The impact of legumes after 21 years of rotation (11 legume crops) continue to positively impact soil health. The addition of a legume in a cotton rotation impacts all aspects of the cotton production. The evidence was clear from the soil N status and plant accumulation of N (nitrogen use efficiency (NUE), Carbon:Nitrogen (C:N) ratio and Apparent Nitrogen Fertiliser Recovery (ANFR)), yield, fibre quality to seed nitrogen and protein content. The most significant benefit of including a legume was evident when a legume followed a summer and winter fallow before the next cotton crop was sown (CV~C and CF~C). The long fallow allowed for the legume crop to reach optimum biomass production and time for the fixed atmospheric N to be mineralised and available for the following cotton crop. The addition of wheat into the cotton rotation also had a positive impact when compared to the continuous-cotton rotation. The impact from adding wheat was not as beneficial as a legume, however there were benefits by adding wheat into the cotton system.

This long term rotation experiment at ACRI, established by the late Dr Ian Rochester in 1994, has assisted with the understanding of soil and plant nutrition interactions. It has been a pivotal site in developing the industry crop nutrition requirements (especially N) but has evolved over the years to include the following research initiatives and outcomes:

- Advancing the understanding of legumes and their role in cotton crop systems.

- The role of stubble management in soil fertility.

- Improving the understanding of fertilisers (the role of timing, application and type) in increasing cotton productivity.

In recent times the site has also been important to the following research initiatives

- The dynamics of soil carbon in various rotation systems. (Data published from 1998 to 2008 reported the only increase in soil carbon under a cotton farming system in Australia).

- Greenhouse gas emissions from various cotton rotations and N application practices.

- Increasing attention on soil microbiology because of the sites high fertility.

- The response of cotton genotypes to various fertility regimes.

The 2016/17 season was the first time a Bollgard® 3 variety had been sown in the long term trial. The new variety responded to high rates of N and positively to the legume rotations. The fitted curves for N rates and N uptake revealed a positive linear relationship for the new Bollgard® 3 variety for the cotton-vetch-summer and winter fallow-cotton rotation. This is something to investigate for Bollgard® 3 response under a legume rotation. Under the continuous cotton rotation, the optimum N rate was calculated as 407 kg N/ha, however only yielding 12 bales/ha. In comparison the CV~C rotation only required 147 kg N/ha to achieve a much higher 14 bales/ha. The new variety responded positively to both vetch and faba bean in rotation and continued to follow the trend to the previously sown Bollgard® II varieties.