Determining deep root water uptake patterns with tree age in the Chinese loess area

Highlights

• Understanding RWU from deep soil is critical for evaluating the sustainability of afforested trees.

• Annual deep soil water below 5 m contributed 9–39% to the total RWU over the tree lifetime.

• Shallow soil contributes 61% of total water uptake, showing a compensated root water uptake.

• Trees use more deep soil water in drier years, but did not shift their water use to shallow soil right after large rainfall.

Abstract

Deep soil water is important for trees to combat droughts and thus is an important consideration for assessing sustainability of afforestation. However, the extent to which trees could depend on deep soil for root water uptake (RWU), remains poorly understood. Here we selected five apple orchards, planted in 2008, 2005, 2001, 1998 and 1994 (named A2008, A2005, A2001, A1998, and A1994, respectively) from the Chinese Loess Plateau and measured water isotopes from tree xylem and soil to the depth up to 23 m. We then used the Bayesian mixing model MixSIAR with dual isotopes (²H, ¹⁸O) to quantify the seasonal contribution ratio of each soil layer (0–0.4 m, 0.4–2 m, 2–5 m, and 5 m to maximum rooting depth) to RWU in normal years 2017, 2018 and wet year 2019. Results showed that with increasing orchard age, rooting depth increased from 10.2 m to 23.2 m, resulting in cumulative deep soil (below 5 m) water deficits from 74.5 mm in 9-year-old orchard to 1191.8 mm in 25-year-old orchard. And annual deep soil water below 5 m contributed 9–39% to the total RWU over the orchard lifetime. Although fine roots in shallow 0–2 m soils in old orchards A1998 and A1994 only accounted for 20% of that in the entire profiles, these roots contributed, on average, 64% of the total absorbed water in 2017–2019. Relative to the normal year, apple trees relied less on deep soil water in wet year. Our findings is of particular significance to the ongoing eco-restoration on the Chinese Loess Plateau (CLP).

Introduction

In the terrestrial ecosystem, vegetation plays important roles in the hydrological cycle, transpiring incident precipitation (Schlesinger and Jasechko, 2014), and altering surface and subsurface water flow patterns (Bond et al., 2002, Huang et al., 2003). In the past decades, a series of revegetation projects were conducted in China, contributing 25% of the global greening during that period (Chen et al., 2019). The “Green for Green” project, which was implanted on the CLP in 1999, is the largest of these projects. Although this project has made pronounced progress increasing vegetation coverage, the revegetated trees have also altered the water cycle, and disrupted the water balance that existed between soil and plants prior to revegetation (Feng et al., 2016, Zhang et al., 2018).

Apple trees have been commonly used as an economical revegetation strategy, where the conversion from cropland to apple orchards not only increase the income of local farmers, but also reduces soil water erosion of the loess hills and existing gullies (Chen et al., 2015, Gao et al., 2018a). Many studies have been conducted to assess the changes in soil water following afforestation (Jia et al., 2017, Wang et al., 2011, Yan et al., 2015). Notably, recent studies (Li et al., 2018a, Li et al., 2019) revealed the connection between deep soil depletion and rooting depth and presented a conceptual model for the interaction between rooting depth and deep soil water depletion: under limited precipitation (150–800 mm year⁻¹), trees growing in a deep unsaturated zone, develop root networks progressively deeper in order to access water in deeper soil. However, little focus has been placed on understanding how changes in deep soil water and rooting depth will impact RWU over the lifetime and over different precipitation years. This is important for understanding how the soil water balance is affected by land use change and afforestation over the long term, which is critical, but poorly understood for assessing the sustainability of afforestation.

RWU patterns have been identified among various vegetation types (Wang et al., 2019, Wu et al., 2016), as well as in mixed-species settings where heterospecifics compete for water (Tang et al., 2018, West et al., 2007). Previous studies with isotopes method found that plants may shift their RWU patterns under different precipitation conditions, and the patterns can be species-dependent (Liu et al., 2019b, Nie et al., 2019). However, few RWU studies have included trees of varying ages located in a variety of different soil water conditions over a multi-year scale.

RWU patterns have been assessed by measurements of water potential (Cook and O'Grady, 2006, Nehemy et al., 2020), root distribution (Song et al., 2020), and/or chemical or isotopic tracers (Wang et al., 2017, Zhang et al., 2017). Water isotopes in particular, have been widely used to trace the movement of water from the soil to the plant xylem (Rothfuss and Javaux, 2017). A critical assumption in the application of water stable isotopes (²H, ¹⁸O) is that no isotope fractionation occurs when plants transport water from the soil to the xylem. This assumption has been convincingly verified recently with a variety of plants (Chen et al., 2020). Furthermore, water isotopes can be used to quantify and partition RWU with respect to both space (shallow soil, deep soil, groundwater, etc.) (Evaristo and McDonnell, 2017, Rossatto et al., 2012) and time (winter precipitation, summer precipitation, etc.) (Allen et al., 2019, Zhang et al., 2017). Relative to neutron probe method (Wang and Wang, 2018, Wang et al., 2015), the advantage of isotopes based method is that they can capture the seasonal water use pattern accurately associated with deep-rooted plants.

Although roots are commonly found at depths of more than 20 m on the CLP (Li et al., 2018a, Yan et al., 2015), the depths of isotopic investigation of most studies are far less than the maximum rooting depth, and are often as shallow as 3 m or less (Gao et al., 2018b, Huo et al., 2018, Tang et al., 2018, Wang et al., 2019, Wang et al., 2017). For example, recent studies concerned with RWU on the Loess Plateau identified rooting depths of 2 m for Jujube trees (Huo et al., 2018), 3 m for Hippophae rhamnoides and Spiraea pubescens (Wang et al., 2019), and 5 m for Robinia pseudoacacia (Zhao et al., 2019). However, the apple trees in this study have much greater rooting depths to the depth of 23.2 m in the old orchard (Li et al., 2018b). Therefore, in the case of apple trees, shallow investigation depths will likely result in an incomplete understanding of the contributions of deep soil water to RWU. Furthermore, as apple trees on the CLP age, their fine roots shift from an exponential distribution to a more uniform distribution with depth, thus increasing the proportion of fine roots located in deep soils (Li et al., 2018b). Whether fine-root water uptake in deep soil influences RWU pattern remains unclear. However, a small quantity of RWU in each deep soil layer certainly has the potential to increase the amount of water available for transpiration.

The objective of this study is to quantify the extent to which afforested apple trees may rely on deep soil water for RWU with the increasing rooting depth in a water-limited area. A Bayesian mixing model (MixSIAR) coupled with dual isotopes was used to determine the seasonal RWU proportion and the contributions of shallow and deep soil water to the transpiration of apple trees of varying age, under varying rooting depth and precipitation. The results of this study are projected to improve our understanding of the water use and drought mitigation strategies of trees facing deep soil water deficits.