Improving the apple carbon balance model MaluSim by integrating and testing water deficit and fruit drop effects

Abstract

Apple (Malus x domestica Borkh.) is the major tree fruit in Germany, but there still exist major uncertainties in commercial apple production. Thinning has been practiced for many years and is an essential part for successful apple production, but is still an unpredictable part with large variations from year to year and even within years. Another uncertainty for apple growers is the need of apple trees for adequate water supply. Due to the advancing climate change apple growers are uncertain whether additional irrigation will be necessary during the establishment phase of young apple trees.

Therefore, to facilitate these decisions in future, the motivation of this study was to acquire a universal apple model to be used in Germany and Central-Europe. The apple carbon-balance model MaluSim which has been developed in the USA was chosen as a basis. At first, the model has been parameterized for German growing conditions and first simulation runs were conducted. In comparison to the original model, the seasonal assimilate production for German standard trees was about half of the one for standard trees in the USA, while standard trees in Germany were only about one third of the size of US ones.

Subsequently, the previously untested fruit growth and abscission submodel has been modified and tested. Simulated final fruit numbers were compared to counted fruit numbers of unthinned trials in Zornheim (GER), Jork (GER), Lindau (CH), Güttingen (CH) and Wädenswil (CH). The modification ‘G4-4’ showed good simulation results with very low deviations to actual fruit numbers. Average deviation to fruit numbers was only 3.7 %. Results indicate that the model is able to very adequately calculate final fruit numbers of natural fruit drop of Central-European apple trees and that it could be a tool used for thinning advices.

Based on published data, a new water submodel has been included into the MaluSim framework to additionally improve the model. Midday stem water potential has been used as an indicator for water status of the tree. A water stress effect on long shoot growth, respiration, and photosynthesis based on midday stem water potential was used in the water submodel. First simulation runs using the new water submodel indicate that general behavior is realistic. Modeling results generally corresponded to findings of published experimental results. The submodel is able to simulate variable intensities of water deficit effects at variable times.

To study the effects of additional irrigation and water deficits on young apple orchards a field experiment was conducted during 2012 and 2013 in Geisenheim, Germany. Trees of the cultivars ‘Fresco’ (‘Wellant®’), ‘Jugala’, and ‘AW 106’ (‘Sapora®’) were planted in autumn 2011 in a randomized plot design. In 2012, three irrigation treatments, CT (control treatment, only rain fed), NT (normal treatment, irrigated with 2 L/tree/day), and ET (evapotranspiration treatment, irrigation based on calculated water balance), were applied and effects on tree growth and physiology were recorded in 2012 and 2013. Besides soil moisture, water status of the trees has been recorded using predawn and midday stem water potential measurements.

In both years, amounts of precipitation were higher than the 30-year average for the site, with 2013 being wetter than 2012. Between treatments, statistically significant differences developed in both seasons, but were less profound in 2013. In 2012, vegetative growth recorded in NT and ET was statistically significantly higher than in CT for all cultivars. In the wetter season 2013, similar tendencies evolved. But only in ‘Jugala’ and ‘Fresco’ long shoot growth and wood surface area was significantly higher, while for ‘AW 106’ differences in long shoot growth were not statistically significant. Similar results were obtained for the first harvest in 2013.

Light interception and calculated leaf area, measured in ‘Fresco’, was significantly lower in CT trees (2012: 17.4 %; 2013: 24.9 %) compared to ET (2012: 26.3 %; 2013: 42.1 %) and NT (2012: 24.2 %; 2013: 38.3 %), in both years. Additionally, differences between CT and the two irrigated treatments were found for leaf gas exchange rates during the seasons. A relationship between midday stem water potential and leaf gas exchange has been confirmed. Grouping of photosynthesis measurements according to midday stem water potentials led to statistically significant differences between groups. Photosynthesis started to decrease between -1 and -1.5 MPa.

In conclusion, the results show that although both years have even been wetter than usual, the additional irrigation was highly beneficial for tree growth and had a positive effect on orchard establishment. This effect is supposed to be considerably higher in drier years.