Impacts of Downed Dead Wood Poplar Trees on Forest Regeneration in the Semi-Arid Region of Northern China

Abstract

In the past few decades, due to rising temperatures and changes in precipitation, the degree of drought in semi-arid areas has increased, leading to a large number of tree deaths and threatening the natural forests distributed in the semi-arid areas of North China. This article takes the forest ecosystem of Saihanwula Nature Reserve in the southern section of Greater Khingan Mountains in China’s semi-arid region as a research area and studies the distribution of downed dead wood and its impact on forest renewal in the area. We used the sample plot survey method, investigated the number of downed dead wood, decay class, dumping direction, existence form, and the number of regenerated seedlings in the sample plot, and calculated the density of regenerated seedlings in different plots. The renewal density is 4050 ± 824, 2950 ± 265, plants/ha, and 2625 ± 237 plants/ha, respectively, in the sample plots for Later-death plot, Mid-death plot, and Early-death plot. The average storage of downed dead wood in Saihanwula Nature Reserve is 58.51 ± 16.56 m³/ha. The distribution densities of downed dead wood are 50 ± 21, 806 ± 198, 189 ± 76, and 22 ± 5 plants/ha for decay classes II, III, IV, and V respectively. The main form of downed dead wood in the research area is “trunk base fracture”, accounting for 68.78% of the total number of downed dead wood. A large number of downed dead wood had serious negative effects, such as crushing and injuring the regeneration seedlings and other plants under the forest at the moment of dumping and for a long time after dumping. The crushed and injured rate is 5.3~7.8%, with downed dead wood accumulated in the forest from the early stage of downed dead wood. It had negative effects on the regeneration of seeds, seedlings, and young trees, such as obstructing and hiding the light from the soil surface and inhibiting the regeneration and growth of seedlings. However, after the trees were dumped, large gaps appeared in the forest, increasing the sunlight area on the soil surface. In the later stage of tree death, moderately high decayed downed dead wood changed the soil structure in terms of soil softness, water holding capacity, and nutrient content, thus promoting the growth of seedlings and young trees. Reasonably utilizing the relationship between downed dead wood and forest renewal can effectively promote the healthy development of forests.

1. Introduction

Forest renewal is a natural biological process of natural forest reproduction in the dynamics and balance of forest ecosystems [1]. It is not only a process of replacing old forest stands with young ones, but also an important means to complete this process [2,3]. With the shift towards a multi-objective management mode in forest management, researchers are also diversifying their thinking on the elements that affect forest renewal. This study takes woody debris (dead woody substances present in forest ecosystems in the form of dead standing trees, fallen branches, downed dead wood, etc., called woody debris (WD) [4,5]) from different stages of tree death as the research object. Previous studies have shown that forest renewal is often influenced by external factors, such as light, soil, moisture, and temperature [6,7]. For example, litter and soil thickness are key factors affecting seedling regeneration in typical forest communities of ecological public welfare forests in the middle section of Zhongtiao Mountain [8]. The impact of forest disturbance on natural forest renewal in tropical changing environments were investigated, and factors such as light availability, soil conditions, climate, nutrient availability, competition, herbivorous animals, and seed predation affected forest renewal [9].

Previous research on downed dead wood has mainly focused on the relationship between downed dead wood and natural regeneration and succession [10,11]. In recent years, extensive research has been conducted on the dynamic changes and ecological effects of downed dead wood and other wood debris after large-scale abnormal death in semi-arid forests [12,13]. Logging can affect the diversity of forest plants [14], while moderate logging can promote forest renewal [15]. Fungal activity of wood decomposers and moss interacted with the regeneration of spruce seedlings on rough wood debris [16]. In recent years, a large number of forest deaths have occurred in research areas in northern China [17,18,19], leading to the accumulation of a large amount of downed dead wood in the forest. The impact of this downed dead wood on forest development (whether the forest is declining or becomes derivative) requires observation of the relationship between the different stages of downed dead wood death and forest renewal and exploration of the impact of the characteristics of downed dead wood at different stages on forest renewal.

It is interesting to note that different stages of downed dead wood have different inhibitory and promoting effects on forest renewal, which contradicts previous research findings [20]. In order to gain a deeper understanding of the impact of downed dead wood on forest renewal, we designed experiments and conducted research in the Saihanwula Nature Reserve in the southern section of the Greater Khingan Mountains in Inner Mongolia. By investigating the storage, spatial distribution pattern, existence form, decay class, regeneration density, and squeezing effect of downed dead wood on seedlings, we explored the basic characteristics of downed dead wood and their impact on forest renewal.

2. Materials and Methods

2.1. Research Area

The research area is a poplar forest observation area of the Saihanwula Forest Ecosystem National Positioning Observation and Research Station for the semi-arid region of northern China, which is located in Balin Right Banner, Chifeng City (longitude 118°18′~118°55′ E, latitude 43°59′~44°27′ N, hereinafter referred to as the “Saihanwula Nature Reserve”). It is a transitional area from East Asian broad-leaved forests to cold temperate coniferous forests, as well as a transitional area from grasslands to forests [19]. The study area belongs to the Zhongshan Mountains of the Alshan Branch of the southern end of the Daxing’anling Mountains and has a northeast–southwest slope direction and an average elevation of 1000 m. Its climate belongs to the mid-temperate continental monsoon climate zone, with abundant surface freshwater resources. Due to its unique geographical and natural conditions, it has formed a unique plant community and is one of the few mountainous comprehensive nature reserves in China that mainly protects ecosystems, such as forests, grasslands, and wetlands, species diversity, and important water source conservation areas at the source of the western Liaohe River [21]. The climate conditions are semi-arid and semi-humid, with strong winds and little rain in the spring and hot and rainy summers, with mostly southerly winds; and with rapid cooling in autumn and dry and cold winters, with mostly northerly winds [22]. The research area has an average annual precipitation of 400 mm and an annual potential evaporation of 2050 mm, making it a typical semi-arid area. The main forest vegetation is the secondary forest of poplar and birch. The sample plot for this experiment was selected within the poplar forest, accompanied by trees such as Betula platyphylla Suk. and Quercus mongolica Fisch. ex Ledeb. Shrubs included Spiraea pubescens Turcz. and Ostropsis davidiana Decaise, and herbal plants mainly included Carexpeidefermis and Sanguisorba officinalis L. (Figure 1).

2.2. Experimental Design

According to the duration and degree of tree death, three types of sampling plots were set up: Early-death plot (meaning that within the forest plot, Populus davidiana grew normally with a small amount of fallen or dead trees), Mid-death plot (the forest consisting of slightly withered, dying, or already dead trees), and Later-death plot (most of the trees have already died and formed a large number of downed dead wood), representing different death degree gradients. The sample plot area is 20 m $\times$ 20 m, with 3 replicates set for each plot type for a total of 9 experimental plots (

Table 1. Basic information of test plots.

Plot	Longitude	Latitude	Altitude (m)	Aspect	Slope (°)	Canopy Closure	Downed Dead Wood Diameter
E1	118°43′47.19″	44°12′59.36″	1217.2	WN	12°	0.83	12.83 ± 3.05
E2	118°43′47.39″	44°12′58.30″	1227.6	WN	10°	0.76	12.24 ± 1.15
E3	118°43′46.27″	44°12′56.90″	1237.2	WN	13°	0.60	10.69 ± 1.98
M1	118°43′46.33″	44°13′00.37″	1212.8	WN	16°	0.68	9.32 ± 2.89
M2	118°43′48.10″	44°12′57.74″	1231.8	WN	19°	0.51	8.75 ± 3.46
M3	118°43′47.38″	44°12′57.20″	1234.1	WN	12°	0.78	8.91 ± 1.93
L1	118°44′08.53″	44°12′56.46″	1227.2	WN	18°	0.65	10.64 ± 3.59
L2	118°44′10.55″	44°12′56.65″	1232.9	WN	17°	0.56	9.69 ± 3.41
L3	118°44′07.52″	44°12′55.62″	1239.2	WN	23°	0.66	9.84 ± 2.32

Note: “WN” means a northwest slope. “E” is Early-death plot, “M” is Mid-death plot, “L” is Later-death plot. “1, 2, 3” are replicates.

).

2.3. Field Survey

2.3.1. Field Investigation Methods

①

The downed dead wood reserves were surveyed and recorded according to the standards of international downed dead wood investigation and research: each downed dead wood with a diameter ≥ 10 cm in the sample plot was surveyed. We used a breast diameter ruler to measure the diameter of the big and small heads (the big head was not more than 10 cm away from the roots, and the small head was not more than 10 cm from the crown of the tree). The diameters of the large and small heads of the tree were measured using a diameter gauge. A laser rangefinder was used to measure the length of downed dead wood, and we measured the X and Y coordinates of the wood within the sample plot, where X represents the east–west coordinates and Y represents the north–south coordinates.

②

The existing forms of downed dead wood were investigated and divided into five forms: uprooted (the downed dead wood is pulled out entirely from soil, with its roots attached to it), break in the middle of the trunk, break in the base of the trunk (also called trunk base fracture), tree segments, and root piles (

Table 2. The existing forms of downed dead wood.

The Existence of Downed Dead Wood	Features
Uproot	Uprooted death wood
Trunk base fracture	The height of the broken pole < 1 m
Break in the middle of trunk	The height of the broken pole ≥ 1 m
Root pile	Standing dead trees, length < 1 m
Tree segment	Large dead branches and dead wood without head and tail, length > 1 m

, Figure 2, orange font).

③

Investigation of the decay class of downed dead wood was based on a five-level classification (from I to V, indicating slightly decomposed to heavily decomposed), referring to suitable indicators used in other studies [23]. The conditions of organisms attached to the downed dead wood, seedling growth, and root invasion were combined to determine the class, and a small knife was inserted into the downed dead wood debris to test its mechanical hardness (

Table 3. Downed dead wood decay classification.

Features	Decay class
	I	II	III	IV	V
Leaves	Exist	None	None	None	None
Branches	All twigs exist	Big branch exists	Big, thick branches exist	The branches have fallen off, and the trunk is still there	None
Bark	Exist	Exist	Mostly exist	Mostly fallen off	None
Backbone shape	Round	Round	Round	Round to oval	Oval to flat
Invaded by the root	None	None	Sapwood area	Invade all	Invade all
Plant growth	None	Little plant growth	Few shrub seedlings and moss	Large area of moss	Bush moss and big tree

, Figure 2).

④

The forest renewal density and regeneration mode or approach were investigated, and the species and quantity of seedlings were surveyed using the full sample plot method [24]. The soil cutting method was used to determine and count the renewal mode of the new seedlings. The surface soil was cut open to directly observe whether there was a connection point between the mother plant root system and the young plant root system. A young plant connected to the mother plant root system in a “┴” shape is called a clone plant (asexual reproductive ramet), while a young plant without a connection point with the mother plant root system and with its main root growing vertically into the soil is called a seedling plant (sexual reproductive ramet). Finally, by calculating the proportion of seedlings with different renewal modes through survey data, the proportion of asexual reproduction can be used to infer the vitality of the root system of withered trees.

2.3.2. Indoor Testing and Data Processing

The wood reserve for downed dead wood is the volume of the truncated stem body and branches and is calculated by the following formula:

$$\mathrm{V}={\displaystyle \frac{\mathsf{\pi }({\mathrm{d}}_{\mathrm{r}}^2+{\mathrm{d}}_{\mathrm{R}}^2)\mathrm{L}}{8}}$$

(1)

where V represents the volume of downed dead wood and its branches; π is the circumference; d_r is the diameter of the small head end of any stem body or branch; d_R is the diameter of the large end; and L is the length of woody debris. The reserve volumes of all the downed dead wood in a plot was summed to give the total reserve in the plot, and the reserve per unit area was then known.

2.3.3. Data Analysis

Microsoft Excel 2016 was used for preliminary organization of the field survey data and for calculation and analysis of the woody debris reserves, regeneration seedling density, and woody debris density at different decay classes in 9 plots. IBM SPSS 25.0 was used for statistical variance analysis, i.e., performing the one-way, ANOVA, and Duncan variance analyses on the woody debris reserves in different plots, different decay classes of woody debris reserves, and density of regeneration seedlings. SigmaPlot 14.0 software was used to draw and graph. The boxplot was used to analyze the dispersion of regeneration seedlings in different death stages of the plot and to examine whether there is a correlation between different plots. We used Auto CAD 2016 software to draw a spatial distribution map of downed dead wood using the method of direct coordination. The significance of differences in any variable, such as the regeneration density between two plots, was tested by Student’s t-test method, with the p threshold value being 0.05.

3. Results

3.1. Distribution Characteristics of Downed Dead Wood in the Poplar Forest of Saihanwula Nature Reserve

3.1.1. Downed Dead Wood Reserves

There are significant differences in the reserves of downed dead wood among different plots (Figure 3). The reserve at the Later-death plot is 79.11 ± 23.56 m³/ha, the reserve at the Mid-death plot is 53.72 ± 13.78 m³/ha, and the reserve at the Early-death plot is 42.74 ± 12.36 m³/ha, showing a reserve pattern of Later-death plot > Mid-death plot > Early-death plot. The average reserve of downed dead wood in the Saihanwula Nature Reserve poplar forest is 58.51 ± 16.56 m³/ha.

3.1.2. Spatial Distribution Pattern of Downed Dead Wood

In the Early-death plot, 47% of the downed dead wood tilted towards the northeast direction (the direction pointed by the thin head of the downed dead wood is the direction of the downed dead wood), and in the Mid-death plot, 55% of the downed dead wood tilted towards the northeast. In the Later-death plot, 25.5% of the downed dead wood tilted towards the northeast and 34.8% towards the southeast (Figure 4). Apart from slope, wind is one of the biggest factors affecting the direction of downed dead wood. The direction of downed dead wood in the poplar forest of Saihanwula Nature Reserve is mostly northeast and southeast, indicating that the Saihanwula Nature Reserve was mainly blown by southwest and northwest winds.

3.1.3. Existing Form of Downed Dead Wood

The main existing form of downed dead wood in the Saihanwula Nature Reserve poplar forest is the “trunk base fracture” (

Table 4. The dumping modes of downed dead wood in the Saihanwula Nature Reserve forest.

Plot	Uproot	Percentage	Trunk Base Fracture	Percentage	Break in the Middle of Trunk	Percentage	Tree Segment	Percentage
Later-death plot	26 ± 5	6.38%	281 ± 15	69.04%	96 ± 7	23.58%	4 ± 1.2	0.98%
Mid-death plot	13 ± 2	3.96%	209 ± 12	63.71%	104 ± 6	31.71%	2 ± 0.5	0.61%
Early-death plot	17 ± 4	4.91%	245 ± 13	70.81%	83 ± 11	23.98%	1 ± 0.3	0.28%

); about 70.81% of the downed dead wood in the Early-death plot were in the form of trunk base fracture. Secondly, the break in the middle of trunk in the Mid-death plot of downed dead wood accounted for approximately 31.71%; The proportion of dead wood in the Later-death plot in the form of uproot and tree segments is higher compared to the Early- and Mid-death plots, accounting for approximately 6.38% and 0.98% of the total number of downed dead wood, respectively.

The decay class of downed dead wood in the poplar forest in Saihanwula Nature Reserve is mainly class III (Figure 5), with an average density of 806 ± 198 trees/ha. Next is class IV, with an average of 189 ± 88 plants/ha, and a few existed in classes II and V, with an average of 50 ± 21 and 22 ± 5 plants/ha. From the decay class of downed dead wood, it can be seen that the growth decline or tree death of the Populus sylvestris forest in Saihanwula Nature Reserve is relatively common, and the death time is relatively long, leading to the accumulation of a large number of class-III decomposed downed dead wood in the forest.

3.1.4. The Main Factors Contributing to the Death of a Large Number of Trees

According to the rainfall and temperature data over the past 60 years, although the annual precipitation did not have an overall trend over 60 years, it decreased during 1997–2010 compared to other periods (Figure 6; with the precipitation in most years of 1997–2010 being lower than the long-term regression line). The annual mean air temperature had a clear increasing trend over 60 years and especially increased strongly before 2010. These changes in climate condition would have led to a large area of abnormal forest death in the poplar sylvestris ecosystem of Saihanwula Nature Reserve. The extremely dry and warm years of 1997–2010 were about 14 to 27 years ago, which is the reason for the high reserve of downed dead wood in the forest and the highest reserve of downed dead wood in decay class III, accounting for 75.5% of the total proportion of downed dead wood reserves.

3.2. Regeneration Characteristics of Poplar Forest

3.2.1. Forest Regeneration Density after Death

The seedlings are 162 ± 33 at the Later-death plot, 118 ± 11 at the Mid-death plot, and 105 ± 10 at the Early-death plot, as investigated by the soil dissection method. The regenerated seedlings in the sample plots are all in the form of budding, and sprouting regeneration is the main regeneration approach or mode in the study area, with certain spatial differences. The renewal density is 4050 ± 824 plants/ha, 2950 ± 265 plants/ha, and 2625 ± 237 plants/ha for the Later-death, Mid-death, and Early-death plots, respectively (Figure 7). The study plots with different death gradients showed significant differences in renewal density. The renewal density of the Later-death plot was higher than that of the Mid-death plot and Early-death plot, indicating that the forest had a rapid renewal rate after death.

3.2.2. Impact of Downed Dead Wood on Forest Regeneration

The average number of seedling trees that were crushed to death by the downed dead wood reached 175 plants/ha at the study area, accounting for 10.9% of the total seedlings of the sample plots. The average number of seedlings that were damaged or hurt by the downed dead wood reached 33.3 plants/ha, accounting for 1.9% of the total seedlings (Figure 8). Especially, the crushing-to-death and the damage in the Later-death plot reached 216.6 and 316.6 plants/ha, accounting for 8.1% and 18.25% of the sample plot seedlings, respectively. The crushed impact was medium in the Mid-death plot, with the Early-death plot being relatively the least impacted.

4. Discussion

4.1. A Large Number of Downed Dead Wood Occurred in the Poplar Forest after Abnormal Death

There are many studies on the large-scale forest death in semi-arid areas of northern China [25,26,27], but there is little research on the impact of forest dead woody debris on forest renewal. Some research has shown that the complete decay time of downed poplar dead wood is approximately between 15 and 35 years [28]. From Figure 6, we know that in the 1990s and 2000s, severe droughts occurred in North China [29], leading to tree death, which was about 20 years ago. Since the study area is in the national nature reserve, it is prohibited to cut and clear dead trees and conduct other external interference [30]; therefore, dead trees showing different classes of decay have been accumulating in the forest without clearance. The dead trees experienced decomposition and eventually fell down under various influences, such as strong wind, rainstorm, and animal impact [31]. During this period, the crushing impact and damage from dead trees to surrounding plants or seedlings, the blocking and interception to the soil surface, and other negative effects restricted the regeneration and growth of plants and delayed the regeneration time to the late-death stage. Thus, the reserves of downed dead wood in the Later-death plot were significantly higher than those in the normal succession area (the Early-death plot), and the number of regeneration seedlings was also higher than that in the Early-death and Mid-death plots. There are quite a few forests in the reserve that have died in large areas. The long-term decay of downed dead wood can improve the growth pattern of the forest. However, this situation exacerbates a negative impact on forest fires. Therefore, the amount of downed dead wood is crucial for the health of forests.

4.2. Inhibition of Forest Renewal by Downed Dead Wood in the Secondary Forest

Most of the downed dead wood in the study area is dumped in the northeast and southeast directions. The monsoon wind direction in the survey area is perennially southwest and northwest [32,33], especially in the non-growing season. As a result, the dumped direction of downed dead wood (northeast and southeast directions) in our plots matched to the wind direction. Therefore, for the regenerated seedlings growing in the northeast and southeast positions of the dead trees, the probability of being crushed or injured is higher. The direction of downed dead wood can be different due to differences in forest types, tree species, and terrain conditions [34,35], and these conditions often become key factors affecting forest renewal, restoration, and sustainable development. Downed dead wood mainly existed in the form of “trunk base fracture”, with relatively few trees fallen in forms of “break in the middle of trunk” and “uprooting” (Figure 2, Table 4). Usually, the tree died and formed a dead standing tree, which continuously decomposed and later mechanically broke under the influence of strong winds in spring, becoming downed dead wood. In summary, the accumulation and distribution characteristics of downed dead wood in the Saihanwula Nature Reserve forest area are mainly marked by the existence of large areas of dead or abnormal stands, which are somewhat different from other natural succession forests without such a tree death situation. For the management of downed dead wood, the negative impact caused by their abnormal accumulation should also be considered.

4.3. Secondary Forests Have the Ability to Quickly Renew after Death

The regenerated seedlings in the sample plots were all grown via sprouting regeneration. There was germination regeneration for Populus in the survey area, which may be related to the limited soil seed bank. Related studies have shown that when the seeds are inputted to the surface of the soil in Sanhanwula Nature Reserve forest, only about 28% to 30% enter the surface to form a soil seed bank, while the probability of germination and seedling development is even smaller [36,37]. The seed vitality (or survival rate) of poplar or birch is not strong when the external environmental conditions are harsh [38]. Therefore, the contribution of the forest itself to the soil seed bank is relatively small, and its contribution to regeneration is relatively small, as well. The regeneration approach after the death of the secondary forest is mainly through sprouting regeneration. In terms of regeneration ability, it is manifested as being stronger at the Later-death plot than the Mid-death plot and the Early-death plot. The high regeneration density of the Later-death plot (4050 ± 824 plants/ha) indicates that downed dead wood has a promoting effect on forest renewal. The reason why the regeneration density of the Later-death plot is higher than that of the Mid-death plot and the Early-death plot may be because after a large area of forest death happens, although the aboveground part has already withered, the underground root system has not yet died. At the same time, the withering of the tree crown brings more sunlight to directly stimulate the dormant buds of the root system, forming a large number of clones. With the passage of death time, the amount of regeneration increases. In addition, the decomposition of downed dead wood in the later stage enriches soil nutrient elements and promotes seedling regeneration [39,40,41]. Therefore, the forest of Saihanwula Nature Reserve has the ability to quickly regenerate to compensate for tree death after the death, and this research finding is comparable to some previous studies [21]. Moreover, the number of downed dead wood is large, the forest renewal is fast, and the renewal density is high. Downed dead wood has a promoting effect on forest renewal, which is consistent with previous research results [42].

5. Conclusions

The reserve of downed dead wood in the poplar forest of Saihanwula Nature Reserve in the semi-arid region of North China is 58.51 ± 16.56 m³/ha. The direction of downed dead wood is mainly influenced by northwest and southeast winds. The main form of existence is “trunk base fracture”, accounting for 68.8% of the total number of downed dead wood. Its decay class is mainly class III, and the renewal approach is sprouting renewal. Although the formation of downed dead wood has some negative effects on seedling regeneration, such as crushing and injuring, the downed dead wood has the ability to promote forest renewal in the later stages of death. Taking all factors into consideration, a large amount of dead wood accumulates in the forest and can only be utilized after a long period of decay. Moreover, a large amount of high decay class downed dead wood also poses potential risks to forest fire prevention and other aspects. Therefore, appropriate measures should be taken to manage dead wood, providing a basic or favorable condition for maintaining the healthy and sustainable development of the forest eco-system.