Introduction

Tree growth is an important factor in both fiber production and carbon storage, and the rate of growth will affect how quickly carbon can be stored or fiber can be produced. The relationship between tree growth and environmental factors is complicated, but drought is one of the most important as it is the main way that moisture can limit growth and one of growing concern as the rate of occurrences increases due to climate change. Many studies into the mechanisms of drought related mortality and growth limitation have been recently conducted, but the effect of drought over longer periods of time is not fully understood. Therefore, the goal of this study is to further define the multi-year effect of drought on tree growth across Canada.

Trees have a variety of requirements for growth, namely: light (Porter, 1999), water (Chaves, et al., 2002), heat (Wahid, Gelani, Ashraf, & Foolad, 2007), and nutrients (LeBauer & Treseder, 2008). In particular, water plays a key role in the absorption of nutrients, photosynthesis, and the transport of sugars and hormones (Pfautsch, Hölttä, & Mencuccini, 2015). Soil moisture becoming too low can create embolisms in xylem tissues, reducing a tree’s capacity to transport water (Venturas, Sperry, & Hacke, 2017). Thus, the relationship between a tree’s growth and drought can be affected by a tree’s resistance to embolism and how quickly the stomata close in response to a lack of water (Nadal-Sala, et al., 2021). The response can vary greatly between genera, species, and even individuals (Hacke, Sperry, Wheeler, & Castro, 2006). This study will attempt to separate the effect of these differences by presenting species separately. Additionally, there is evidence that suggests trees within the same species will have varying adaptations based on their environment (Serrano & Peñuelas, 2005). Canada has a wide variety of climatic conditions (Majorowicz & Skinner, 1997), so results will be separated by ecozone as well.

In a 2011 study, Lloret, Keeling, & Sala proposed indices related to various components of resilience. These are characterized by the growth of a tree before, during, and/or after an event. This allows for easier characterization of the effect of drought on growth and allows for comparison between species and sites. Resistance (Rt) shows the reduction in performance during an event and corresponds with the ratio of growth during the event to growth after the disturbance. Recovery (Rc) is the capacity to recover relative to the damage experienced during the disturbance and is calculated as a ratio of growth after the event to growth while it was in effect. Finally, Resilience (Rs) is the ability to reach pre-disturbance levels of performance and is characterized as the ratio between growth after the event and before it took place. This study will use these indices to represent the impact of drought.

It is well known that tree growth varies with age (Matsushita, et al., 2015). There are also many other causes of long-term trends in tree growth. In this study, the main factor of interest is local variances, so long-term trends should be removed from consideration. A common practice to remove these trends is called spline smoothing, where a line is drawn over a full time series and then the dependent variable is replotted relative to this line. There are many models for this spline, but Friedman’s algorithm is better than many for looking at short term variance (Roosen & Hastie, 1994), which was used in this study.