integrity of the experimental design.

Dryland cotton fanning systems are complex and usually comprise of a number of different rotational crops in conjunction with summer and winter fallows. The standard system employed in dryland cotton production is the planting of cotton after a 10-month fallow from a winter cereal(Marshall 2002). However, there are a number of alternative planting options, such as sorghum, sunflowers or Innize, which for reasons of price or available soil moisture should be planted in place of cotton. Although originating from hot and regions and tolerant of long dry periods, modern cotton varieties require warm to hot growing conditions and reliable rainfall during the growing season. Cotton should be planted into a minimum of 60cm of wet soil, where as in lower rainfall areas 90cm of wet soil is preferred to provide sufficient moisture between rainfall events (Bange et al. 2002). For these reasons, dryland cotton production tends to be limited to areas receiving greater than 600mm of rainfall annually, with approximately 40% of that rain falling during the summer growing season. Currently, cotton is grown commercially from Hillston in the south-west of NSW to Emerald in central Queensland, although trial plantings have been established in the Northern Territory and the northwest regions of Western Australia. Dryland production accounts for approximately 20% of the total area planted to cotton, with the junior production regions in the lower and upper Namoi Valley, Moree, the Darling Downs, south-west Queensland and central Queensland (Marshall et al. 2002).

Adequate soil moisture and the likelihood of receiving planting rains are critical factors for growers in deciding which crops they are likely to plant, particularly in the summer cropping phase. The critical nature of soil moisture to dryland production has seen the evolution of a number of different planting configurations, including solid, single skip row and double skip row planting that maybe employed by growers to conserve soil moisture throughout the season.

Weed management in farming systems involving dryland cotton is then by nature equally complex. A number of the residual herbicides used in rotational crops may damage cotton, in particular the sulfonylurea and triazine herbicides such as chlorsulfuron and atrazine. Equally, a number of the common cotton herbicides have long plant backs to either winter cereals or other summer cereals used in rotation with dryland cotton. To preserve soil moisture, many dryland growers have adopted minimum or zero tillage systems that are almost solely reliant on herbicides for weed control . Control measures therefore must be flexible to allow last minute changes in the crops grown due to soil moisture limitations or price fluctuations and need to provide adequate levels of protection against weeds in the chosen crop, as well as in the planting configuration being used.

To better understand the weed management issues of this complex farming system and to provide direction for future research efforts, a scoping study was initiated in July 2001 by the Cotton Research and Development Corporation (CRDC), Australian Cotton CRC, Grains Research and Development Corporation (GRDC) and CRC for Australian Weed Management. The scoping study was a collaborative project involving scientists and technical staff from Queensland Department of Primary Industries and New South Wales Agriculture.

The main recommendation arising from six and a half years of research into how to

gin cotton with an emphasis on quality relates to practices outside of ginning, rather

than ginning itself.

The research in part produced a rigorous comparison of Australian cotton against

cotton from a similar industry (USA). We lost. While Australian cotton had several

good attributes, it was shown to be appreciably higher in Nep content.

Other individual studies showed the quality aspects of UNR cotton, of combinations

of cotton moisture and heating levels, and the relevant importance of a wide range

of quality attributes of lint in predicting a wide range of quality attributes of yarn

and fabric. Initial work was also carried out into how much decisions taken during

growing, affect cotton quality post harvest.

The main outcome of this research is to show that there is a Ginner's Paradox.

There are two main groups of quality attributes for cotton lint. The first group is

concerned with cleanliness and appearance (leaf, colour where weathering is not an

issue, preparation, etc). The second group relate to performance in the hands of the

industrial buyer (several length attributes, plus Neps and immature fibre content).

The former affect bale price strongly, but quality in the hands of the industrial buyer

little or not at all, where as the latter affect quality in the hands of the industrial

buyer strongly but bale price little or not at all.

The paradox for the ginner is that he can gin for best results in one or other of these

groups of attributes, but not both.

Current classing practice emphasizes the former, but punishes or ignores the latter.

For pragmatic reasons, ginning practice follows classing practice. The end result is

that an opportunity to produce a step upwards in quality across the Australian

industry is going begging.

Those classing practices must change before the results of this research can be

adopted.

This project identified a method of evaluating crop growth to predict yield response to Pix. * Two nutritional disorders were examined. Firstly, the long fallow (Galathera) syndrome is undoubtably due to poor infection by mycorrhiza. Zinc fertilizer strategies and possible soil management strategies were identified to minimise the problem. Secondly, waterlogging induced iron chlorosis was identified, but the condition was not completely solved by iron fertilizer: removing foliar symptoms did not necessarily improve yield. *Nutrient diagnosis. A database has been established to indicate desirable levels of all nutrients in cotton leaf tissue. In conjunction with experiments where deficiencies are confirmed, this data can be used to assist with diagnosis of crop nutrient status.

The aim of the project was to produce more virulent strains of a naturally occurlng entomopathogenlc nuclear polyhedrosls virus of Heliothis species.