Planted forests are forests formed by planting and/or intentional seeding of native or introduced tree species (FAO 2005). Planted forests have become an important part of terrestrial ecosystems, and bring considerable economic and ecological benefits (Stevens et al. 2015, Chandler and McGraw 2017, Spicer et al. 2020). Meanwhile, planted forests have also been recognized as an effective way to mitigate climate change (Cheng et al. 2023). China has been implementing afforestation and reforestation projects since the 1970s, establishing more than a quarter of the world's planted forest area (FAO 2021). The results of China's eighth consecutive national forest resources inventory (2009–2013) show that the planted forests cover an area of 69.33 million hectares, accounting for 36% of the country's forested land. Compared with natural forests, the planted forests tend to have fewer species, younger stands, and a higher forest management intensity (Rollinson et al. 2020, Cheng et al. 2023). As a result, the forest ecosystems dominated by the planted forests are prone to a series of ecological problems, especially affecting the healthy growth of understory vegetation (Fig. 1). Understory vegetation is an important part of forest ecosystems (Scott et al. 2000, Chandler and McGraw 2017). Although understory vegetation accounts for a small proportion of the total biomass of the forest ecosystem, it plays an important role in maintaining forest diversity and conserving water and soil (Bohlman 2015, Bell 2016, Benes and Carpenter 2016; Zhao et al. 2016). Increased shrub and herb coverage or diversity may benefit various species groups, such as birds and invertebrates (Díaz 2006). Therefore, the growth of understory vegetation is of great significance to the sustainable management of the ecosystems in the planted forest.
Some studies have suggested that understory vegetation does not grow due to allelopathy, that is, coniferous forests secrete tannins that inhibit the growth of understory vegetation (e.g. Benes and Carpenter 2016). However, other studies have reported that the excessive canopy density of artificial coniferous forests hinders the growth of understory vegetation (Scott et al. 2000, Ligot et al. 2014, Benes and Carpenter 2016, Chandler and McGraw 2017). Canopy density is a key factor that affects the microclimatic like light transmission and forest precipitation, and it has an important impact on the development of understory vegetation (e.g. Bell et al. 2016). However, the relationship between the canopy density and the understory vegetation is very variable and depending in climate conditions, and the thresholds of the relationship have not been clearly determined. Previous studies have been performed using different thinning intensities or retention densities to analyze the relationship between the canopy density and the understory vegetation (Li et al. 2009, Chandler and McGraw 2017). It is difficult to ensure the continuous change of the control variables in the quantitative control experiments, which leads to different results in different studies, thus making it difficult to obtain an accurate threshold of canopy density (Li et al. 2009, Stevens et al. 2015, Cao et al. 2018). In addition, understory vegetation is possibly affected not only by the canopy density but also by forest age, topography, and soil fertility (Ligot et al. 2014, Stevens et al. 2015, Spicer et al. 2020). Researchers have used correlation analysis and regression analysis to study the relationship between canopy density and understory vegetation, for example, Zhao et al. (2016) found that the diversity of understory plants increase with the decreasing canopy density by the regression analysis. These statistical methods obtain only the average relationship between the variables; however, factors other than canopy density could influence understory, and the statistical results mixed the effect of other factors may elicit the wrong result (Laiho et al. 2014, Chandler and McGraw 2017). Additionally, the result of correlation analysis represents a monotonic relationship, but the relationships between ecological variables are nonlinear and have thresholds (Bohlman 2015). For example, canopy density of Masson pine (Pinus massoniana) that is too low (< 0.2) or too high (> 0.8) is not conducive to soil conservation, and a canopy density of 0.6 has the best effect in terms of preventing soil erosion (Zhao et al. 2016, Cao et al. 2018, Li et al. 2020).
The constraint line provides a useful method to analyze the complex nonlinear relationships between ecological variables based on massive data (Hao et al., 2017). The concept of the constraint line was first proposed and used to represent the best state of population growth (Holling 1987, Guo and Rundel 1998, Hao et al. 2017). Due to the interaction of multiple factors in a complex ecological process, the two variables studied in the real situation commonly show scattered point clouds, but the ecological processes represented by different parts of the cloud-shaped data are not the same and contain rich ecological information (Thomson et al. 1996, Cade and Noon 2003). When the supply of all factors, except the measurement factor, is at a free level, it is considered an ideal situation. In this case, only the measurement factor has a restrictive effect on the related relationship, which could be represented by the upper constraint lines (Evanylo and Sumner 1987, Schnug et al. 2008). Therefore, the upper constraint line represents the boundary at which the response variable is primarily limited by the constraint variable, rather than by other factors (Guo and Rundel 1998). In contrast, the lower constraint line represents the change trend of the response factor when the other impacting factors are at their worst levels (Cade and Noon 2003). The mean line represents the general relationship between the measurement factor and the response factor in the real situation.
The selection of planted forest species varies according to the region, climate and other conditions (Li et al. 2009). In the hilly areas of southern China, planted forest species are selected from Chinese fir (Cunninghamia lanceolata), cypress (Cupressus funebris), Masson pine (Pinus massoniana), and eucalyptus (Eucalyptus sp), etc.; in the plains of northern China, poplar (Populus) and paulownia (Paulownia fortunei) are the main species; in the mountains of northeastern China, larch (Larix gmelinii) is the main species (Cao et al. 2018, Li et al. 2020). In the early 1990s, to speed up the progress of wasteland eradication, aerial seeding and high-density artificial afforestation were adopted (Tan et al. 2016). Several dense forests have been formed with an average density of 3000 trees per hectare (Li et al. 2009, Cao et al. 2018). Most of the planted forests are pure forests with sparse understory vegetation and single biodiversity metrics, resulting in serious understory ecological problems (Fig. 1), such as severe soil erosion and nonpoint source pollution (Paletto and Tosi 2009, Li et al. 2009, Cao et al. 2018, Li et al. 2020). Especially, under the planted forests of pine and cypress, because pine needles are hard and unpalatable, shrubs and herbs do not grow well in the forest understory, resulting in poor stability of the entire ecosystem, which is not conducive to the sustainability of the regional ecosystem (Li et al. 2009).
In this study, Hubei Province, located in the southern China, was chosen as the study area. The three planted tree species (Chinese fir, cypress, and Masson pine), commonly found in the hilly areas of southern China, were selected to analyze the effects of canopy density on the understory vegetation, meanwhile, the thresholds of canopy density were explored. The analysis results of this paper would provide suggestions for the plantation density and management of the planted forest in the hilly areas of southern China. The research methods and ideas in this paper could also be used in the decision analysis of optimal canopy density for the planted forests in China and around the world. Specifically, this paper has the following two objectives: 1) explore the relationships between the canopy density and the indicators of understory vegetation by using the mean lines, upper constraint lines, and lower constraint lines; 2) reveal the optimal canopy density values of the three planted tree species that are conducive to the growth of understory shrubs and herbs.