Introducing CLM-FruitTree to model carbon allocation in fruit orchards with the Community Land Model

Carbon allocation is a major driver of plant growth and plays a key role in shaping ecosystem processes and the global carbon (C) cycle. In contrast to annual crops, fruit trees store and remobilize C in their perennial plant components, have long canopy durations, relatively low respiratory costs, and remain productive for decades. To predict C dynamics in fruit tree orchards under global change, it is essential to expand the understanding of carbon allocation in fruit trees and to improve its representation in comprehensive modelling environments such as land surface models (LSMs). LSMs simulate the exchanges of matter and energy between the terrestrial biosphere and the atmosphere. They are widely used in C cycle and climate change studies, and typically include representations of various types of natural vegetation and annual crops. Despite the importance of fruit orchards in regions that are strongly affected by climate change, such as the Mediterranean, they are rarely considered in LSMs, thus leaving an important gap in the representation of C allocation and related biogeophysical and biogeochemical processes of these agro-ecosystems. In this work, we present the new fruit tree sub-model CLM-FruitTree within the Community Land Model version 5 (CLM5). Herein, a fruit tree is described by a perennial deciduous phenology with C allocation to standing woody biomass components and annual organs such as leaves, fine roots, and fruits that are either shed or harvested within the yearly cycle. Two different pools, the storage and the photosynthetic pool, contribute to tree growth while C allocation to the individual plant components is based on allocation coefficients that vary depending on the specific phenological phase. CLM-FruitTree was tested using multiple years of field measurements of above- and belowground biomass components, leaf area index (LAI), yield, soil respiration, and eddy covariance (EC) data from an apple orchard in South Tyrol, Italy. We found that biomass allocation was captured within 1-5 % of the measured values, with about half of the assimilated C allocated to fruits. Growth from C storage thereby played a significant role in shaping initial leaf development and growth of fine roots. Simulated ecosystem C fluxes showed a high correlation (r > 0.84) with the EC measurements and the seasonal dynamics were well represented. Average annual gross primary productivity was predicted within 1.5 % of the measured values while net carbon uptake was overestimated by on average 21 % mostly due to an underestimation of soil respiration in the orchard caused by necessary simplifications in the microbial respiration, orchard structure, and management practices. Overall, the new sub-model CLM-FruitTree allows the exploration of the dynamics of C allocation and fluxes in fruit orchards, and may advance C cycle and climate change studies of such agro-ecosystems at larger scale.