International Journal of Applied Earth Observation and Geoinformation
Diverse responses of different structured forest to drought in Southwest China through remotely sensed data
Graphical abstract
Difference in drought response of forest with various forest height and stock volume. (a) The histogram of forest height frequency in damaged, dead and normal regions. The taller the trees are, the higher the ratio of damaged/dead forest is; (b) the relationship between damaged/dead ratio of forest and its stock volume. The ratio was elevated as the increase in forest stock volume.
Introduction
In the context of global climate change, the frequency, intensity, and duration of drought stress have gradually increased (Cook et al., 2014), which substantially impact the structure and function of forest ecosystems (Choat et al., 2012). Drought not only affects the vigor and growth of trees, but also induces tree death and forest degradation (Choat et al., 2012). The effects of drought on forest ecosystems are multifaceted, and can include decreases in forest productivity (Ciais et al., 2005) and advancement of the dormant period (Xie et al., 2015). Accordingly, trees respond to drought stress by adjusting the growth of their root systems, and by reducing stomatal conductance and leaf area (Nepstad et al., 1994; Delzon and Loustau, 2005; Niu et al., 2014). Forest ecosystem responses to drought and their internal adjusting mechanisms vary according to drought intensity and duration (Niu et al., 2014), which add to the uncertainty of forest responses to drought (Guarín and Taylor, 2005; Nepstad et al., 2007). In addition, human activities such as deforestation and afforestation also increase the variation in forest attributes such as forest age structure. Therefore, to better mitigate and adapt to the impact of human activities and climate change on forest ecosystems, a more accurate evaluation of drought stress in forests with different attributes (such as forest age) is required (Piao et al., 2011; Steinkamp and Hickler, 2015).
Lack of rainfall is the primary cause of droughts, and precipitation is the most direct index for analyzing the characteristics of drought. When evaluating the intensity of drought, multiple indexes for drought assessment have been constructed using precipitation and temperature data. The Standardized Precipitation Index (SPI), mainly based on precipitation data, is easy to calculate and can be utilized to indicate drought at different time scales (McKee et al., 1993). Without precipitation, the temperature increases which leads to water evaporation, aggravating the drought intensity (Sheffield and Wood, 2008). Therefore, the drought indexes that contain temperature information, such as Palmer Drought Severity Index (PDSI) and standardized precipitation evapotranspiration index (SPEI), are more useful for analyzing drought characteristics (Sheffield and Wood, 2008). Due to the time hysteresis (Wu et al., 2015) and the accumulative effect (Huang et al., 2015) of climate responses in plants, it is necessary to apply a meteorological index at multiple time scales for drought evaluation (Li et al., 2015; Huang et al., 2015; Luo et al., 2016). As SPEI integrates both the advantages of PDSI (where the effect of temperature trends and fluctuations in evaporation are considered) and SPI (which is easy to calculate and multiple time scales are considered), it is widely used at global and regional scales (Vicente-Serrano et al., 2010).
The differences in forest attributes constitute another important reason that forests differ in their responses to drought. The age of forests, age-related tree height, and stock volume all significantly affect the water absorption and consumption of trees. Previous research demonstrated that soil water was less available to small trees with shallow roots compared to its availability to large trees with developed root systems, thereby making small trees more sensitive to drought stress (Guarín and Taylor, 2005; Nakagawa et al., 2000). Other studies have revealed that water transportation paths are longer in large trees, that their consumption is higher to maintain respiration, and that evapotranspiration on the leaf surface is more vigorous in large trees, which leads to a higher water demand; thus, drought would appear to have a greater impact on large trees (Nepstad et al., 2007). The forest age is an important reference index in all the attribute factors influencing forest water balance. According to the growth theory of trees, diameter at breast height (DBH), tree height, and stock volume are all related to forest age (Zhang et al., 2014), and these factors are reflected in the photosynthetic and productive activity of forests (Zhou et al., 2013; Zhou et al., 2015). Human activity (such as afforestation) is a significant driving factor in changes to forest age, so revealing the differences in drought responses of forests with various forest ages may facilitate better management to control forest damage risk in future climate change. At the regional scale, the drought response of forest can be affected by the spatial heterogeneity of the meteorological drought index and the intrinsic forest attributes. Thus, the influence of spatial heterogeneity for meteorological drought should be removed to reveal the effect of vegetation attributes (such as forest age) on the response to drought stress (Luo et al., 2016).
Remote sensing technology, which is a high-efficiency modern data acquisition tool, could provide data with high spatial and temporal resolution for forestry research (Assal et al., 2016; Dorman et al., 2013; Xie et al., 2015). The optical properties of green vegetation, which shows strong absorption in the red wave band and strong reflectance in the near infrared wave band, allows for easy calculations of metrics to identify changes in vegetation. Multispectral sensors mounted on different satellites provide a wealth of reflectivity data for studies on the earth's surface. For example, the moderate-resolution imaging spectroradiometer (MODIS), mounted on the satellites of Terra and Aqua, supplies data for various vegetation indices based on the calculation of remote-sensing reflectance, which is widely used in the study of ecology. Of these, the most commonly used index is the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI). Compared with NDVI, EVI is more adapted to study areas with high vegetation coverage at the regional scale (Matsushita et al., 2007). Trees respond to drought stress by reduce leaf area (Nepstad et al., 1994), which could be reflected in the information from satellite observation (Huang et al., 2015).
In this study, by integrating field measurement data, remote sensing data, national forest inventory data, and climate data, we analyzed the drought responses in different forests with various forest ages, stock volumes, and canopy height at local and regional scales to reveal different drought responses in forests of different forest ages. The objectives of this study are to solve two scientific questions: Eliminating the interference of spatial heterogeneity, 1) do differences in forest age significantly affect forests’ response to drought, 2) and is there a consistent pattern over different spatial scales?
Section snippets
Study region
The study region is located in Yunnan Province in southwest China (Fig. 1). The total area is 390,000 km2 (97.51–106.18°E and 21.13–29.25°N) with an annual average temperature of 16 °C, annual precipitation of 1105 mm (Luo et al., 2016). In recent years, several severe drought events occurred in Yunnan province, especially a short drought event in 2005 and a long drought event during 2009–2013, which had substantial effects on the local forest ecosystem (Luo et al., 2016).
According to the 8th
Comparison of the relationship between forest age and EVI deficit under dry and wet conditions
The SPEI data in Yunnan during 2001–2014 demonstrated that the phenomenon of wet and dry alternation was obvious in Yunnan (Luo et al., 2016). There was no drought occurred in the whole province in 2001 (Fig. 4), and the wet condition in the northwestern region where the sampling plots were located was similar. In contrast, a wide range of drought was seen in 2012 with over 80% of the area exhibiting at least mild drought, and the sample located in northwestern region showing moderate drought (
Differences in drought responses of forests with various ages
Drought will affect the function and structure of forest ecosystems (Choat et al., 2012), and the forest age is one of the important attributes of forest ecosystems (Zhou et al., 2015). There are differences in the drought responses of forests with various ages, the effect of which can be reflected by multiple potential mechanisms (McDowell et al., 2008). Some studies demonstrate that soil water is less available to the shallow root system of younger trees, resulting in weaker drought tolerance
Conclusion
Human activities such as afforestation significantly change the age structure of forest ecosystems. Investigating the characteristics of drought responses in forests with different ages is the basis for accurately assessing the effects of future climate change on forest ecosystems. In this study, we used the EVI to calculate the ED, which can show the effects of drought on the growth status of forests. In addition, we analyzed the characteristics of drought responses in forest ecosystems with
Acknowledgments
The authors greatly appreciate constructive comments on this article from Dr. Kaicheng Huang and two anonymous reviewers. We also thank Q. Guo and H. Ren for providing the field measurement data and M. Simard for providing the forest height dataset. This study was supported by the National Natural Science Foundation of China (41571185 and 41621061), the Fundamental Research Funds for the Central University (2015KJJCB33), the New Century Excellent Talents in University (NCET-10-0251), and the
References (49)
- et al.
Spatial and temporal trends of drought effects in a heterogeneous semi-arid forest ecosystem
For. Ecol. Manage.
(2016) - et al.
Age-related decline in stand water use: sap flow and transpiration in a pine forest chronosequence
Agric. For. Meteorol.
(2005) - et al.
Forest performance during two consecutive drought periods: diverging long-term trends and short-term responses along a climatic gradient
For. Ecol. Manage.
(2013) - et al.
Estimating forest canopy height and terrain relief from GLAS waveform metrics
Remote Sens. Environ.
(2010) - et al.
Optical-biophysical relationships of vegetation spectra without background contamination
Remote Sens. Environ.
(2000) - et al.
Drought triggered tree mortality in mixed conifer forests in Yosemite National Park California, USA
For. Ecol. Manage.
(2005) - et al.
Overview of the radiometric and biophysical performance of the MODIS vegetation indices
Remote Sens. Environ.
(2002) - et al.
From space to species: ecological applications for remote sensing
Trends Ecol. Evol.
(2003) - et al.
Plant growth and mortality under climatic extremes: an overview
Environ. Exp. Bot.
(2014) - et al.
The ecology of lianas and their role in forests
Trends Ecol. Evol.
(2002)