Predicting the effect of fertiliser on 2nd rotation decline in hardwood plantations

Project No: PNC304-1213

Almost 1 million hectares of hardwood plantations are growing in Australia. Dominated by Eucalyptus globulus (55%) and E. nitens (24%), most are cultivated for pulpwood on 10 to 15 year rotations and many are now ready for harvesting and subsequent replanting.

However, while previous experience shows that applying appropriate amounts of fertiliser can prevent second rotation decline, there is no reliable, cost-effective method to predict how a plantation will respond to fertiliser or when it is best applied.

Researchers from The University of Melbourne, together with several forest growers and funding from FWPA, have recently compiled and analysed historical fertiliser trials to provide new insight. 

The research considered data from 49 sites across south-western and south-eastern Australia, with the site measurements collected over almost 20 years volunteered by a variety of industry collaborators. 

Growth responses to fertiliser applied at establishment (age 0 and 1; ~250 kg ha-1 of nitrogen (N) and ~80 kg ha-1 of phosphorus (P)), mid-rotation (age 4-5; 250 kg ha-1 of N-only) and combinations of the two were determined for 49 experiments across southwest and southeast Australia. 

Growth responses to fertiliser typically lasted 4 years and substantial gains in final yield could be achieved, particularly at sites receiving mid-rotation fertiliser.  

Growth responses to fertiliser varied substantially between sites, highlighting the need for a systematic way of identifying responsive sites. Models were developed to predict growth response to fertiliser from soil and/or foliar tests in combination with climate variables.  

Potentially mineralisable N, from 0-10 cm soil samples, in combination with long-term average annual rainfall, could predict short-term growth responses to establishment (N and P) fertiliser with 70% accuracy.  

Initial attempts to model  growth responses to mid-rotation fertiliser were unsuccessful to any level of accuracy. Although the best establishment fertiliser models only predicted short-term growth response (i.e. one year after application) with any level of accuracy, they show significant value in identifying sites highly unlikely to respond to fertiliser.

To run any simulation model, a plantation manager needs to acquire appropriate site data (such as climate variables) and send soil or leaf samples to a laboratory for analysis. The results are used by the model to predict the percentage of volume growth following the application of fertiliser. At this stage, the researchers suggest plantation managers use the models as tools to rank sites according to predicted fertiliser response, allowing managers to deploy fertiliser resources to sites more likely to show a significant volume growth response and avoid applying fertiliser where it is not needed.

Predictive relationships to assist fertiliser use decision-making in eucalypt plantations (PNC304-1213)

View the webinar

Read the report.