Light competition affects how tree growth and survival respond to climate

Abstract

1 INTRODUCTION

Climate change may cause a decline in tree growth and survival due to changes in temperature and in water regimes (Lindner et al., 2010; McDowell et al., 2020). In forests, individual trees are directly affected by the local climate, which depends on weather conditions, but they are also indirectly affected through competition from neighbouring trees (Jump et al., 2017; Ruiz-Benito et al., 2013). Climate is likely to influence species competitiveness as well as the individual tree's response to competition, leading to changes in tree dynamics (Clark et al., 2011, 2014). Therefore, studying the effect of competition along climatic gradients is key to understanding how individual trees respond to climate and hence, to better grasp how forest species assemblages and structures will vary with climate change (Magalhães et al., 2021).

Several authors have suggested that the effect of competition varies along abiotic stress gradients and is weaker in stressful environments (Bertness & Callaway, 1994; Craine, 2005; Grime, 1979; Maestre et al., 2009; Tilman, 1980). Grime (1979) argued for a weaker competitive effect in less productive and more stressful environments since, for the tree, conserving energy may be a better strategy than competing for the limited resources available when stress is high. However, soon thereafter, Tilman (1980) emphasised the need to clarify which resources are involved. This is especially important in productive environments, where the abundance of below-ground resources leads to intense asymmetric competition for access to light; on the contrary, in stressful environments, the limited availability of water or nutrients in the soil leads to intense competition for access to the below-ground resources. Later on, the stress gradient hypothesis considered not only competition but also facilitation processes (Bertness & Callaway, 1994; Maestre et al., 2009). These authors put forward the idea that the net competition effect should decrease with increasing abiotic stress due to an increase in the frequency of facilitative interactions. One example of direct facilitation is canopy photoprotection in arid areas: direct exposure to strong sunlight could lead to greater heat and desiccation, and excessive irradiance or UV radiation stress (Demmig-Adams & Adams III, 2006; Valladares & Niinemets, 2008). Another example is the beneficial effect of a dense canopy in cold areas, where the canopy layer protects tree organs from fatally low temperatures by limiting the upward dissipation of heat and by reducing the cooling effect of the wind (Charrier et al., 2015).

So far, studies that have attempted to assess how the effect of competition on tree growth and survival varies along climatic gradients have all concluded that there is a significant and important interaction between climate and competition (Coomes & Allen, 2007; Fernández-de-Uña et al., 2015; Ford et al., 2017; Gómez-Aparicio et al., 2011; Kunstler et al., 2011; Rollinson et al., 2016; Ruiz-Benito et al., 2013; Taccoen et al., 2021). However, they have reported conflicting directions for this interaction, highlighting the need to be more specific about exactly which climatic gradient is being analysed and which resources underpin the competitive interactions. For example, studies in the Mediterranean area have found a greater competition effect in water-limited environments (Gómez-Aparicio et al., 2011; Ruiz-Benito et al., 2013), while studies in temperate regions have observed a greater competition effect in more productive sites where access to light is limited (Ford et al., 2017; Kunstler et al., 2011).

The first step towards disentangling the interactions between climate and competition is to analyse competition for a specific resource, rather than use a generic competition index. Previous studies have used crowding indices as proxies for the competition experienced by an individual tree. These models assume that the more neighbours an individual is surrounded by, and the larger these neighbours are, the more competition it faces. The main drawback of crowding indices is that they aggregate many processes and may be misleading when we are trying to understand the effect of competition for a specific resource along large abiotic stress gradients, where stress factors are likely to vary between bioclimatic zones (Magalhães et al., 2021). Some studies have attempted to distinguish the effects of light competition from those of competition for below-ground resources by using an asymmetric crowding index to represent access to light and a symmetric index to represent access to water and nutrients (Ford et al., 2017). However, asymmetric versus symmetric crowding indices are not process-based and remain poor proxies for competition for a specific resource. We propose to focus on the role of the interaction between climate and light competition. Light competition is known to be a key process in forests and is an important determinant of both forest structure and tree dynamics (Pacala et al., 1996), with tree species strongly varying in their level of shade tolerance (Valladares & Niinemets, 2008) and their sensitivity to light competition (Kunstler et al., 2011). In addition, models for estimating tree light competition are more advanced than models representing competition for other resources, such as water or nutrients, making it possible to study variations in light competition effect at larger scales (Craine & Dybzinski, 2013).

This brings us to the second line of inquiry: using large-scale studies covering broad environmental gradients to further understand forest responses to climate (Ruiz-Benito et al., 2020). Large-scale studies are crucial if we wish to include species' climatic margins, where demographic performance is likely to vary for a given species (Kunstler et al., 2021). Another major advantage of large-scale studies is that they make it possible to compare numerous species, which helps to answer the question of whether the sensitivity to competition varies among species depending on their climatic niche or ecological strategies. The direction and intensity of the climate–competition interaction effect is likely to vary with a species' tolerance to resource limitation (Maestre et al., 2009). For instance, Kunstler et al. (2011) found that the importance of competition for tree growth decreases with increasing productivity along the bioclimatic gradients of temperature and aridity, and that the mean importance of competition is higher for shade-intolerant species than for shade-tolerant species. These results emphasise the importance of not only examining competition-climate interactions for multiple species on a large geographical scale but also comparing the sensitivity of different species to light competition based on their ecological strategies.

Herein, we present a large-scale study of the effect of light competition on individual tree growth and survival across Europe, made possible by the availability of a database of over 1 million trees, including nine European countries from Spain to Scandinavia. Firstly, to analyse the effect of light competition, we derived a tree-level light competition index from the SamsaraLight ray tracing model (Courbaud et al., 2003, 2015), a spatially explicit and tree-based model that estimates the amount of light intercepted by a given tree based on light beam interception and attenuation by the 3D crowns of each tree in the stand. We used species-specific crown allometries to represent the tree crown structure in space. Then, we considered two climatic gradients: temperature and aridity. We used a water balance model based on soil structure, monthly water fluxes and snow melt to derive a plot-level aridity index. Finally, we fitted species-specific tree-based growth and mortality models as a function of climate and light competition. We then used the models to predict annual tree growth and survival under different climates and levels of light competition to test whether the effect of light competition varied along the two climatic gradients, with high aridity in drier climates and low temperatures in colder climates constraining tree dynamics. The main hypothesis is that the net light competitive effect will be dependent on the balance between the negative effect of shading (reduced carbon assimilation) and its positive effects (reduced evaporative demand, reduced frost stress, …). This balance is likely to change along climatic gradients, depending on the relative importance of these climatic stresses. In addition, the benefit of being in full light will not be the same depending on the occurrence of other climatic constraints. The performance improvement should be smaller when water supply is low or when low temperature stress is high. We addressed this question at two different scales: (i) within species—between the climatic margins of a given species, and (ii) among species—by comparing responses for different species with different levels of shade tolerance and different mean climatic niches. We hypothesised that (i) within a given species' climatic niche, the effect of light competition on growth and survival would be weaker at the cold or dry species stress margins, and (ii) among species and across Europe, the species mean sensitivity to light competition on growth and survival would be weaker for shade-tolerant species, and for species whose mean climatic niche is located either in the hot, dry Mediterranean region or in cold boreal regions.