Better prediction of Mediterranean olive production using pollen-based models

Abstract

Olive oil is a major economic resource of the Mediterranean region. Olive crop management can be improved by models that forecast the variable reproductive biology of olive tree. However, the processes controlling olive harvest are complex on large scales. Here, we study the parameters that influence olive fruit production for developing accurate forecasting models. Seventeen aerobiological sampling points have monitored olive pollen grains in Spain, Italy and Tunisia from 1993 to 2012. Six crop models have been developed at two provinces and country scales. The modelling has been done in two steps: (1) typification and (2) modelling by partial least square regression. Results show that higher pollen indexes and water availability during spring are related to an increase of final fruit production in all the studied area. Higher pollen indexes are also positively correlated with air temperature during early spring and autumn. Furthermore, a decrease of fruit production is related with increasing air temperature during winter and summer. To conclude, we have designed accurate models that allow accurate predictions of olive production.

Forecasting olive crop yields based on long-term aerobiological data series and bioclimatic conditions for the southern Iberian Peninsula

“…Summer evapotranspiration is another important variable that has an influence on crop yield. An increase in evapotranspiration during the dry period, when the water need is at its maximum, can significantly reduce the quantity, and also the quality, of the olive tree fruit (Ribeiro et al, 2008;Oteros et al, 2013).…”

Multi-model ensemble simulations of olive pollen distribution in Europe in 2014: current status and outlook

“…A similar error was found for Andalusia (Galán et al, 2005) and two sites (Perugia and Ascoli Piceno) in Italy (Frenguelli et al, 1989) -4.8 and 4.33 days of the standard error, respectively. A recent study (Aguilera et al, 2014) constructed three independent statistical models for Spain, Italy, and Tunisia and ended up with over 5 days of standard error for the Mediterranean. In another study, the authors admitted the scale of the challenges: "The specific moment for the onset of the olive heat accumulation period is difficult to determine and has essentially remained unknown" (Aguilera et al, 2013).…”

“…One of the candidate processes is the chilling-sum accumulation suggested by some studies, e.g. Aguilera et al (2014). A switch to different types of phenological models with genetic differentiation of the populations following Chuine and Belmonte (2004) is another promising option.…”

“…Several wider-area studies were also undertaken to aim at more general statistical characteristics of the season, e.g. Aguilera et al (2014Aguilera et al ( , 2013, Galan et al (2016).…”

Response of maize and olive to climate change under the semi-arid conditions of southern Spain

“…To evaluate the olive flowering dates, different models have been published based on the pollen density (Galan et al, 2005;Oteros et al, 2014) or in phenological observations (Orlandi et al, 2005). There are some models that calculate different processes of the olive productivity, as the photosynthesis, transpiration or water use efficiency (Diaz-Espejo et al, 2002Connor and Fereres, 2005).…”