Aboveground Biomass Productivity and Nutrient Use Dynamics of Clumping Tropical Bamboos in Northern Thailand

Abstract

Bamboo has great potential in restoring degraded lands while providing multiple environmental benefits and harvestable products at regular intervals due to its unique characteristics of rapid growth. However, a comprehensive and species-specific knowledge of biomass productivity, nutrient dynamics, and potential harvest-induced nutrient losses is lacking. This study compared the biomass and nutrient dynamics of four bamboo species Bambusa oldhamii Munro, Dendrocalamus latiflorus Munro, Bambusa tulda Roxburgh, and Dendrocalamus brandisii (Munro) Kurz in Northern Thailand. Field measurements, laboratory analyses, and statistical modeling were used to estimate their aboveground biomass (AGB), nutrient use efficiency (NUE), and nutrient stocks. Culm diameter at breast height (DBH) and age were identified as the most reliable predictors of AGB. The study revealed that D. brandisii had superior productivity and NUE compared to the other species, particularly the introduced non-native bamboo species. These findings emphasize the need for species-specific strategies that consider both biomass productivity and nutrient dynamics. Furthermore, D. brandisii can be suggested as a native candidate for bamboo plantations in Northern Thailand and similar environments, given its high productivity and efficient nutrient use, underpinning its potential contribution to environmental rehabilitation and rural livelihoods. However, more research is required to minimize nutrient losses and maintain a productive age structure.

1. Introduction

Bamboo cultivation is increasingly recognized for its environmental benefits, including carbon sequestration, soil conservation, and biodiversity enhancement, alongside its rapid growth and potential for degraded land rehabilitation [1,2]. Thus, the establishment of bamboo plantations has been promoted throughout the tropics, especially among smallholders [3]. In terms of species selection, certain species may have superior timber properties or higher market demand, thereby providing better income opportunities. However, care must be taken to ensure that the chosen species is well-suited to the local environment to maximize growth and productivity [4]. Its rapid growth, regenerative ability derived from its rhizome, and versatile uses enable smallholders to earn a regular income by harvesting shoots and culms, especially in combination with agricultural crops [5,6]. However, the fast-growing nature of bamboo depends on substantial nutrients, and regular harvesting may lead to nutrient depletion and decreased productivity over time, particularly on nutrient-poor sites [7,8]. As bamboo species are increasingly introduced outside their native range due to their economic importance, some might become invasive and require monitoring [9,10]. Thus, understanding nutrient dynamics within bamboo plantations of both native and non-native species is crucial for developing and promoting sustainable management practices that protect local biodiversity, soil fertility, and rural livelihoods.

Bamboo culms do not undergo secondary growth but reach their full size within their first growing season. The resulting aboveground biomass (AGB) can range from 25 to 300 Mg ha⁻¹ depending on the species, spacing, stand age, nutrient and water management, as well as harvesting [4]. Concerning timber production, culms are harvested when they exhibit optimal physical stability and flexibility, characterized by dense cell walls without deterioration. This typically occurs at the age of 3–4 years [4]. Younger culms are characterized by high water content and limited strength, whereas older culms become more susceptible to fungi infestations, losing stability. This rapid development is fueled by efficient nutrient uptake, primarily of the macronutrients (N, P, K, Ca, and Mg) required for physiological processes like photosynthesis, energy transfer, and cell wall formation. Consequently, the availability of nutrients in the soil is positively related to yield [11,12]. In addition, effective bamboo management involves regular harvesting of mature culms to prevent overcrowding, which could otherwise trigger malformation and reduced yields [13]. Therefore, regulating standing-culm density and culm-age structure is pivotal [14]. Due to its extensive rhizome system, bamboo swiftly resprouts, enabling continuous productivity [15,16]. This high productivity paired with efficient nutrient use can allow bamboo stands to maintain soil fertility even under regular harvesting [16]. However, it should be noted that bamboos have a high demand for nutrients, particularly during their rapid growth phase in the first year, indicating a need for effective nutrient management strategies.

Bamboo influences soil fertility through its unique biogeochemical processes and biomass allocation patterns. Its nutrient cycling abilities make it an asset in rehabilitating degraded land and enhancing productivity in various crop systems. Bamboo allocates most of its biomass to culms (70%–80%) and branches (13%–20%), while leaves constitute only 7%–10% of the total plant biomass [17]. However, due to their high nutrient concentration, leaves are a major nutrient reservoir [7]. The decomposition of bamboo leaves and nutrient mineralization from litter occurs generally faster in tropical climates, where 50% of litter biomass decomposes within one year, adding to the soil organic matter (SOM) [18]. Nevertheless, the harvesting and extraction of culms reduces the soil nutrient content. Although fertilization can boost bamboo productivity in commercial bamboo plantations or nutrient-poor sites, optimal rates vary and should be considered alongside the short- and long-term impacts on productivity and site conditions [19]. The productivity of bamboo is influenced by various soil properties, including its chemical and physical characteristics. Although bamboo tends to be less impacted by acidic soils compared to other crops, its productivity is still affected by the reduced availability of nutrients [20]. Hence, soil improvement measures can enhance bamboo productivity by optimizing these properties.

Bamboo plays a significant role in the socio-economic and environmental landscape of Thailand, serving as a multipurpose resource for the local communities [21]. Smallholders widely cultivate it for its rapid growth and diverse uses, including food, timber, energy, and ornamental uses. Bamboo woodlots are typically managed through a selective cutting system in which only mature, weak, or dead culms are removed. It is labor-intensive but provides a steady yield maintaining a multi-aged stand [22]. The development of bamboo resources and their sustainable management is further supported by the Royal Project Foundation and the Royal Forestry Department [21]. One initiative was to experimentally introduce non-native bamboo from Taiwan to study their potential for biomass production and restoration of degraded areas focusing on species with larger culm dimensions [23,24].

The aim of this study was to compare the biomass and nutrient dynamics within experimental bamboo plantations of native and non-native species. Firstly, the potential nutrient loss due to bamboo harvesting was investigated relative to soil nutrient stocks in order to better understand the nutrient balance of bamboo plantations and their long-term sustainability. Secondly, the nutrient use by the bamboo plants and the overall nutrient cycling was analyzed to understand how efficiently nutrient inputs are converted into biomass. The study was conducted in 20-year-old monospecific plantations of four sympodial or clumping bamboo species: Bambusa oldhamii Munro, Dendrocalamus latiflorus Munro, both native to subtropical East Asia, as well as Bambusa tulda Roxburgh and Dendrocalamus brandisii (Munro) Kurz, indigenous to Southeast Asia but also found in Southern China [25]. B. oldhamii thrives in porous soils rich in SOM, reaching 10–16 m in height with a 3–9 cm culm diameter, producing around 10–11.7 t ha⁻¹ yr⁻¹ of culm with AGB of 104 Mg ha⁻¹ [26]. D. latiflorus reaches a height of 14–25 m, with culm diameters of 5–20 cm, and prefers moist, sandy loam soils, sequestering 14.2 MgC ha⁻¹ yr⁻¹ [27,28]. B. tulda prefers fine-textured, moist alluvial soils reaching heights of 8–30 m and AGB ranging from 73 to 127 Mg ha⁻¹ [29]. D. brandisii, a large bamboo species, achieves a height of 19–33 m, with culm diameters of 13–20 cm [30]. Their altitudinal distribution varies as D. latiflorus is found between 100 and 1000 m, B. oldhamii approximately reaches up to 1600 m, B. tulda is found between 600 and 1500 m, and D. brandisii grows up to 1300 m [31]. For each bamboo stand, the nutrient stock in the living culms and in the soil was calculated by estimating the AGB of all culms, the relative biomass allocation of each plant component, and the nutrient concentration in each component. Thus, this study provides a comprehensive assessment of nutrient dynamics in bamboo plantations and their implications for sustainable bamboo management and soil fertility.

2. Materials and Methods

2.1. Study Site

Our study was conducted in the Royal Angkhang Agricultural Station located in the Chiang Mai province (19°50′–19°57′ N, 99°01′–99°03′ E). It is nestled at an elevation ranging between 1400 and 1920 m above sea level with an average annual temperature standing at 18.9 °C (13.5 to 22.3 °C) and an annual rainfall averaging 1705.7 mm. The station is characterized by two primary soil groups, siltstone and weathered shale. The soil texture varies from clay loam to clay, enriched with a high amount of organic material [32]. Prior to the station’s establishment, the area was characterized by deforestation and the subsequent transformation to an herbaceous grassland dominated by Eupatorium adenophorum Spreng. and Imperata cylindrica (L.) P.Beauv. In 1993, four bamboo plantations featuring B. oldhamii, B. tulda, D. brandisii, and D. latiflorus were established in this area using separated rhizomes. The bamboos were planted at 5 m × 5 m spacing. Subsequently, the stands were managed for timber production as culms were harvested selectively after reaching the age of four years, while leaves and branches were removed during the harvest and left on site. Apart from these post-harvest residues, no additional nutrients were applied.

2.2. Data Collection

2.2.1. Forest Survey

Within each bamboo stand, two square sample plots were established, each encompassing an area of 400 m². We measured and recorded the diameter at breast height (DBH) for each bamboo culm, categorizing them into four age classes: 1 year, 2 years, 3 years, and 4 or more years. The age of the culms was determined based on distinct morphological characteristics, with the assistance of experienced local staff who were well-acquainted with these species and their growth patterns. Generally, younger culms have a more vibrant color and smoother texture, while older culms exhibit faded coloration, rougher texture, fungal growth, and developed branches. Data collection was carried out in two phases. The initial phase spanned from March to April 2013, followed by a second phase from January to April 2015. It should be noted that the B. oldhamii stand was only measured once in 2015. These data collection periods correspond to 20 and 22 years, respectively, post the initial establishment of the bamboo stands.

2.2.2. Biomass Sampling

Allometric equations were developed to estimate the AGB of the four bamboo species. In each plot, culms were sampled for biomass measurements based on a stratification of culm age (4 classes) and DBH (5 classes). In each DBH class, for every bamboo species, two culms from each age class were harvested as close to the rhizome as possible. The sample size comprised 80 culms for each species—40 in 2013 and 40 in 2015. However, B. oldhamii was only sampled in 2015, with a total of 40 culms. For each harvested culm, we measured DBH and height, and determined the fresh weights of the culms, branches, and leaves in the field. To determine the relative moisture content and dry weight, culm, branch, and leaf samples were oven-dried at 80 °C for 24–48 h or until the sample weight remained constant. These procedures were conducted at the Forest Soil Science Laboratory, Department of Silviculture, Faculty of Forestry, Kasetsart University. The total AGB of individual culms was calculated as the sum of the dry weight of all aboveground components, including culms, branches, and leaves.

2.2.3. Nutrient Concentration in Plant Components

For each biomass sample, the nutrient content, specifically, nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg) was determined subsequently. N-content was determined using the Dumas method or dry combustion technique using a PerkinElmer Model 2400 Series II CHNS/O Elemental Analyzer (PerkinElmer, Waltham, MA, USA). P, K, Ca, and Mg were extracted through wet ashing using a mixture of HNO₃ and HClO₄ acids in a 5:2 ratio. P was analyzed using the vanadomolybdate yellow color method and a UV mini 1240 spectrometer at a wavelength of 440 nm. The contents of K, Ca, and Mg were determined using a PerkinElmer PinAAcle 900 F atomic absorption spectrometer. Nutrient concentrations were calculated based on the amount of nutrients relative to each sample’s dry weight.

2.2.4. Soil Sampling and Analysis

Most bamboo rhizomes and roots are concentrated in the topsoil [33]. Mineral soil cores from five depths (0–10, 10–20, 20–30, 30–40, and 40–50 cm) were collected using a split tube sampler at three random locations from each replication of the bamboo species plantations, and soil samples were divided into two parts for analysis of the physical and chemical properties. Bulk density (BD) was determined using the core method [34]. A composite of three samples was prepared for each depth, air-dried, ground, and passed successively through 2 and 0.5 mm sieves. Soil texture was determined using the hydrometer method [35]. Soil pH was determined using a pH meter at a 1:1 soil-to-water ratio [36]. SOM was analyzed using the Walkley and Black rapid titration method [37]. Soil total N was measured using the Dumas method [38] and a CHNS analyzer (PerkinElmer 2400, Series II CHNS/O Elemental Analyzer). Available phosphorus (P) was extracted using the Bray II method [39] and analyzed using a spectrometer (UV mini 1240). Exchangeable K, Ca, and Mg were extracted using ammonium acetate (NH₄OAc) 1 N pH 7.0 [40] and analyzed using an atomic absorption spectrometer (PerkinElmer, PinAAcle 900 F). The total amount of each macronutrient in the topsoil (kg ha⁻¹) was calculated by multiplying BD (kg m⁻³), nutrient concentration (kg kg⁻¹), the depth of the soil layer (0.5 m), and the area (1 ha).

2.3. Data Analyses

2.3.1. AGB and Biomass Allocation

In this study, we developed species-specific allometric equations to estimate the AGB of individual bamboo culms. The selection or combination of input parameters included DBH, log-transformed DBH, age of the culms, and a combined measure of D²H. We considered a range of model forms including linear, power, and exponential equations [41]. Additionally, we also tested log-transformed versions of the linear and power models to ascertain if these provided a better fit to the data. A weighted fitting process was used for all models [42]. For model assessment, we employed the Monte Carlo cross-validation method, which involves randomly splitting the dataset into a training set (80%) and a remaining validation set [42]. This process was repeated 100 times to ensure robustness. The performance of each model was assessed using Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), adjusted R², the bias, the Root Mean Squared Error (RMSE), and the Mean Absolute Percentage Error (MAPE, [41]). Models that ranked best across all metrics were selected to predict the AGB of the respective stands.

For each sampled culm, the biomass allocated to each aboveground component—culms, branches, and leaves—was estimated by calculating the relative proportion of each component relative to the total AGB based on their dry weight. Subsequently, species-specific Ordinary Least Squares (OLS) models were developed to estimate the proportion of plant components based on the culm age. Culm DBH did not significantly affect biomass allocation and was not incorporated into the model. The final model was used to estimate the biomass allocation among all culms within the sample plot.

2.3.2. Biomass Nutrient Concentration, Stock, and Use Efficiency

Using the sampling data, species-specific linear regression models were developed to estimate the concentration of each macronutrient (N, P, K, Ca, and Mg) in each plant component based on culm age. The previous estimations of AGB, plant component allocation, and nutrient concentration were used to estimate the nutrient stock in the living biomass for each culm. This was carried out by multiplying the total AGB of culms, the proportion of plant components, and the nutrient concentration in the respective component. Subsequently, the nutrient stock of harvestable culms (4 yr), remaining litter following harvesting, and younger stand biomass (<4 yr) could be estimated for each species and plot to assess the potential nutrient removal through harvesting. By estimating the nutrient stock in the living biomass, the nutrient use efficiency (NUE) of the aboveground components could be calculated for each bamboo stand by dividing the total stand biomass by the total stock of nutrients [43]. NUE is a measure of how effectively a plant uses available nutrients to produce biomass and can indicate different nutrient use strategies [44]. The entire analytical process is outlined in Figure A1.

2.3.3. Statistical Analyses

The Kruskal–Wallis test was used to test species-specific differences in growth parameters, productivity measures, soil conditions, and plant nutrient concentrations. Dunn’s test for pairwise multiple comparisons was used as a post hoc analysis to compare all possible pairwise differences. Non-parametric tests were used as most properties were not normally distributed according to the Shapiro–Wilk Normality Test. All analyses were performed in R version 4.3.1 [45] using the package rstatix [46].

3. Results

3.1. Aboveground Biomass

Remarkably, the best-fitting allometric models for AGB predominantly relied on culm DBH and age, rather than height, which is often difficult to measure accurately and therefore prone to error (

Table A1. Allometric equations for aboveground biomass of four bamboo species ranked according to their overall goodness-of-fit. All intercepts and coefficients were highly statistically significant (p < 0.001).

Species	Model	Predictors	BIC	Adj. R2	Bias	RMSE	MAPE	Intercept	Coef. (1)	Coef. (2)
B. oldhamii	(1) Log-Transformed Linear	logDBH, logAge	12.590	0.914	2.623	0.476	15.667	−1.877	1.936	0.358
	(2) Power	D2H, Age	32.840	0.913	3.910	0.392	15.332	0.076	0.690	0.436
	(3) Power	DBH, Age	39.140	0.887	3.521	0.442	15.342	0.159	1.876	0.412
B. tulda	(1) Log-Transformed Linear	logDBH, logAge	−4.242	0.910	1.839	1.022	16.207	−1.515	2.133	0.209
	(2) Log-Transformed Linear	logDBH, Age	−3.807	0.911	2.542	1.053	16.516	−1.584	2.131	0.096
	(3) Log-Transformed Linear	logDBH	6.721	0.889	1.841	1.170	18.902	−1.241	2.048	-
D. brandisii	(1) Log-Transformed Power	D2H, Age	−31.080	1.000	0.974	8.244	13.242	0.608	0.215	0.050
	(2) Log-Transformed Power	logDBH, Age	−27.690	1.000	1.651	8.578	14.323	0.836	1.538	0.054
	(3) Log-Transformed Power	DBH, Age	−26.406	1.000	1.166	8.819	14.459	0.771	0.594	0.028
D. latiflorus	(1) Log-Transformed Linear	logDBH, logAge	6.628	0.860	3.708	2.393	18.656	−2.157	2.035	0.237
	(2) Log-Transformed Power	logDBH, Age	13.466	0.998	2.198	2.604	19.273	0.487	1.968	0.102
	(3) Log-Transformed Linear	logDBH, Age	7.309	0.856	3.792	2.522	18.928	−2.240	2.040	0.105

). The adjusted R² values of the top models ranged from 0.86 to 1, indicating that these models accounted for a high amount of variation in the data. Furthermore, bias values were low, suggesting little systematic error. RMSE and MAPE were also generally low, indicating that AGB could be predicted accurately with minimal error (Figure A2).

D. brandisii developed the largest culms, with D. latiflorus culms being intermediate in size. B. oldhamii and B. tulda exhibit small average diameters (

Table 1. Comparison of key growth parameters, productivity measures, and soil conditions among four bamboo species, including diameter at breast height (DBH), mean annual increment of aboveground biomass (MAI_AGB), and soil organic matter (SOM). Values are means with standard deviations in parentheses. Letters indicate significant differences between mean values of species as determined by Dunn’s test (p < 0.05).

	B. oldhamii	B. tulda	D. brandisii	D. latiflorus
DBH (cm)	3.36 (0.96) a	3.75 (0.88) b	13.92 (2.11) c	7.71 (2.05) d
Culm age distribution (%)	33.8/23.8/ 27.7/14.7	10.8/9.9/ 23.4/55.9	10.5/17.1/ 14.9/57.5	23.8/18.8/ 24.8/32.6
Culm density (ha−1)	4512.5	7475	1968.8	1862.5
New culms (culm ha−1 yr−1)	1283.3 (320) a	1097.9 (759.2) a	279.2 (97) b	418.8 (116.3) b
Harvestable culms (Mg ha−1)	2.48 (0.64) a	43.48 (16.16) ab	116.08 (55.35) b	10.64 (3.55) a
MAIAGB (Mg ha−1 yr−1)	1.83 (1.98) a	7.17 (3.14) bc	17.37 (6.75) b	4.8 (2.94) ac
SOM (%)	5.61 (3.85) a	2.79 (1.65) b	4.76 (2.35) ab	7.73 (2.71) a
pH	4.29 (0.19) a	5.38 (0.1) b	5.32 (0.21) b	4.72 (0.2) a

). Conversely, the culm density and the number of newly produced culms was higher in the B. oldhamii and B. tulda stands compared to the two Dendrocalamus species. Similarly, the age structures differed notably across species. While B. oldhamii had a relatively even distribution across age classes, which is only slightly skewed towards younger culms, the majority of culms in B. tulda, D. brandisii, and D. latiflorus are three or four years old, i.e., mature. However, despite lower culm density and new culm production, D. brandisii led in harvestable timber and mean annual increment of above-ground biomass (MAI_AGB), indicating its superior productivity per culm. The soil in all bamboo stands was dominated by a clay texture, but differed in terms of SOM and pH levels. The growth of the non-native species, B. oldhamii and D. latiflorus, was potentially affected by significantly lower pH levels. These species also showed higher levels of SOM.

3.2. Biomass Allocation

Among all four species, the total culm biomass increased significantly with growing age (Table A1). Furthermore, the proportional allocation to the culm component declined significantly as the plants’ age increased, while allocating progressively more biomass to their branches (p < 0.01; Figure 1). Thus, bamboo branches typically started to appear and grow after the culm had reached its full height. The patterns for leaf biomass allocation were somewhat mixed. For B. tulda, the allocation to leaves slightly increased over the years, but this trend was not statistically significant (p > 0.05). D. brandisii and B. oldhamii showed a small but statistically significant increase in leaf biomass allocation (p < 0.05). In contrast, D. latiflorus showed a significant increase in leaf biomass allocation over the years (p < 0.01). Thus, as these bamboo species matured, they tended to allocate more resources to their branches and leaves, potentially to maximize photosynthesis and reproductive output. Additionally, the relative amount of post-harvest remains, or litter, increased as culms aged. However, DBH had no effect on the relative biomass fraction (p > 0.05).

3.3. Nutrient Stock in Harvestable Biomass and Soil

Nutrient stocks varied substantially across different bamboo species, whereas nutrient distribution patterns remained similar across the different macronutrients assessed (Figure 2). D. brandisii demonstrated the highest nutrient stocks in the juvenile stand, mature culms, and post-harvest residue, indicating efficient nutrient uptake and turnover rates. However, compared to other species, D. brandisii had lower topsoil stocks for all macronutrients. Nevertheless, most differences in nutrient stocks were not statistically significant (ANOVA, p > 0.05). Only, B. oldhamii showed significantly lower N-stocks and higher Ca-stocks compared to B. tulda (Tukey HSD, p < 0.05). B. tulda exhibited a more balanced nutrient distribution, with moderate levels in the biomass relative to soil nutrient stocks. Mirroring their low levels of AGB, B. oldhamii and D. latiflorus showed lower nutrient stocks in their biomass, particularly harvestable culms, but appeared to maintain high nutrient stocks in the topsoil. Across all species, the comparison of harvestable culms and post-harvest residue composed of leaves and branches shows that leaves and branches store a high level of nutrients. Overall, these results indicate species-specific differences in nutrient uptake, storage, and cycling. Moreover, the difference in nutrient stocks between juvenile (age 1–3 years) and mature culms (4 years) is likely also a consequence of the skewed age structure (Table 1).

3.4. Nutrient Concentration and NUE

Low nutrient concentrations may indicate deficiencies. B. oldhamii generally had lower nutrient concentrations in all plant components compared to the other species but showed high N- and K-concentration in its leaves (2.7%; Figure 3). Moreover, D. brandisii and D. latiflorus generally had higher concentrations of most nutrients in all plant parts compared to B. oldhamii and B. tulda. Thus, each species had a distinct nutrient profile suggesting species-specific differences in nutrient uptake and utilization strategies, or other physiological factors. Across all species, the concentration of N, K, Ca, and Mg tended to be higher in leaves compared to culms and branches. In contrast, P was found at similar concentrations across all parts.

For N, P, and K, D. brandisii consistently demonstrated the highest NUE across all nutrients, while accumulating the highest AGB, indicating that the efficient utilization of these nutrients likely supported their growth (Figure 4). In contrast, B. oldhamii exhibited the lowest NUE for P and N, and an intermediate NUE for P. Despite its higher NUE for N, its AGB was lower than that of B. tulda and D. latiflorus, suggesting that other factors might be impacting its biomass productivity. B. tulda and D. latiflorus showed intermediate levels of both NUE and AGB.

4. Discussion

While bamboo cultivation offers numerous environmental benefits and income opportunities, rapid growth and harvesting can lead to nutrient depletion, particularly on nutrient-poor sites. Additionally, the introduction of both native and non-native bamboo species, driven by economic factors, requires careful monitoring to prevent invasive behavior and maintain sustainable practices. Thus, this investigation of the nutrient dynamics within bamboo plantations of both native and non-native species may enhance our understanding of the conversion efficiency of nutrient inputs into biomass and the potential nutrient loss from harvesting. The species-specific comparison of productivity among four bamboo species revealed significant differences between native and non-native species. D. brandisii and B. tulda, both native species, demonstrated higher productivity and greater AGB and grow well in moist tropical regions with temperatures of 20–27 °C [47]. In contrast, the non-native species, B. oldhamii and D. latiflorus, demonstrated lower productivity despite maintaining high nutrient stocks in the soil. Moreover, the standing biomass and productivity of the studied plants did not nearly reach the values reported in the literature [26,48]. Instead of insufficient soil nutrient stocks, our results indicate that other environmental factors might be at play, emphasizing the complexity of plant–soil interactions. Although bamboos are generally more tolerant to acidic soils, the observed low soil pH levels (4.3–4.7) likely limited nutrient availability and, consequently, productivity. The low pH among B. oldhamii stands (4.3) likely prevented the P-uptake as indicated by the significantly lower P-concentrations in the leaves. However, such nutrient deficiencies were not as evident among D. latiflorus. Thus, other environmental factors causing such lower productivity could include factors like the suitability of the local climate, the presence of pests or diseases, or over-harvesting [13]. For instance, it has been reported that heavy clay or acidic soils, as found in the study site, might not be suitable for D. latiflorus [31]. However, these non-native bamboo species seem less adapted to the specific growing conditions of the study site, potentially rendering them unsuitable. The considerable effort and expense required to enhance their productivity to match that of native species may not be justified given their apparent inferiority.

The high productivity of D. brandisii and B. tulda might lead to its preferred selection, but should also warrant caution as it could potentially increase harvest-induced nutrient losses, potentially reducing long-term soil fertility. While mature culms store most nutrients and are regularly harvested, leaves and branches are not extracted upon harvest. They decompose on-site, releasing their nutrients back into the soil, which contributes to maintaining soil fertility. D. brandisii exhibited the fastest annual litter decomposition rate among the studied species [18]. Nevertheless, this nutrient recycling process may not be sufficient to prevent nutrient losses induced by the regular harvest of mature culms. Among D. brandisii plants, litter and branches only accounted for 5% and 11.5% of AGB during harvest, similar to other bamboo species [4]. It also demonstrated the highest nutrient stocks in the juvenile stand, mature culms, and post-harvest residue. Its soil nutrient stocks, in comparison, were relatively low, suggesting a potential nutrient loss through harvest. Nevertheless, D. brandisii demonstrated consistently higher nutrient concentrations in its leaves across all nutrients, indicating a sufficient nutrient supply and, unlike B. tulda, the absence of potential nutrient deficiencies [4]. The high productivity and stable soil fertility of D. brandisii might be a consequence of its high NUE across all nutrients. Thus, D. brandisii was particularly efficient at utilizing the available nutrients to support growth, which could mitigate the potential risk of nutrient depletion, and sustain its productivity even on nutrient-poor sites. Nevertheless, the management of this species should focus on reducing soil nutrient depletion and maintaining its high productivity. If necessary, nutrients could be replaced using organic amendments, applying fertilizers, practicing intercropping, or adjusting the timing and intensity of harvests to minimize nutrient loss [49].

In most stands, including D. brandisii, the age distribution of culms was unbalanced with most culms reaching maturity and senescence. This could present a complex challenge for the management of bamboo plantations considering both nutrient dynamics and economic viability. A large-scale harvest of mature culms could lead to a considerable decline in overall biomass and the sudden removal of a substantial amount of nutrients [13]. The absence of mature culms would impact the microclimate and could even lead to increased soil erosion. With varying nutrient stocks at different ages, this would also disrupt the nutrient distribution across the stand. Specifically, an abrupt removal could lead to nutrient deficiency among remaining and newly sprouting culms, which often access nutrients from senescent culms [15]. Furthermore, an unbalanced age structure poses a serious challenge for smallholders who depend on the harvestable timber from these plantations to maintain a steady flow of income [50]. Addressing this imbalance involves modifying the stand’s age structure either by felling culms at specific ages or varying the length of the felling cycle. Notably, both structures skewed towards old and young ages may constrain stand productivity reducing either photosynthetic capacity or photosynthetically active leaf area [4]. Consequently, a balanced age structure, maintained through strategic harvesting practices, is crucial for optimizing both the ecological sustainability and economic productivity of these bamboo plantations [14].

Our findings highlight the considerable benefits of developing allometric models for biomass estimation that rely predominantly on culm DBH and age rather than culm height. Such an approach not only simplifies the estimation process but also enhances its accuracy and reliability [51]. Culm height measurements can often be challenging due to the bending nature of bamboo culms, leading to inaccuracies that affect traditional biomass estimation models. In contrast, models centered on DBH and age offer a robust method for estimating biomass with minimal error, as supported by the high R² values and low bias, RMSE, and MAPE values shown here. Culm age, moreover, can be determined based on morphological characteristics, although it might require some training [17]. These models could provide improved information for managers that want to gain a better understanding of the biomass productivity and carbon sequestration potential of their plantations.

The rapid growth and potential for annual harvest make bamboo a viable alternative for smallholders seeking regular income. However, it might be necessary that some of their income is reinvested into nutrient management to sustain the yield of their bamboo woodlot. As fertilizers are often too expensive for smallholders, alternative practices such as agroforestry, involving the integration of nitrogen-fixing or nutrient-pumping trees, might maintain or improve soil fertility [16,52,53]. Diversification, with a focus on native bamboo species due to their apparent higher productivity, is recommended, whereas non-native species should be carefully selected, and their ecosystem impact monitored.

This study, while providing valuable insights, has several limitations. Firstly, the scope of the study, both in terms of sample size and geographic area, could restrict the generalizability of the findings to other environments and bamboo species. However, the focus on four bamboo species provides a comprehensive view of these species’ behaviors under the studied conditions. Secondly, estimating the total AGB and nutrient stock of each stand relied on modeling which may contain uncertainties [42]. Yet, the robust models utilized in this study, which demonstrated high explanatory power and minimal error, were instrumental in overcoming these uncertainties. Thirdly, belowground biomass (BGB) was not included in this study despite representing approximately a third of the total biomass and despite a substantial nutrient stock playing an important role in nutrient uptake and soil nutrient cycling [4]. Nevertheless, the study presents a comprehensive analysis of AGB that is crucial in understanding the productivity and carbon sequestration of these bamboo species. Fourthly, the growth and nutrient dynamics could not be monitored due to the destructive sampling methods necessary. Instead, culms were sampled across all age classes. Finally, estimating the age of individual culms based on morphological characteristics may be prone to errors [17]. Here, we relied on the local knowledge of the plantation manager, which enhanced the reliability of age-based comparisons.

Bamboo plantations and their nutrient dynamics should be further investigated to ensure the sustainable management of bamboo plantations in Thailand. Besides increasing the overall scope to increase the sample size and environmental conditions, a better understanding of the underlying mechanisms would be needed to improve our general understanding of the plant–soil interactions. Such investigations might enable researchers to develop more targeted management recommendations for plantation managers and owners. Specifically, long-term experimental studies would be beneficial as these would allow manipulating soil nutrients, culm density, and age structure, as well as harvesting timing and intensity. A robust comparison of different bamboo species and their impact on the soil would only be possible by controlling and altering these conditions.

5. Conclusions

This study investigated nutrient dynamics within bamboo plantations of both native and non-native species, aiming to enhance our understanding of the conversion efficiency of nutrient inputs into biomass and the potential nutrient loss from harvesting. The findings revealed significant differences in productivity among the four bamboo species studied, with the native species, D. brandisii and B. tulda, demonstrating higher productivity and NUE compared to the non-native species, B. oldhamii and D. latiflorus, potentially due to the superior adaptation of native species to the local conditions. These insights underscore the need for careful species selection in bamboo cultivation to optimize productivity and sustainability, as the introduction of less suited, non-native species may lead to suboptimal productivity and potential nutrient depletion, even with high soil nutrient stocks. The study contributes to the field by introducing robust allometric models for biomass estimation, relying predominantly on culm DBH and age, enhancing the accuracy and simplifying the estimation process. Despite these advancements, the study acknowledges its limitations, such as the restricted scope and sample size, potential uncertainties in modeling, and the exclusion of BGB from the analysis. Future research should expand the scope, include BGB, and employ long-term experimental studies manipulating various factors like soil nutrients, culm density, and age structure, to provide a more comprehensive understanding of bamboo plantation management. Such research is also necessary to develop better management recommendations, promoting the cultivation of native bamboo species over non-native ones, incorporating age structure balancing in harvest strategies, and considering alternative nutrient management practices like agroforestry.