Influence of different cultivars–locations on maximum daily shrinkage indicators: Limits to the reference baseline approach

• We registered four olive irrigation treatments in two locations and for 3 or 4 years.
• MDS must be transformed to relate it to other parameters.
• Relationship between MDS and SWP showed two parts, with a threshold around −1.8 MPa.
• MDS is less sensitive to water stress than SWP.
• MDS data presented high variability. A great numbers of sensors are necessary.

The use of biosensors for irrigation scheduling is an interesting way to improve water use efficiency. Trunk diameter variations (TDV) have been studied in several species as a biosensor capable of detecting water stress. The most frequently used parameter for this purpose from TDV is maximum daily shrinkage (MDS). In the present study, MDS was studied in olive trees in two locations in Spain (Badajoz and Ciudad Real), with different cultivars in each one (‘Morisca’ in Badajoz and ‘Cornicabra’ in Ciudad Real), for three years in one location and four years in the other. In order to establish a reference baseline for both locations the relationship between vapour pressure deficit (VPD) and MDS was calculated. Four irrigation treatments were used: one control, irrigated by the FAO method, two in relation to stem water potential threshold and one based on both the trunk growth and MDS from linear variable differential transducers (LVDT). Heteroscedasticity was found in MDS data, so the transformation of the square root (SQR) of the data was necessary be done. The reference baseline with SQR of MDS vs VPD was significantly different between varieties/locations. A local calculation is thus needed to obtain an accurate reference baseline. The relationship between SQR of MDS vs stem water potential (SWP) showed a lineal fit with values higher than −1.8 MPa. In order to ensure that the relationship between SWP and MDS is not due to their relation to VPD, SWP and MDS were divided by their theoretical values in function of VPD. It was observed that the response of MDS to water stress is much weaker than SWP and in this range of SWP (higher of −1.8 MPa) variations of MDS are only due to VPD. The number of sensors necessary for precise measurement of MDS was, on average, more than thirty.