Potential of plurispecific agrosystems: the case of agroforestry
Agroforestry systems (AFS), as the intentional combination on one plot of trees and/or shrubs with crops or livestock, represent a good opportunity for agrosystem redesign (Dupraz and Liagre, 2008) and has been recognized as a sustainable agricultural practice for half a century (Garrity, 2012). Beneficial outcomes of agroforestry include reduced nutrient and pesticide runoff (Davis et al., 2012), increased biocontrol (Gliessman, 1985), improved soil quality, erosion control, carbon sequestration (Cardinael et al., 2017) and alleviate hazards linked to extreme climatic events (Leakey, 2014).
AFS, like other multi-species systems, are agrosystems designed to maximise resources (light, water, nutrients) usage in time and space by maximising positive interactions and minimising negative interactions (Jose et al., 2004). Therefore, using different indicators such as the land equivalent ratio (LER), it is possible to compare mono-specific agrosystems and multi-specific ones. Nevertheless, if we aim to achieve higher yields in an agroforestry system, we need to understand what interactions take place and how they will influence plants growth. It is possible to achieve higher yields in an agroforestry system if the interspecific competition is lower that then intraspecific competition (Gliessman, 1985). However, the relations between the different components of the agrosystems will be modified as the plants are aging. To apprehend these complex systems, we can compartment interactions in aboveground and belowground interactions.
Impact of aboveground interactions among plants
Interactions in AFS depend on the species and the disposition of the trees. The interactions intensity will be different if the trees are on the border of the field or inter-cropped (Jose et al., 2004). The most noticeable aboveground interaction is the competition for light between the species (Jose et al., 2004) which causes plants to experience a modification in light quantity and quality because of light absorption by surrounding vegetation. The chlorophyll of neighbouring plants filters out the red (600–700 nm) and blue (400– 500 nm) wavelengths of the sunlight while reflecting and transmitting most of the far-red (FR) wavelengths (700– 800 nm). In response to a drop in the red to far red ratio (R:FR) (Vandenbussche et al., 2005), two major strategies have been recognized for maximizing fitness under shaded or partially shaded conditions (Henry and Aarssen, 1997; Gommers et al., 2013). The first strategy, known in literature as the shade avoidance syndrome, consists in maximizing light interception through morphological and phenological traits that contributes to space occupation (Ballaré et al., 1997). It includes traits as stem and petiole elongation, increased specific leaf area (ratio of leaf area to leaf dry weight), apical dominance, hyponasty, reduced branching and accelerated flowering (Smith and Whitelam, 1997; Foulkes et al., 2010). The second strategy found in shade tolerant species aims at maximizing net carbon fixation through shade-adapted leaf physiology (Givnish, 1988). While shade and non-shade species optimize light capture and utilization through what is known as the carbon gain hypothesis (increased specific leaf area, increased photosystem II:I ratios and lower chlorophyll a:b ratios) (Givnish, 1988; Valladares and Niinemets, 2008), shade-tolerant species suppress shade avoidance traits (Niinemets and Valladares, 2004). However, shade adaptation and its effect on plant development have been mainly studied on annual plants and in a controlled environment and little is known of perennials response to shade (Matsubara, 2018; Maron, 2019).
Still, aboveground negative interactions can be limited by diminishing the density of trees, their disposition, precocity or growth period (Chirko et al., 1996). However, the shade provided by the trees can also benefit the annual crops by limiting excessive summer radiations as stated before (Lin, 2007, 2011; Quinkenstein et al., 2009). Trees will also impact the microclimate that can benefit the shaded plants depending on the climate. For example, the row of trees can act as a windbreaker that will affect the evapotranspiration demand and therefore improve the water use efficiency (Quinkenstein et al., 2009). But a humid microclimate is also going to favour cryptogamic diseases (Gliessman, 1985) which can be a problem for apple trees because of Venturia inaequalis the pathogen responsible for apple scab if the selected cultivar is not resistant. The introduction of trees will also create new ecological niches by modifying the landscape that can offer new habitat that can contribute to increase the number and the diversification of natural enemies and pest (Jose et al., 2004; Quinkenstein et al., 2009).
Impact of belowground interactions among plants
Belowground interactions between perennial and annual plants depend of the spatial location of their roots. In an ideal situation where the roots of each different species are present in different compartments of the soil, competition will be less important than where there are in the same (Gliessman, 1985; Cardinael et al., 2017).
Perennial plants usually have the majority of their fine roots in the first thirty centimetres of the soil and so are in competition with the annual crop (Jose et al., 2006). However, most of the trees used in agroforestry have deep roots that will explore, if the depth of soil allows it, horizons of soils inaccessible to the annual plants (Rowe et al., 1998; Jose et al., 2001). Studies have shown that in agroforestry tree roots colonized deeper soil layers and were more vertically oriented (Cardinael et al., 2015). Thereby, tree’s roots can act as an interception net for the leached nutrients (Allen et al., 2004). These nutrients will then be available to the annual plants after decomposition of the litter in the case of deciduous trees and if the leaves are left on the plot. In the same way, trees will also be able to absorb nutrients coming from the bedrock alteration (Schroth, 1995).
Deep roots could also act as a hydraulic lift if the top horizons are dryer than the bottom (Caldwell et al., 1998; Jose et al., 2004). If the quantity of water moved by this phenomenon is important enough, it could limit competition for water in mixed species systems. Even small amount can have a positive impact such as (i) making available nutrients that are not in a dry soil, (ii) facilitate root exploration and (iii) keeping roots active in a temporary dry soils and allow a quick recovery of activity (Pierret et al., 2016). Furthermore, roots exploration can be improved thanks to the pores created by the tree’s roots on one hand and biological activity improving soil structure on the other hand (Hulugalle and Lal, 1986).
Usually, tree roots occupy every soil horizon and thus are in competition with other plants for water and nutrients when they become a limiting factor. Even if this interaction can favour the separation of root systems (Pierret et al., 2016), yield will be negatively impacted (Smith et al., 1999) as the trees develop especially when they are still young (Parker and Meyer, 1996). A study has shown that walnut tree shallow roots in agroforestry grew mainly during the spring-summer period which could increase the competition with other plant (Germon et al., 2016). Some species will also exudate allelochemicals in the rhizosphere that can harm the annual crop (Rizvi et al., 1999). For example, apple trees have been reported to be sensitive to juglone, the phenolic compound that is the agent of Juglans spp. allelopathy (Galusha 1870; McWhorter et al., 1874) cited in (Jose, 2011). Soil under 10-year-old black walnut trees (Juglans nigra) alley cropping system can have significant amounts of juglone if release rates are greater than the abiotic and microbial transformation rates (von Kiparski et al., 2007). However, the concentration of juglone drops significantly with distance from the walnut tree row (Jose and Gillespie, 1998a) and the highest concentrations of juglone measured do not exceed the concentration inhibition threshold of crops typically considered for intercropping (Jose and Gillespie, 1998b). More recent studies showed that there are several processes that can be altered by lower concentration of juglone which can limit water and nutrient uptake (Hejl and Koster, 2004; Böhm et al., 2006). In the light of this knowledge, it is safe to hypothesise that apple trees planted at 6 metres or more of walnut trees are probably not influenced by walnut allelopathic effect, but it could be a confounded factor for interpreting the effect on apple trees planted near walnut trees.
Prototyping fruit tree based agroforestry: the case of an apple orchard
Our study was developed on apples that is one of the most important fruit production of temperate climate and whose architecture and functioning has been extensively studied in conventional orchards (Volk, 2017). Apples are among the oldest and most important fruit crops in the world (Harris et al., 2002). They have been cultivated since ancient times, in fact, archaeological studies have shown that they were cultivated already in 1000 BC (Juniper et al., 1998).
Cultivated apples are a result of extensive ancient hybridization of various species of the genus Malus Mill., a member of the Rosaceae Juss. family, subfamily Pomoideae (pome fruits) (Jackson, 2003; Webster, 2005). Over hundred botanical names have been published for the cultivated apple (Qian et al., 2010; Cornille et al., 2012, 2014), however, Malus domestica Borkh. is now the correct binomial nomenclature for the cultivated apple (Qian et al., 2010). Some morphological characteristics shared by apple cultivars in the world are: woolly pubescence on young stems and on the lower surface of the leaves, dull green leaves, elliptic-ovate in shape, with irregularly saw toothed margins, woolly pubescence on flower stalks and calyx, and pome fruits indented at the base with persistent calyx (Webster, 2005).
While interesting to reduce the negative impact of conventional orchard management agroforestry systems designed around fruit trees in temperate climate are poorly developed and studied and usually put the fruit tree in the upper strata (Lauri and Simon 2019) as typically illustrated in mixed fruit tree and vegetable farms that combine fruit trees and market gardening (Paut et al., 2021). In tropical climate agroforestry designed around fruit trees (e.g. coffee and cocoa which are shade-adapted species) is a common practice. A recent study on coffee shows that light use efficiency increases with shade leaving net primary productivity fairly stable across all shade levels (Charbonnier et al., 2017). Furthermore, shade has been proposed as a solution to improve tree water status and water use efficiency during drought periods (Nicolás et al., 2005; Girona et al., 2012). While light interception has been reported to be a primary factor to fruit yield (Palmer et al., 2002) some studies have shown that under a moderate water stress net shading improved yield in apple (Lopez et al., 2018). The benefit of shade was multi-factorial, it improved the tree water status, delayed fruit maturity hence giving more time for fruit growth and reduced photo-inhibition. Therefore, there is an incentive to study increasingly complex woody plant combination in temperate climate by combining timber trees, fruit trees and annual crops or shrubs on different strata (Lovell et al. 2018; Lauri et al. 2019).
Apple tree based agroforestry systems (AT-AFS) could be an interesting solution to reduce pesticide use since current high-density monoclonal orchards are usually highly susceptible to pests and diseases and, therefore, dependent on pesticides (Simon et al., 2017). In addition to the aforementioned interests, an apple-based agroforestry systems could also be of interest in the Mediterranean area to limit the adverse effects of recurrent excessive summer radiation (light and temperature) which are responsible for annual field losses (Racsko and Schrader, 2012) and increase water use efficiency (Mupambi et al., 2018). Indeed, these regions are regularly exposed to high solar irradiance and dry climate, even if competition for light and nutrients can affect fruit tree growth, the main expected benefits are related to the mitigation of microclimatic stresses (Lauri, Mézière, et al., 2016). However, the inherent complexity of agroforestry systems is the primary hurdle to achieving their potential benefits. To optimize an agroforestry system, apart from selecting species with no allelopathic effects or strong interspecific competition, it is necessary to study them extensively to draw temporal and spatial assembling rules (Gliessman, 1985).
In our experiment apple trees are fertilized and irrigated according to the organic farming recommendations. As a result, we expect that aboveground competition (i.e. competition for light) will most probably be the limiting factor for the apple trees and we focused on the impact of shade of apple trees architecture, morphology, phenology and functioning.
Nowadays, apple trees are almost exclusively compound trees consisting of a scion grafted on a rootstock (Jackson, 2003; Tromp, Webster and Wertheim, 2005). Rootstocks are used to avoid juvenility, to control vegetative growth, to promote flower-bud formation, to improve cropping efficiency and quality of the fruits, and in some cases to provide winter hardiness and provide resistance or tolerance to some telluric diseases (Tromp, Webster and Wertheim, 2005; Hanke et al., 2007).
The scion is the productive part of the tree that bears the different buds and two different type of shoots: (i) vegetative shoots and (ii) reproductive shoots composed of a flower cluster and one or more bourse shoots. All shoots are indeterminate in growth (Tromp, Webster and Wertheim, 2005) and emerge from buds which have the potential to produce leaf primordia only (vegetative shoots) or both leaf and flower primordia (reproductive shoots:
Figure 1: Apple flower (fruit) bud in diagrammatic longitudinal section, showing foliar appendages and flower buds (Abbot, 1970).
Figure 1). Flower buds are found terminally on all types of shoots and terminally or axillary on long shoots after vegetative growth has stopped (Jackson, 2003; Tromp, Webster and Wertheim, 2005). Floral induction refers to the change of from the vegetative phase to the reproductive phase and occurs mainly in early summer, but it can be extended to early autumn under certain conditions. Floral initiation begins when the meristem flattens and continues as primordial sepals, petals, stamens and pistils form centripetally on the apex and grow into fully formed appendages (Pratt, 1988). Although flower induction can be inhibited by heavy cropping, some cultivars being notorious for their ‘biennial bearing’ habit, environmental factors also affect induction and initiation (i.e. solar radiation).
How shade can affect apple tree
Apple tree cultivars have all been selected and studied under optimal conditions, and their acclimation to different degrees of shade has mainly been studied under shade nets (Zibordi et al., 2009; Morandi et al., 2011; Bastías and Corelli-Grappadelli, 2012; Lopez et al., 2018). While an alteration of leaf morpho-physiological traits (i.e. palisade thickness, stomatal aperture, and chlorophyll content) and an increased elongation is expected (Bastías et al., 2012), little is known of the other architectural traits (i.e. number of ramifications, bud types) that apple trees will express in natural and fluctuating shade produced by upper trees and their adaptation to a changing environment. An important reduction in light intensity and at critical timing can affect apple production at different development stages leading to a decrease in fruit quantity and quality. The first negative consequence of shade on fruit production is the inhibition of floral initiation (Corelli-Grappadelli, 2003). Floral initiation is under the control of diverse environmental stimuli such as temperature and photoperiod and endogenous factors. The reason has not yet been fully elucidated as why shade reduces flower-bud initiation (Corelli-Grappadelli, 2003) but five genetically defined pathways have been identified that control flowering among which the photoperiod pathway refers to a regulation of flowering in response to day length and quality of light perceived (Srikanth and Schmid, 2011). The effects of shade nets on fruit growth development gave different results depending on climate and cultivars in relation to a reduction in light availability. In South Africa, for example, 20% shade nets reduced fruit growth for “Royal Gala” and “Cripp’s Pink” (Gindaba and Midgley, 2005) and increase fruit growth for “Fuji” (Smit, 2007). Studies in Spain concluded that 20% shade nets did not affect fruit growth in “Mondial Gala” (Iglesias and Alegre, 2006). Furthermore, a reduction of light intensity in the period from 15 to 30 days after full bloom may greatly reduce fruit set (Byers et al., 1985). During early stages of fruit growth, a decrease in photosynthesis and tree carbon assimilation (Zibordi et al., 2009) can reduce fruit growth rates and induce fruit drops (McArtney et al., 2004). Light availability can also affect fruit growth by affecting/changing carbohydrate partitioning between sinks (i.e. fruit and shoots). Shoots in full sun light are able to export photo-assimilates to fruit three weeks after full bloom while similar export for shaded shoots is reached only five weeks AFB for 70% of the shoots (Corelli-Grappadelli, 2003), suggesting that under shade shoot growth has priority over the fruit for photo-assimilate (Bepete and Lakso, 1998). Light quality also impacts fruit development, while shade has been reported to reduce fruit growth (Morandi et al., 2011) another study reports an increase of maximal fruit growth up to 20% under blue shade nets that reduced the R:FR ratio and increased in the Blue:Red ratio (Bastías et al., 2012).
Tree architecture as a framework to analyse adaptation to agroforestry system
Tree architecture is a discipline of botany that was developed by Hallé (Halle and Oldeman, 1970). Combining four main criteria (i.e. growth and branching process, morphological differentiation of axes and position of reproductive organs) 23 architectural models were established (Barthélémy and Caraglio, 2007) each characterized by a unique combination of modalities off the four criteria. Apple tree combines features of two architectural models, Rauh and Scarrone. Branches are monopodial with lateral flowering in Rauh and sympodial with terminal flowering in Scarrone. Apple trees combines both lateral and terminal flowering (Lauri and Laurens, 2005; Costes et al., 2006). Based on the observation of tree shape and branching Lespinasse proposed ideotypes to describe the ’bearing habit’ of apple trees by combining the branching pattern and the distribution of fruiting (Lauri and Laurens, 2005). However, the variability of apple tree growth and fruiting patterns in each ideotype proved the limit of this approach so more detailed studies were developed on the behaviour of individual fruit-bearing shoots of various lengths with the objective to characterize the contribution of the fruit bearing shoot and the branch in bearing regularity (Lauri et al., 1995, 1997) which led to three main results. First, cultivars are differentiated by the frequency of ‘extinction’ (i.e. axillary flower clusters whose bourse-shoots die) which showed that tree architecture also results from interactions between growing and non-growing organs (Lauri et al., 2009). Second, cultivars can be deciphered by the frequency of bourse-over-bourse formation and third there is a positive correlation between bourse-over-bourse frequency and extinction. Upscaling these findings from the branch to the whole tree, three main strategies have been identified: (i) high bourse-over-bourse frequency indicates the ability to have a regular bearing pattern, (ii) low bourse-over-bourse could be related to irregular bearing if there is a synchronisation in shoots and (iii) lower bourse-over-bourse could be related to regular bearing if shoots are desynchronized (Lauri et al., 1995, 1997; Lauri and Laurens, 2005). However, the actual bearing pattern of apple trees results both from its endogenous potential and the way it reacts to the training and pruning procedures (Breen, 2016).
Table of contents :
Chapter 1: Introduction
1. New stakes for agriculture
1.2. New paradigms for apple farming systems
2. Potential of plurispecific agrosystems: the case of agroforestry
2.1. Impact of aboveground interactions among plants
2.2. Impact of belowground interactions among plants
3. Prototyping fruit tree based agroforestry: the case of an apple orchard
3.1. Apple tree
3.2. How shade can affect apple tree
3.3. Tree architecture as a framework to analyse adaptation to AFS
5. Materials and methods
5.1. Study site
5.2. Light quantification
5.3. Apple tree traits
5.4. Sap flow measurement
Chapter 2: A critical assessment of neighbourhood crowding index: application study in an agroforestry system of timber and fruit tree
2. Materials and methods
2.1. Study site and plant material
2.2. Apple tree traits
2.3. NCI Computation
2.4. Light quantification
2.5. Data analysis
3.1. Estimation of PAR under the canopy of walnut tree
3.2. Comparison between PARTLS, PARHP and NCI
4.1. Comparison of PARHP and PARTLS methodologies
4.2. Agrosystem management as key to control above- and belowground interactions
Chapter 3: Apple tree adaptation to shade in a fruit tree based agroforestry system
2. Materials and methods
2.1. Study site
2.2. Light quantification
2.3. Data collection
2.4. Data analysis
3.1. Shade adaptation traits
3.2. Floral initiation
4.1. Changes in morphology, architecture and phenology
4.2. Changes in the reproductive strategy
Chapter 4: Can agroforestry improve apple trees water use? – An essay combining environmental variables and sap flow
2. Materials and methods
Chapter 5: Discussion
1. Overview and main results of the thesis
1.1. Light management is essential in AT-AFS
1.2. Impact of shade on apple trees morphology, architecture and phenology
1.3. Shade as a tool to limit transpiration during summer and to buffer extreme heat
1.4. Potential of AT-AFS
2. Limits of the study