Radial growth and non-structural carbohydrate partitioning response to resin tapping of slash pine (Pinus elliottii Engelm. var. elliottii)

Baoguo Du·Qifu Luan·Zhanglin Ni·

Honggang Sun3·Jingmin Jiang3

Abstract Slash pine (Pinus elliottii Engelm. var. elliottii) is a resin-producing species grown worldwide for significant economic benefits for wood production. Resin tapping creates a carbon sink at the expense of carbon allocation for growth and consequently, wood production may be reduced. Non-structural carbohydrates comprising starch and sugars stored in plant organs, may serve as intermediate pools between assimilation and utilisation. However, the effect of resin tapping between tree growth and non-structural carbohydrates is not well understood. This study investigated (1) the effects of resin tapping on radial growth, (2) the effects of resin tapping on non-structural carbohydrate pools in different compartments, and (3) the feasibility of resin production without disruption of tree growth. Twenty one-yearold slash pines were subjected to resin tapping over two successive years. Non-structural carbohydrate concentrations in needles, branches, stem phloem, and roots of tapped and untapped trees in summer and winter were determined after the second year of resin harvest. The results showed that tapping had no significant effects on annual increments. Starch was the dominant non-structural carbohydrate fraction, regardless of tissues and season, and constituted up to 99% of the total non-structural carbohydrates in the phloem and roots. Glucose and fructose were the dominant sugars; sucrose was negligible. Compared with the controls, tapped trees showed 26% lower non-structural carbohydrate concentration in the phloem above the tapping wound in summer, which was attributable to the decreased abundance of starch, glucose, fructose, and sucrose. In winter, the altered nonstructural carbohydrate profiles in the phloem above the tapping wounding were minimised as a result of recovery of the sugar concentrations. In contrast to free sugars, which accumulated substantially in needles and branches during winter, starch was enriched in the phloem, roots, and current-year needles. The results provide evidence for a localised effect of resin tapping, and highlight the observation that resin extraction does not always cause a sacrifice in wood growth under a moderate resin-tapping intensity in slash pine plantations.

Keywords Slash pine·Localised effects·Non-structural carbohydrate·Resin tapping·Radial growth

Introduction

Resin-producing tree species are of considerable commercial importance worldwide as a source of flavour and fragrant materials (Soerianegara and Lemmens 1993; Rijkers et al. 2006). Pine oleoresin from pine species plays a significant role in the aromatic/agarbathi (incense) industries (Murugesan et al. 2011). High oleoresin yields are desirable because the derivatives constitute valuable feedstock for renewable chemicals widely used in products that compete with petroleum-derived feedstock (Rodrigues-Corrêa et al. 2012; Susaeta et al. 2014). Currently, increased demand for natural resins has boosted prices on global markets (Génova et al. 2014).

Resins are carbon-rich compounds synthesised in epithelial cells (Gershenzon 1994; Rijkers et al. 2006; Cown et al. 2011), requiring considerable metabolic energy (Maaten et al. 2017), and stored in a network of specialized canals located throughout the wood, bark, roots and needles (Trapp and Croteau 2001). Studies have shown that resin yield is positively related to diameter at breast height (DBH) (Rodrigues et al. 2008; Rodríguez-García et al. 2014). Resin flow may be influenced by a number of factors such as light, temperature and moisture, and the maximum resin flow has been associated with late spring and summer months (Rodrigues et al. 2008; Rodríguez-García et al. 2014). The exudation of resin is generally mediated by anatomical and metabolic changes in both radial and axial canals which occur in the outer layer of the earlywood and latewood (Boschiero and TomazzelloFilho 2012). Only small quantities of resins exude naturally on the trunk surface due to wind injury, lightening or wounds caused by insects or fungal pathogens (Nair 2000). However, mechanical tapping for commercial purposes by a repeated wounding process (Rodrigues et al. 2008; Tümen and Reunanen 2010) represents a drain on carbon reserves that limits the carbon pool of the trees and requires preferential allocation of carbohydrates for resin synthesis at the expense of allocation for generative growth (Lombardero et al. 2000; Rijkers et al. 2006; Mengistu et al. 2012). Negative effects of tapping on radial growth have been reported in several pine species (Chen et al. 2011; Papadopoulos 2013; Génova et al. 2014; Rodríguez-García et al. 2016), as well as in frankincense (Boswellia papyrifera(Del.) Hochst) with reduced leaf area, fewer healthy and filled seeds, decreased germination and sexual reproduction (Rijkers et al. 2006; Mengistu et al. 2013). Resin tapping also induces local wood anatomy changes, including the formation of traumatic resin canals (Martin et al. 2002; Luchi et al. 2005; Boschiero and TomazzelloFilho 2012; Papadopoulos 2013; Rodríguez-García et al. 2015, 2016), and the activation of constitutive axial resin canals refilling their resin reservoirs (Ruel et al. 1998; Phillips and Croteau 1999). It has been found that canals in the vicinity of the wound start to empty their contents earlier than those in distal parts from the wound, which are mediated by the induced systemic responses (Rodríguez-García et al. 2014), and the de novo resin synthesis process is activated for 3-4 d after wounding (Rodríguez-García et al. 2015). In spite of considerable research, knowledge is lacking of the effects of resin tapping on the ability of trees to maintain their vegetative status and photosynthetic capacity and consequently, their ability to acquire carbon and produce additional resin and wood in the future (Mengistu et al. 2013; Génova et al. 2014; Susaeta et al. 2014).

Compared to anatomical structural differentiation processes, which are slow and deplete a large amount of carbohydrate reserves, metabolic processes are rapid (Bonello et al. 2006; Moreira et al. 2015). Non-structural carbohydrates (NSCs) play distinct pivotal roles in plants, including transport, energy metabolism, and osmoregulation, and provide substrates for the synthesis of defence compounds (Hartmann and Trumbore 2016). The NSC pools are frequently investigated to evaluate variations in carbon balance (Bansal and Germino 2009) which depend on dynamic exchanges between assimilation and consumption. The stored carbohydrates play an important role in tree growth and development as they can be used when photosynthesis is inadequate to meet carbon requirements for maintenance and growth, particularly under stress conditions (Bansal and Germino 2009; Hartmann and Trumbore 2016). Starch is a major constituent of the NSC pool; its concentration is positively correlated with the ratio of photosynthesis to respiration in conifer seedlings (Hoch et al. 2003; Bansal and Germino 2009), as well as in rubber trees (Hevea brasiliensisMüll. Arg.) (Silpi et al. 2007; Chantuma et al. 2009). Soluble sugars are normally the second-most abundant NSC fraction, dependent on the supply of carbohydrates from source organs to sinks. Sucrose and glucose act as either substrates for cellular respiration or as osmolyte to maintain cell homeostasis (Gupta and Kaur 2005), whereas fructose is more predominantly associated with secondary metabolite synthesis (Rosa et al. 2009). To our knowledge, in comparison to the widely studied influence of tapping on conifer anatomical structures and radial growth, little information is available regarding the responses of NSC profiles in relation to resin tapping. Several studies only recently elaborated on the effects of frankincense tapping onB. papyrifera(Del.) Hochst (Mengistu et al. 2012, 2013) and latex tapping on rubber trees (Chantuma et al. 2007, 2009; Silpi et al. 2007) on NSC pools.

Slash pine is an important species originating from the southeastern United States and introduced to China some 80 years ago. Owing to its remarkable characteristics, i.e., rapid growth, wide adaptability, and high resin yield (Wen et al. 2004; Susaeta et al. 2014), the current plantation area exceeds 2 million ha and it is thus a major timber and resintapping species in southern China (Zhang et al. 2016). However, information on the physiological processes underlying the effects of regular resin tapping on the growth ofPinusspecies is limited. Comparative studies are needed to assess the effects of spatiotemporal and management variations on physiological regulations and mechanisms. In the present study, we assessed the effects of resin tapping of mature slash pine on radial growth and NCS pools of different compartments in summer and winter. We hypothesised that: (1) resin tapping will reduce radial growth due to direct carbon drain and more carbohydrates being invested in resin production rather than growth; (2) NSC allocation patterns will be altered as a result of resin tapping and specifically, compared with controls, tapped trees will show lower NSC reserves in summer and the difference will dissipate during winter; and, (3) tapping will induce prominent localised effects close to the tapping wound rather than induced distal systemic responses.

Materials and methods

Study sites

The study sites were located on Tianmu Mountain in Hangzhou, China (30°42’ N, 120°30’ E), which experiences a subtropical monsoon climate. From 1990 to 2000, mean annual, monthly minimum (January), and monthly maximum (July) temperatures were 16.7, 5.6, and 27.3 °C, respectively. The average length of the growing season is 246 days, and mean annual precipitation is 1697 mm. Slash pine plantations were established in January 1993 with spacing of 2 m × 2 m between rows and between trees within a row. Average height of 21-year-old trees was 17.1 m before the start of tapping. Averaged DBH varied from 20.3 to 21.2 cm (Table S1). The experiments were replicated in three adjacent plots (S1, S2, and S3, each 0.11 ha in area) subject to the same silvicultural practices but with different orientations from east to southwest (Table 1). Soil properties, pH, total nitrogen (N), ammonium (NH4+), available phosphorus (P), and organic matter contents, are given in Table 1.

Table 1 Site information and soil characteristics of the three study plots

Tapping regime, growth measurements, and sampling of trees

Tapping was performed from mid-May to mid-October (＞ 10 °C daily mean temperature) in 2014 and 2016 according to the bark chipping method (Tümen and Reunanen 2010). The V-shaped bilateral tapping method exposing the sapwood was used to obtain crude resin (Rodríguez-García et al. 2015). The tapping face was on the uphill side of each trunk at 2.0 m above the ground. The width of the tapping face, based on technical regulations of resin tapping (LY/T 1694-2007) (State Forestry Bureau of China 2007), did not exceed 40% of the stem circumference to avoid detrimentally affecting stem growth (Fig. 1). Resin tapping was carried out every second day between 10:00 and 14:00. The total amounts of resin collected in 2014, 2015 and 2016 were 4.2 ± 0.5, 3.8 ± 0.6 and 3.6 ± 0.6 kg per tree, respectively. Both control trees (without tapping) and tapped trees were selected randomly in plots S1, S2 and S3, and the number of control trees and tapped trees were 12/42, 10/23 and 11/23, respectively.

Fig. 1 Resin tapping and plant sampling; the downward V-shaped bilateral tapping method was used to obtain resin; phloem samples were harvested from the upper right resin tapping face and the opposite side at the same height

Increment cores were taken from the tapping face of the untapped controls and tapped trees using a 5 mm increment borer 2.1 m above the ground in December 2018. The cores were dried, glued to core holders, and surfaces smoothed. Annual radial increment was measured by scanning the core surface with a visible core microtome, and then ring widths were measured using a Lintab? 6 measurement system (Rinntech?, Heidelberg, Germany) with 1/1000 mm resolution. Finally, the measured ring widths were detected with TSAP-Win? and COFECHA software for tree ring crossdating (Vannoppen et al. 2019).

Needles, branches, stem phloem, fine roots from both controls and tapped trees were analyzed for carbohydrate contents in July and December, 2018, representative of summer and winter. Current and previous year needles and branches, without visible insect/pathogen injuries, were sampled from the middle of the crowns using 20-m-high retractable pruning shears. Stem phloem samples 8 cm long × 3 cm wide were taken with a chisel and hammer from the same tree in summer and winter above the tapping wound (Phloem-AT), as well as from the rear side opposite the tapping wound (Phloem-OT) (Fig. 1). With the controls, phloem samples were taken from the same position on the trunk as the tapped trees. The bark layer was removed and only ca 2.0 mm thickness from the phloem was harvested. Root samples from the 10-cm depth consisted entirely of fine roots ＜ 2 mm. The sampling took place from 10:00 to 14:00. All samples were stored on dry ice to avoid tissue respiration and transported to the laboratory. Samples were oven dried at 60 °C for 72 h until constant weight was attained, ground into powder using a Wiley Mini Mill (Thomas Scientific, Swedesboro, NJ, USA) and passed through a 0.5 mm-mesh sieve.

Soil sampling and chemical analyses

In spring 2014, five soil sampling points arranged in an “S”-shape were sampled in each plot using a hand auger. A total of 500 g of surface soil (upper 15 cm) was collected from each sampling point, transported to the laboratory, air-dried, ground and passed through a 100-mesh (0.15 mm) sieve before analyses.

So who was I if I was no longer capable and busy? How could I respect myself if all I did was sit on the couch everyday and watch the snow fall?Learning how to receive the love and support that came my way wasn t easy

Total soil nitrogen (total N) and total phosphorus (total P) were measured using the semi-macro Kjeldahl digestion and the molybdenum-antimony colorimetric method followed by H2SO4digestion, respectively (Allen 1974). NH4+-N content was measured using the titration method after MgO extraction-diffusion. Total potassium (total K) was measured by inductively coupled plasma-optical emission spectrometry (ICP-OES), followed by concentrated H2SO4and HClO4digestion (LY/T 1234-2015). Available K was measured using ICP-OES after extraction with CH3COONH4solution (LY/T 1234-2015). Available P was analysed by Mo-Sb colorimetry after extraction with diluted H2SO4-HCl (LY/T 1232-2015). Soil organic matter content was analysed by titration after oxidation with K2Cr2O7solution (Hesse 1971).

Analysis of non-structural carbohydrate concentrations in plant samples

The determination of NSC content followed Hoch et al. (2003). For extraction, 20 mg of dried plant powder was boiled for 30 min in 2 mL deionized water. After centrifugation at 8000 rpm for 5 min, 500 μL of the supernatant was transferred to new micro tubes and treated with phosphoglucose-isomerase and invertase (Sinopharm Chemical Reagent Co., Ltd, Beijing, China) to convert fructose and sucrose to glucose. Total glucose was measured in a microplate photometer (BioTek Instruments, Inc., Winooski, VT, USA) at 340 nm after conversion of glucose to glucose 6-phosphate with the glucose hexokinase assay (G3 292, Sigma). The balance of the water extract, including starch, was incubated with a crude fungal amylase (Sinopharm Chemical Reagent Co., Ltd., Beijing, China) at 40 °C for 15 h to decompose starch to glucose. Total glucose concentrations were then determined as described above. Concentration of starch was calculated indirectly by subtracting the concentration of free sugars (i.e., glucose, fructose and sucrose) from the total glucose concentration after digestion of starch. Pure glucose, fructose, and sucrose solutions were used for calibrations. The replicability of glucose determination was checked by analysingCitrusleaves (Institute of Geophysical and Geochemical Exploration of China, Beijing, China). The NSC concentrations are reported as a percentage of dry matter.

Statistical analysis

For an initial exploratory analysis of the effects of site and resin tapping on non-structural carbohydrate (NSC) concentration, partial least squares discriminant analysis (PLS-DA) (Chong et al. 2018) was carried out on the publicly available platform MetaboAnalyst 4.0 (http:// www. metab oanal yst. ca/ Metab oAnal yst/) (Xia and Wishart 2016). Raw data were log transformed and mean-centred for normalization.

Data were statistically analysed by one-way analysis of variance (ANOVA) followed by an analysis of least significant differences (LSD) and Student’st-test (when only two groups were considered, i.e., summer vs winter, and control vs resin tapping) using the software SigmaPlot 11.0 (Systat Software GmbH, Erkrath, Germany). Raw data were log10transformed if required. Bar charts were generated using SigmaPlot 11.0, and colour scales of fold changes were created with Microsoft Excel 2013 (Microsoft Corporation, Redmond, WA, USA) using conditional formatting tools. Concentrations of all parameters were based on the tissue dry weight (DW).

Results

Soil characteristics and tree growth on three plots

Table 2 Increment cores from 2014 to 2018 of control and tapped trees on three adjacent plots, as well as three plots of pooled data

Effect of resin tapping on non-structural carbohydrate concentrations (NSC) in summer and winter

In a preliminary analysis of the effects of resin tapping on NSC profiles in summer and winter at the three sites, PLSDA was applied to NSC data (incorporating starch, glucose, fructose, and sucrose) as determined in different tissues. No clear tapping- and site-related clusters were resolved, but apparent variations between summer and winter were demonstrated (Fig. 2). Since there was little difference between the three study plots in terms of NSC profiles as well as the response to tapping (Figs. S1-S5), data from the three plots were pooled to have more vigorous results.

In summer, total non-structural carbohydrate (TNC) concentrations, representing the sum of starch, glucose, fructose and sucrose in the phloem above the tapping wound (phloem-AT) of resin-tapped trees significantly decreased (73% of control,p＜ 0.01) (Fig. 3), concurrent with significant declines in starch and sugar concentrations (Figs. 4 and 5). Total sugar as well as fructose concentrations in current year needles of tapped trees were 15% and 18% lower than controls, respectively (p＜ 0.05) (Fig. 5). Resin tapping had no other effects on NSC concentrations in summer. The decline in TNC and starch in phloem-AT of resin-tapped trees observed in summer were still restrained in winter, i.e., 12% and 13% lower than controls, respectively (Figs. 3 and 4), whereas the differences in sugars between controls and tapped trees were insignificant (Fig. 5).

Partitioning of non-structural carbohydrate (NSC) in summer and winter

Starch was the most abundant NSC irrespective of season and tissue and ranged from 57 to 99% of the TNC in the phloem and roots, respectively (Figs. 3 and 4, Table S2). Glucose and fructose were the two most abundant sugars regardless of season and tissue (Fig. 5) and constituted up to 15% and 24% of the TNC, respectively (Table S2). Sucrose was only abundant in the bark and stem in winter, ranging from 0.8 to 9.3 mg g-1dry weight (Fig. 5C). In summer, glucose and fructose were more abundant in the phloem, and in conjunction with higher starch concentrations, resulted in higher TNC. In winter, TNC and starch accumulated in higher concentrations in the phloem and roots; higher quantities of glucose and fructose accumulated in the needles. Compared to summer, TNC concentrations in winter were significantly higher in needles, current year branches, phloem and roots (Fig. 6). Similar patterns were also found for in starch. Concentrations of sugars were significantly higher in the phloem in the summer, but accumulated in needles, branches and roots in winter (Fig. 6).

Fig. 2 Clustering of carbohydrate concentrations in Pinus elliottii trees computed separately per season and treatment; figures are score plots from a partial least squares discriminant analysis (PLS-DA) computed per season (A and B), and per treatment (C and D) of trees from three plots (S1, S2, and S3). A and B show plots for control (Control) and resin-tapped (Tapped) trees in summer (A) and winter (B). C and D show plots for clustering between summer and winter for control (C) and resin-tapped trees (D). Semi-transparent shading indicates 95% confidence regions

Fig. 3 Three sites pooled for total non-structural carbohydrate (TNC) concentrations in different tissues of controls and resin-tapped trees in summer and winter. Data are means ± SE(n = 10-12 for controls and 23-42 for resin-tapped trees). ** and *** indicate significant differences between controls and resin-tapped trees at p ＜ 0.01 and 0.001, respectively; AT, above the tapping wound; OT, opposite the tapping wound; 0, current-year; 1, previous-years growth

Fig. 4 Three sites pooled for starch concentrations (mg g-1 DW) in different tissues of controls and resin-tapped trees of slash pine in summer and winter. Data are means ± SE (n =10-12 for controls and 23-42 for resin-tapped trees). ** and *** indicates significant differences between control and resin-tapped trees at p ＜ 0.01 and 0.001, respectively. AT, above the tapping wound, OT, opposite the tapping wound, 0, current-year, 1, previous-year

Fig. 5 Three sites of pooled concentrations (mg g-1 DW) of glucose (A), fructose (B), sucrose (C) and their sum (D, total sugar) in different tissues of control and resin-tapped trees in summer and winter. Data are means ± SE (n = 10-12 for controls and 23-42 for resin-tapped trees). *, ** and *** indicate significant differences between controls and resin-tapped trees at p ＜ 0.05, 0.01 and 0.001, respectively; AT, above the tapping wound; OT, opposite the tapping wound; 0, current-year; 1, previous-year

Discussion

Effects of resin tapping on radial growth

According to the “growth-differentiation balance” hypothesis, growth and secondary processes (such as resin biosynthesis) represent competitive sinks for assimilated carbon (Herms and Mattson 1992; Rodrigues-Corrêa et al. 2012). Therefore, the trade-off between growth and resin production has been a concern in forest management (Herms and Mattson 1992). Like other pine resins, slash pine resin is predominantly a complex mixture of volatile and non-volatile terpenes, which constitute the largest group of secondary products (Trapp and Croteau 2001; Rodrigues-Corrêa et al. 2012; Zhang et al. 2016). Tapping creates competition in carbon sinks between growth, including vegetative growth and reproduction, and resin production (Tuomi et al. 1988; Chen et al. 2011; Cown et al. 2011; Génova et al. 2014). This is confirmed by the decreased radial growth reported for severalPinusspecies, e.g., maritime pine (P. pinasterAit.) (Génova et al. 2014; Rodríguez-García et al. 2016), Aleppo pine (P. halepensisMill.) (Papadopoulos 2013) and Masson pine (P. massonianaLamb.) (Chen et al. 2016). The lesser leaf area will consequently limit tree growth by diminishing the carbon gain (Mengistu et al. 2012). Moreover, Novick et al. (2012) observed that pines tended to invest higher quantities of carbon, originally assimilated for growth, to oleoresin production as a defence mechanism under adverse situations. This would render tapped pines more vulnerable, or to suffer more strongly the negative effects on growth under stress or distributions. Silpi et al. (2006) reported that tapping caused up to 80% decrease in the cumulative June to November radial growth of rubber trees. The effects on radial growth varied with location around the tapping wound, i.e., a lower radial growth was recorded in the tapped panel than in the untapped panel, and a higher rate above the wound than below. Moreover, the negative impact of tapping on growth was much stronger in the second year (Luchi et al. 2005; Silpi et al. 2006; Boschiero and TomazzelloFilho 2012; Papadopoulos 2013; Rodríguez-García et al. 2016).

In this study, diameter increments were not significantly affected by moderate intensity resin tapping (Table 2). In line with these results, tapping had little effect on the growth ofP. pinasterAit. andP. caribaeaMorelet (Muga et al. 1995; Rodríguez-García et al. 2014). This finding is import for forest management because the present data provide novel insights into the feasibility of resin production without the sacrifice of growth.

Effects of resin tapping on NSC

Understanding carbon allocation patterns in trees provides the knowledge necessary for interpreting how these patterns change with stress and for developing physiologically based management strategies and genetic improvement programs (Dickson 1989). However, the ecological impacts of harvesting wood exudates from natural forests, particularly pine species, remains poorly studied. Resins and other wood exudates mainly contain carbon-based compounds and are frequently complex mixtures of terpenes (Ticktin 2004; Rijkers et al. 2006). With regard to tapped trees, decreased carbohydrate pools are expected because tapping creates an additional carbon sink and higher quantities of carbon are allocated to defence (Herms and Mattson 1992). This is confirmed by research on frankincense trees which showed significantly lower concentrations of TNC, total soluble carbohydrates and starch in stems, bark and root tissues after tapping (Mengistu et al. 2012). Compared to controls, tapped trees of slash pine had 26% lower TNC levels (p＜ 0.001) in the phloem above the tapping wound in summer, attributed to significantly lower concentrations of starch and sugars, i.e., 26% and 24% lower, respectively. However, this impact was only observed in phloem above the tapping wound, indicating a localized effect (Rodríguez-García et al. 2014, 2016), which might be explained by the disturbance of shoot to root transportation and/or stimulated defense induced by wounding (Jacob et al. 1998; Bonello et al. 2006; Moreira et al. 2015). Previous studies also observed, in contrast to other Pinaceae, thatPinusspecies show a more localized traumatic resin canal response that is quickly attenuated with distance from the attack site (Nagy et al. 2006; Krokene and Nagy 2012). Similar localized sink effects for carbohydrates in bark and wood have been observed in tapped rubber trees but noticeable increases of carbohydrate storage in wood and bark at the expense of radial growth (Chantuma et al. 2007, 2009; Silpi et al. 2007), possibly indicating species-specific responses to tapping wounds.

In addition to the localized impact on non-structural carbohydrates, tapping also induced lower fructose levels as well as total sugar in previous year needles in summer (Fig. 5). Similarly, Mengistu et al. (2012) observed that intensive tapping reduced the amount of carbon allocated to foliage and reproductive sinks of frankincense trees. In conifers, allocation of carbon resources is mainly controlled by the relative strengths of various sinks and their proximity to stored or currently produced carbohydrates (Kozlowski 1992). The decline in sugar concentrations in previous year needles observed in the present study were possibly due to enhanced utilization for secondary growth of the axial system (Hansen and Beck 1994), and for new needle development (Du et al. 2014). Altered sugar concentrations in needles were fully restored in winter when the photosynthetic gain was still high but radial growth had more or less finished (Alfieri and Evert 1968) and there was no new needle development. However, TNC and starch concentrations only partially recovered and remained significantly lower (12% and 13% of TNC and starch, respectively) in the phloem above the tapping wound in winter (Figs. 3, 4).

Fig. 6 Comparison of winter and summer of non-structural carbohydrate concentrations in different tissues of controls (C) and resin-tapped (T) trees; AT, above the tapping wound; OT, opposite the tapping wound; 0, current-year; 1, previous-year. Blocks with “ns” indicate no significant differences (p ＞ 0.05) between summer and winter; the remaining blocks without “ns” differed significantly (p ＜ 0.05) between summer and winter. Red and blue shading indicate increase and decrease, respectively. Sucrose concentrations in needles and roots were missing due to below the detect limit; total sugar is the sum of fructose, glucose, and sucrose concentrations

Composition and distribution of non-structural carbohydrates (NSC)

As the total of starch and free sugars (sucrose, glucose, and fructose), TNC commonly comprises more than 90% of the mobile carbon in plants (Hoch et al. 2003). Contrary to Gholz and Cropper (1991) and other species (Hoch et al. 2003; Silpi et al. 2007; Chantuma et al. 2007, 2009), we found starch was consistently the most abundant NSC in slash pine tissues, regardless of season and tapping. It ranged from 57 to 99% of TNC in the phloem and roots, respectively, indicating that starch was the main stored NSC in phloem of slash pine (Popp et al. 1989; Gruber et al. 2013). Similarly, the higher phloem concentrations of starch than sugars were also documented in slash pine (Popp et al. 1989), and forP. cembraL. andLarix deciduaMill. (Gruber et al. 2013), as well as for ash (Fraxinusspp.) species (Hill et al. 2012). Glucose and fructose were the second most abundant TNC fractions in both controls and tapped slash pine trees. Sucrose is a major end product of photosynthesis and plays a central role in plant growth and development, e.g., as a direct or indirect regulator of gene expression (Koch 2004; Rosa et al. 2009). In contrast to many other species (Sung et al. 1996; Hoch et al. 2003; Chantuma et al. 2007; Gruber et al. 2012), sucrose was only abundant in winter in the phloem and branches of both control and tapped trees. It accounted for a negligible proportion of TNC in the present study, indicating it served as major transport assimilate in slash pine (Koch 2004). The low concentration of sucrose and/or soluble sugars is possibly a mechanism to enhance photosynthesis and NSC reserve mobilization (Rosa et al. 2009), as well as resin production, as reported for rubber trees (Chantuma et al. 2009).

In summer, the phloem contained a higher concentration of free sugars than in other tissues, whereas free sugars also accumulated in needles in winter (Table S2). Progressive accumulation of foliar soluble sugars, peaking at the end of the growing season, has been documented in many conifer species (Bansal and Germino 2009) and for rubber trees (Chantuma et al. 2009). With regards to whole tree pools, the trunk contained the highest reserve pools while roots and bark contained equal amounts (Mengistu et al. 2012). Consistent with marked seasonality (Gholz and Cropper 1991), differed NSC profiles were also found in this study, i.e., compared to summer, starch accumulated in the phloem, roots, and current year needles, whereas free sugar accumulated more in the needles, branches, and roots in winter, regardless of resin tapping (Fig. 6). The recovery in stored free sugars is speculated to be in readiness for growth the following season (Bansal and Germino 2009; Chantuma et al. 2009), therefore the growth would not be diminished (Gupta and Kaur 2005; Rosa et al. 2009; Mengistu et al. 2013). We did not observe any apparent effect of resin tapping on NSC composition and distribution among tissues in summer or winter.

Conclusions

In this study, the results indicate that moderate resin tapping has no detrimental impact on tree growth and non-structural carbohydrates, except for a localized effect close to the tapping wound. The study provides additional evidence to support the feasibility of resin production without diminishing growth of slash pine under a moderate tapping intensity. However, considering increasing frequencies of heat and drought under a changing climate, which will have negative effects on tree growth, production, and survival, additional research is recommended before silvicultural and management recommendations are formulated.

AcknowledgementsB. Du was supported by Mianyang Normal University (QD2017A001, MYSY2017PY02) and the Science and Technology Department of Sichuan Province, China (2019YJ0508). We thank Fubin Tang and Runhong Mo for their support in facilitating our project. We thank Dr. Robert McKenzie from Liwen Bianji, Edanz Group China (www. liwen bianji. cn/ ac), for editing the English text of a draft of this manuscript.