Zhigang Zhao · Wei Shen · Chunsheng Wang ·Hongyan Jia · Jie Zeng

Abstract Erythrophleum fordii Oliv. (Caesalpiniaceae) is a rosewood species naturally distributed in southeast Asia and south China, and the heartwood is commonly used for high-quality crafts and furniture. While there are dif ferences in heartwood development among single trees with distinct social status, the relationship between heartwood development and growth performance remains unclear. This information is essential to improve plantation management for high-yield heartwood production. Forty dominant, intermediate and suppressed trees were sampled from E. fordii plantations aged in 32–35 years in Pingxiang City, Guangxi,China. Stem analysis was carried out to determine horizontal and vertical variations of heartwood and sapwood. Number of annual rings, diameter and area of heartwood and sapwood as well as ratios of heartwood diameter and area at breast height were all signif icantly inf luenced by the social status of trees in stands ( P < 0.05). In these mid-aged plantations, E. fordii stems developed heartwood once the xylem diameter reached 5–10 cm, and then heartwood diameter and area increased with increasing xylem diameter. Heartwood ring numbers, diameter and area as well as their ratios decreased with increasing height, while sapwood ring numbers and diameter were relatively constant within the section where heartwood occurred. Heartwood and sapwood diameters were equal at heights of approximately 6-m for dominant, 5-m for intermediate and 3-m for suppressed trees. Dominant trees dif fered considerably from intermediate and suppressed trees in heartwood volume, while heartwood volume ratios were all below 30%, and near 90%in the stem section below 8-m height regardless of social status. Relationship analysis showed that DBH was the most important factor inf luencing heartwood in even-aged stands.The f indings provide evidence for crop tree selection, thinning regimes and reasonable management of plantations of E. fordii.

Keywords Erythrophleum fordii · Heartwood formation ·Spatial variation · Tree dif ferentiation Project funding: This study is supported by the Fundamental Research Funds for the Central Non-prof it Research Institution of CAF (Grant No. CAFYBB2017MB019).

Introduction

Most tree species show dif ferentiation of heartwood and sapwood in mature stems (Pinto et al. 2004; Wang et al. 2010;Lehnebach et al. 2017), and heartwood normally has better durability and aesthetic values than sapwood (Taylor et al.2002; DeBell and Lachenbruch 2009; Moya and Berrocal 2010). In particular, heartwood has greater value in furniture making and decoration, and its formation and development are given more attention (Yang and Hazenberg 1991; Kerr 1998; Woeste 2002)

Heartwood formation is a process which initiates at the base of the stem and gradually forms upwards from central to outside (Taylor et al. 2002). Its percentage relative to the total stem wood declines gradually from the base to a certain height, after which no heartwood is present (Kerr 1998; Knapic et al. 2006). In several studies, sapwood width has been observed as almost constant up to a certain age(Moraisa and Pereira 2007; Miranda et al. 2009; Zhao et al.2015). This is concluded from observations at breast height or at the stem base, and may be improved when considering the entire tree since heartwood development is inadequate in upward parts (Moraisa and Pereira 2007). Therefore, studies on horizontal and vertical distribution of heartwood and sapwood are important for the determination of heartwood yield and relevant forest management practices (Climent et al. 2003; Knapic et al. 2006).

It is more reliable to identify the inf luencing factors on heartwood yields when the heartwood diameter is larger than the sapwood width at breast height (Sellin 1994; Zhao et al. 2015). In this phase, trees in plantations have been highly competitive for resources, leading to inevitable tree dif ferentiation. All sorts of trees with dif ferent social status vary in their resource allocation. Dominant trees devote more resources to branches and leaves than intermediate and suppressed ones, and intermediate trees direct more into stems than dominant ones (Li et al. 2015 ). It is still uncertain whether the dif ference in resource allocation results in variation of heartwood formation or not (Taylor et al. 2002).Some studies have shown that sapwood width of suppressed trees was smaller than that of dominant trees, while their heartwood ratio was higher (Sellin 1991 and 1994; Ojansuu and Maltamo 1995). In contrast, other studies indicated that fast-growing trees with dominant status in stands had larger heartwood ratios (Miranda et al. 2009; Kokutse et al. 2010;Liu et al. 2014). The dif ference between these studies may be related to tree properties, stand density and site, and should be considered in forest management (Yang and Hazenberg 1992; Taylor et al. 2002).

Erythrophleum fordiiis an endangered species with a distribution in Vietnam, Laos and southern China. It is a longrotation hardwood species with harvest age up to 60 years in the tropics and warm sub-tropics. As a traditional rosewood species, its heartwood is utilized for high-quality crafts and furniture making (Zhao et al. 2009). In our previous study, we reported that heartwood formation was initiated at 10 years old, and heartwood diameter and sapwood width at the base of the stem were almost equivalent at 20 years.At this point, heartwood formation went into a stable phase(Zhao et al. 2015), and it is thus suitable to conduct a comparative analysis on heartwood formation. Horizontal and vertical distributions of heartwood and their difference between trees with dif ferent social status were investigated in mid-aged plantations. The objectives were to assess the properties of heartwood formation, to evaluate dif ferences in heartwood development between single trees with distinct social status, and to reveal the relationship between heartwood development and growth performance of the species.The f indings will be important for crop tree selection and for thinning regimes to make full use of site potential for greater benef it in a sustainable and economic way.

Materials and methods

Study site

The study was carried out at the Experimental Centre of Tropical Forestry, Chinese Academy of Forestry in Pingxiang City (22°03′–07′ N, 106°48′–56′ E), Guangxi Zhuang Autonomous Region, P. R. China, in the natural distribution range ofErythrophleum fordiiand also its main plantation area in southern China. This region is in the north tropical monsoon climate zone with an annual mean temperature of 22 °C and rainfall of 1550 mm.TheE. fordiiplantations were established downslope or mid-slope in 1979, 1981 and 1982. The initial spacing was 2 m × 3 m (1667 trees ha −1 ). Two types of thinning were carried out at an intensity of 30% and 40% when 11- and 21-year old, respectively.

Sampling and stem analysis

Ten 20 m × 30 m sample plots were established in these plantations in 2014, and diameter at breast height (DBH),total height, height to crown base and crown diameter were recorded (Table 1). Social status (Kraft class) was determined for each tree in each plot according to the widely adopted system described by Nicholas et al. ( 1991). As the stands had been thinned twice and there were no dying or dead trees, the trees were divided into three classes: dominant, intermediate and suppressed. The growth properties of each class are shown in Table 2.In each plot, two dominant, one intermediate and one suppressed tree were selected for stem analysis, for a total of 40 sample trees. North was marked on each from the base of the stem to 1.5 m before felling, and then extended to top after felling. Discs 5-cm thick were taken at ground level, 0.3 m,1.0 m and 1.3 m and then every 2-m to the top.

Data collection and analysis

The disc surfaces were polished smoothly by angle grinder with a particle size of 100 mesh. Annual rings, widths of xylem, sapwood and heartwood of each disc were measuredin the four radial directions (north, south, east and west).Areas of xylem and heartwood were calculated using the mean value of their widths assuming their cross-section as circular (Zhao et al. 2015).

Table 1 Information of the sampled plots in Erythrophleum fordii plantations

The volume of xylem or heartwood (V) for each stem segment was calculated using the equation below. The sum of all segments provided a whole tree estimate:

where,Vis volume of xylem or heartwood.S1 andS2 are xylem or heartwood areas of upward and downward sections, andLthe length of a segment. Sapwood volume was calculated as xylem volume minus heartwood volume. Oneway analysis of variance (ANOVA) and Tukey’s multiple range test (P< 0.05) were performed to detect dif ferences in heartwood and sapwood distribution among tree social status using SPSS software 11.5.

Results

Variations in stem cross section

There were signif icant dif ferences (P< 0.05) forErythrophleum fordiitrees in all heartwood and sapwood parameters at breast height among classes of trees social status(Table 3). The heartwood ring number, sapwood width,and heartwood diameter ratio at breast height of dominant and intermediate trees did not dif fer signif icantly and were remarkably larger than those of suppressed trees. Conversely, the highest sapwood ring number was observed in suppressed trees. Heartwood diameter and area as well as sapwood area were dif ferent between paired classes of tree social status. Dominant trees were 98.7%, 98.2% and 134.6%greater than suppressed trees in these parameters, respectively. The ratios of heartwood area in all sampled trees at breast height were less than 50% regardless of their social status. Dominant trees (0.42) showed higher heartwood area ratios than suppressed trees (0.30) at breast height, both classes did not dif fer signif icantly from intermediate trees(0.37).

The relationship between heartwood parameters and xylem diameter at any height suggested that heartwood could be observed when xylem diameter reached 5–10 cm (Fig. 1).Heartwood diameter and area increased with increasing xylem diameter in a signif icant linear relationship (Fig. 1 A), and in a binomial relationship for heartwood area and xylem diameter (Fig. 1 B). Compared to dominant and intermediate trees,heartwood appeared in suppressed trees when the xylem diameter was much smaller. The ratios of heartwood diameter and area increased sharply with xylem diameter increment from 5 cm to 25 cm, and remained almost constant (Fig. 1 C, D).However, the variations of heartwood diameter and area ratios for suppressed trees were much larger than those for dominant and intermediate trees.

Table 2 Description of the dominant, intermediate and suppressed Erythrophleum fordii trees

Table 3 Attributes of heartwood and sapwood at breast height of dominant, intermediate and suppressed Erythrophleum fordii trees

Fig. 1 Relationship between heartwood parameters and xylem diameter for mid-aged Erythrophleum fordii trees

Fig. 2 Comparison of heartwood height and its ratio to whole height among dominant, intermediate and suppressed trees of Erythrophleum fordii. No same letters referred to signif icant dif ference between classes of social status at the level of 0.05, uppercase for heartwood height, and lowercase for its ratio

Fig. 3 Heartwood and sapwood variations along stems of dominant ( a, d, g), intermediate ( b, e, h) and suppressed ( c, f, i) Erythrophleum fordii trees

Variation in longitudinal stem sections

Heartwood height

Dominant and intermediate trees did not dif fer signif icantly in heartwood height, both showing significantly higher heartwood height than suppressed trees (Fig. 2). Average heartwood heights were 13.1(± 2.0) m, 11. 8(± 2.7) m and 9.0(± 3.8) m for dominant, intermediate and suppressed trees, respectively. While obvious dif ferences were absent in heartwood height ratios among these social status classes,the ratios were above 75% for dominant and intermediate trees, and above 60% for suppressed trees.

Heartwood ring

Heartwood ring number declined from base to top of the stem, while sapwood ring number f irst slightly increased up to 12 m for dominant trees, 10 m for intermediate trees and 6 m for suppressed trees, and then declined with increasing tree height (Fig. 3 A–C). In the base section, the ring number was larger in heartwood than in sapwood; and dominant and intermediate trees dif fered little from each other in heartwood ring number and its ratio, which were much higher than suppressed trees. Ring numbers of heartwood and sapwood were equivalent at height of approximately 5 m for dominant trees, 3 m for intermediate trees, and 1 m for suppressed trees. Ratios of heartwood ring numbers changed almost the same as heartwood ring numbers.

Heartwood diameter

Similar to ring number, heartwood diameters decreased with increasing height while sapwood widths were more stable within the stem sections in which heartwood occurred(Figs. 3 D–F). Heartwood diameters of dominant, intermediate and suppressed trees were 18.7(± 3.1) cm, 15.1(± 1.9)cm and 10.2(± 3.6) cm at the base, and sapwood widths 9.0(± 1.8) cm, 8.5 (± 2.1) cm and 6.9 (± 1.2) cm, respectively.These f igures show that heartwood diameters and sapwood widths of suppressed trees were lower than those of dominant and intermediate trees. Heights where heartwood diameter and sapwood width were equal were 6 m for dominant trees, 5 m for intermediate trees and 3 m for suppressed trees. They were 1–2 m higher than those for ring number.Heartwood areaSimilar to heartwood ring and diameter, both heartwood areas and heartwood area ratios decreased with increasing height. The ratio at the base was the highest (47.1%) for dominant trees and lowest (34.7%) for suppressed trees. The sapwood area of trees gradually decreased with increasing height, unlike ring number and sapwood widths which were almost constant. The heartwood area was smaller than sapwood area at any height position along the stem (Figs. 3 G–I).

Fig. 4 Comparison of heartwood volume and its ratio to whole volume among dominant, intermediate and suppressed trees of Erythrophleum fordii. Dif ferent letters refer to signif icant dif ferences between classes of social status at level of 0.05, uppercase for heartwood height, and lowercase for its ratio

Fig. 5 Cumulative heartwood volume along stem of dominant, intermediate and suppressed trees of Erythrophleum fordii

Heartwood volume

There were signif icant dif ferences in heartwood volume and its ratio among dominant, intermediate and suppressed trees(Fig. 4). The volume and its ratio in stems of dominant trees were 0.11 ± 0.04 m 3 and 29.3%, remarkably higher than those of suppressed trees (0.03 (± 0.02) m 3 and 20.9%),respectively. Intermediate trees dif fered from dominant and suppressed trees in heartwood volume but not in its ratio.Figure 5 showed the cumulative heartwood volume with increasing height along the stem. As indicated in Table 2,height to crown base was approximately 8 m for all sampled trees. In the section below this height position, 90.0% of the heartwood volume was in dominant trees, 89.6% for intermediate trees, and 87.6% for suppressed trees.

Relationship between heartwood development and tree growth

Heartwood ring number, diameter and area at breast height were positively correlated with DBH, tree height, and crown diameter (Table 4). Heartwood diameter and area at breast height were also signif icantly related with crown length.Heartwood heights were signif icantly inf luenced by DBH and tree height, and heartwood volume by DBH, tree height,and crown diameter and length. Of these growth traits, DBH was the most important factor inf luencing heartwood volume, followed by tree height.

Discussion

In this study, heartwood diameter and area of dominant trees were signif icantly higher than those of intermediate trees in 32- to 35-year-old plantations ofErythrophleum fordii, with no dif ferences in their ratios to xylem between dominantand intermediate trees. These parameters were considerably lower for suppressed trees. This is in agreement with a study of 31-year-old teak (Tectona grandisL.f.) plantations in the same region (Yang et al. 2020). The dif ferences between both studies is that the heartwood area ratio in xylem dif fered signif icantly between intermediate and suppressed trees forT. grandis, while there were no signif icant dif ferences forE. fordii. Kokutse et al.’s ( 2010) study of a 25-year-old teak plantation in West Africa demonstrated that the heartwood area ratio in xylem of dominant trees was higher than that of intermediate trees, and intermediate trees performed better than suppressed trees. The heartwood area ratio at breast height ranged from 0.30 to 0.42 in the present study, 0.58 to 0.68 in the study by Yang et al. ( 2020), and were all above 0.90 in the study by Kokutse et al. ( 2010). These dif ferences may be due to species dif ferences, genetics, developmental stage, site conditions, and/or management regimes (Climent et al. 2002; Taylor et al. 2002; DeBell and Lachenbruch 2009; Miranda et al. 2009).

Table 4 Relationship between heartwood attributes and growth traits of Erythrophleum fordii (n = 40)

The heartwood ofE. fordiiappeared in cross-sections of the stem as diameters reached 5–10 cm, and heartwood diameter and area were in signif icant linear and binomial relationships with xylem diameter. This f inding is similar to Blackwood (Acacia melanoxylonR.Br.) in Argentina(10 cm, Igartúa et al. 2017) and the Portuguese oak (Quercus fagineaLam.) in Portugal (10–20 cm, Sousa et al. 2013).Their heartwood diameters increased with increasing xylem diameter. The ratios of heartwood diameter and area seemed to be almost constant after xylem diameters above 25 cm.This should be verif ied with sampling more large-sized trees.This value is much larger than that ofT. grandis(16 cm) in the study by Yang et al. ( 2020). The dif ferences might be related with inter-species dif ferences in age of initial heartwood formation (Wang et al. 2010), about 10 years forE.fordii(Zhao et al. 2015) and 7 years forT. grandis(Okuyama et al. 2000).

Heartwood ring number, diameter and area as well as their ratios in the stem decreased with increasing height ofE. fordii, while sapwood ring number and width were nearly constant from base to the position of heartwood height. This was dif ferent from studies onA. melanoxylon(Knapic et al.2006),Q. faginea(Sousa et al. 2013 ) andT. grandis(Yang et al. 2020), in which ratios of heartwood diameter and area were relatively stable or near constant in the lower part of the stem. The dif ferences between these studies are because the area of sapwood and its ratio are much higher than heartwood at any height of mid-agedE. fordii, unlike the other species. As Zhao et al.’s ( 2015) study of the natural forests ofE. fordii, heartwood and sapwood areas at stem base were equal to each other when above 40 years old.

The ratios of heartwood height to whole height ofE.fordiitrees were all above 60%, even up to 75% in some instances, while they were not inf luenced obviously by social status in stand although heartwood heights of dominant and intermediate trees were signif icantly higher than that of suppressed trees. The ratios were lower than those forA. melanoxylonin Portugal, ranging from 67% to 85% (Knapic et al.2006), and mature southern blue gum (Eucalyptus globulusLabill.) ranging from 82 to 87% (Moraisa and Pereira 2007).Social status signif icantly inf luenced heartwood volume,and dominant trees produced better than suppressed trees,as their heartwood volume ratios were less than 30%, and approximately 90% occurred below 8 m regardless of social status.

Heartwood volume ofErythrophleum fordiiwas greatly inf luenced by DBH and tree height, which is in agreement with results forA. melanoxylonin Argentina (Igartúa et al.2017). DBH or DBH and tree height can thus be used as variables for developing heartwood volume prediction equations once more samples are evaluated in larger scale of plantations at dif ferent ages and on a variety of sites. In addition, positive ef fects of silvicultural methods such as fertilization, thinning, and weeding, on heartwood development have been conf irmed by Wilkins ( 1991) for the f looded gum(Eucalyptus grandisW. Hill ex Maiden), Yang and Hazenberg ( 1992) forPicea marianaB.S.P. andP. glaucaVoss.,and Pérez and Kanninen ( 2005) forT. grandis. Miranda et al. ( 2006) reported that irrigation and fertilization could improve the growth ofE. globulusand increase its heartwood content, and considered that heartwood formation was more dependent on tree size than age. These silvicultural methods should also be evaluated forErythrophleum fordiito determine if they may be ef fective for the promotion of heartwood development.

Our results may also be used in crop tree management of valuable timber species such asE. fordii. Dominant trees with high growth potential, good stem form and few defects are normally selected as crop trees. When a crop tree release is carried out (Puhlick et al. 2019), it is necessary for the rapid production of target-sized heartwood to maintain crown development by thinning neighboring trees (Schuler et al. 2013), since crown properties are also positively correlated with heartwood development in this study. As the majority of heartwood developed in the lower stem, this area below 8 m or 10 m should be given more attention. It is the prime section for heartwood production by valuable timber species. Since trees can alter their social status in stand, intermediate trees may become dominant (Li et al.2015), and intermediate and suppressed trees had better be remained more as possible if they do not negatively inf luence growth of dominant trees when thinning for crop trees.This can increase not only production of large-size heartwoods at individual levels, but also the increased production of heartwood at the stand level.

Conclusions

Number of annual rings, diameter and area of heartwood at breast height were inf luenced signif icantly by social status of trees in mid-agedErythrophleum fordiiplantations, and dominant trees performed the best. Heartwood was formed when the xylem diameter was 5–10 cm at any height. Heartwood ring number, diameter and area as well as their ratios decreased with increasing height. Social status greatly inf luenced heartwood height and volume, and dominant trees performed better than suppressed trees. Heartwood volume ratios were less than 30% and approximately 90% of heartwood occurred below 8 m regardless of social status. Relationship analysis showed that heartwood volume was greatly inf luenced by DBH. The f indings are useful for plantation management of valuable tropical timber species with valuable heartwood such asE. fordii.To improve heartwood production techniques for this species, age-related heartwood formation processes and prediction of heartwood volumes will be further studied.