A Biomass Estimation Model for Nondestructive Estimation of Guava Tree Biomass

doi:10.21203/rs.3.rs-2310563/v1

Download PDF

Research Article

A Biomass Estimation Model for Nondestructive Estimation of Guava Tree Biomass

https://doi.org/10.21203/rs.3.rs-2310563/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The current method available for nondestructive estimation of tree biomass using diameter at breast height(DBH) is not applicable to fruit trees which are manipulated to branch at a very low height in case of grafted plants or air layered or cuttings generated plants. Therefore guava trees of commercial orchards require independent allometric equation for nondestructive estimation of tree biomass. Guava tree specific allometric equation was developed using destructive sampling using parameters other than DBH and compared with that equation developed for grafted mangos and sapota. The selected allometric parameters were significantly related with the age of the trees. The root to shoot ratio also differed from those reported for forest trees. The biomass expansion factor (BEF) by and large attained stability beyond 20 years of tree age. The equations so developed generally fitted the data well, and in most cases more than 50per cent of the observed variation in biomass was explained by primary branch girth (PBG) × number of primary branches (NPB). There was a good agreement between the observed and the predicted biomass using both multiple linear regression(MLR) and power model equations. Further a comparison was also made to see if mango and sapota equations can predict guava tree biomass. The results of this study confirmed that mango and sapota equations grossly underestimates guava tree biomass by about 50 per cent and hence mango or sapota tree allometric equation cannot be used for estimating guava tree biomass.

allometric equation

biomass expansion factor

grafted trees

guava

mango

Nondestructive method

Tree biomass

The basic requirement for any attempt to estimate carbon sequestration(CS) of a fruit tree is a method without felling to estimate the biomass of standing orchard trees. Allometric biomass models are the tools used for such a nondestructive estimation of tree biomass. These are simple mathematical models developed by using independent variables. Most of the available equations are devoted to forest trees and the same is not applicable to orchard fruit crops such as guava trees(Ganeshamurthy et al, 2016). Nevertheless they are the most adopted methods for tree biomass estimation across the world (Ketterings et al. 2001). Use of such models taking only tree girth irrespective of DBH of 1.2m (which is not possible in orchard fruit trees) lead to compromise on the accuracy.

The carbon-dense nature of tropical fruit trees such as guava are high carbon sequesters. Large conversion of agriculture land in to fruit orchards is a win-win situation to grab the opportunity to sequester carbon in a large scale. Destructive methods are the most accurate methods. But while using destructive methods both live trees and time have to be sacrificed (Ketterings et al.2001). This paper reports biomass distribution pattern and development of guava specific allometric equations and also attempt extrapolation of mango or sapota allometric equation to guava for estimation of CS in guava orchards.

Method followed by Ganeshamurthy et al.(2016) for developing allometric equation for grafted mango and sapota was used as the basis in this study. They utilized a basic tree parameter common to all orchard fruit trees where canopy architecture is managed for convenience through branching of trees near the ground level to maintain a low tree height (Ganeshamurthy et al 2016). Same methodology was used in this study to develop guava specific allometric equation by destructive sampling and using the estimated tree biomass for developing the equation. Further this was also validated by using mango and sapota allometric equations to verify whether this mango or sapota allometric equation can be extrapolated to guava.

2.1. Computation of tree biomass: Twenty guava trees of popular variety Allahabad Safeda, Sardar, L-49, in the age group between 9 years to 41 years were selected, some from the research institute farm in Bengaluru and at Chettalli and some from the farmer’s fields. The trees were destructively sampled by felling and completely uprooting the trees after recording the required allometric data. This information was used to generate data on biomass distribution and to develop guava allometric equations and also to test whether mango or sapota equation can estimate the tree biomass equally well to that of guava specific equation developed under this study.

The measurement included the number of primary branches(NPB) and the girth of primary branches(PBG), tree height and tree canopy spread. After felling the trees different plant parts were separated such as primary, secondary branches, twigs, leaf and trunk and the root. Sub samples were collected for moisture and carbon estimation. The total above ground and below ground biomass of the felled trees was then computed. Different statistical models were used to estimate tree biomass like multi linear regression model(MLR) and power model. These two models are a class of nonlinear regression model, as the derivatives of Y_t with respect to unknown parameters are functions of either of them, suitable nonlinear estimation procedure was followed for parameter estimation(Seber and Wild,1989). SAS codes were developed to fit these nonlinear regression models. Based on the best fit, the multilinear and power models were used for the estimation of tree biomass. The power model is represented by the following equation

Y_t = ε_t + _aX_t^t ,

Where Y_t is the t^th trees AGB (above ground biomass), X_t the t^th trees observations on PBG (primary branch girth) × NPB (No. of primary branches), ε_t the error terms corresponding to difference between observed and expected tree AGB of t^th tree.

2.2. Root : Shoot ratio was worked out from the data generated after felling the trees.This ratio was found to be 1 : 0.29. For below ground biomass(BGB) estimation, it was estimated using the ratio of 1 : 0.29 as found in this study and also reported by Ganeshamurthy et al.(2016) for mango. The samples were dried, powdered and estimated the carbon using a CHNS analyser (Elementar). Regression models were used to test the validity of mango and sapota allometric equation in guava by relating the AGB estimated through destructive sampling and AGB predicted using mango allometric equation to validate the mango allometric equation for guava.

3.1. Component Biomass Distribution: The tree biomass distribution pattern showed that major aboveground biomass is accumulated in the stem and primary branch wood with dry weight on an average representing, 45.8 per cent of the total above ground biomass (Table 1). The secondary branches and twigs constituted about 17.6 per cent and the foliage accounted for 7.6 per cent. All put together the AGB accounted for 71.1 per cent and the remaining 28.9 per cent is accounted for the roots that included all small, medium and large roots.

Table 1

Component dry weight (Kgs) of harvested guava trees of different age
Sl.No.	Tree Age (years)	Main Stem and Primary Branches	Secondary Branches	Foliage	Total above Ground Biomass (kg)	Roots (kg)	Total tree biomass,(kg)
1	9	66.64	29.31	11.49	151.04	43.80	194.84
2	10	101.73	40.70	17.78	233.64	67.76	301.40
3	10	54.46	23.68	9.38	123.31	35.76	159.07
4	11	105.09	44.21	18.00	236.59	68.61	305.20
5	11	91.5	35.94	15.27	200.60	58.17	258.77
6	11	140.74	54.55	23.39	307.39	89.14	396.53
7	14	81.60	30.96	13.34	175.23	50.82	226.05
8	15	95.50	36.50	15.89	208.86	60.57	269.43
9	17	86.34	29.86	13.92	182.90	53.04	235.94
10	17	67.11	23.97	10.91	143.37	41.58	184.95
11	20	98.48	37.20	16.25	213.58	61.94	275.52
12	21	139.29	52.62	23.00	302.08	87.60	389.68
13	22	103.92	39.26	17.15	225.38	65.36	290.74
14	22	70.73	26.72	11.67	153.40	44.49	197.89
15	22	108.27	40.90	17.87	234.82	68.01	302.92
16	22	111.268	42.04	18.36	241.31	69.98	311.29
17	26	136.57	51.60	22.54	296.18	85.89	382.07
18	29	118.34	44.71	19.53	256.65	74.43	331.08
19	37	116.44	44.00	19.22	252.52	73.23	325.75
20	42	121.32	45.84	20.02	263.14	76.31	339.45
Share of total tree biomass		45.8	17.6	7.6	71.1	28.9

3.20. Relationship between Tree Age and Allometric Parameters: The tree age wise mean and standard deviations of the biometric parameter PBG x NPB of the thirteen age groups examined is presented in Table 2. With tree age as independent variable the changes in PBG x NPB with time was predicted by applying logarithmic regression model:

Y = a ln(X) – b. The results of the best predictive growth models is presented in Figs. 1.

Table 2

Biometric parameters of the different age trees examined in grafted guava
Tree age	No of trees	PBG × NPB
Tree age	No of trees	Mean	SD
9	17	17.78	6.73
10	14	20.04	5.98
11	21	19.74	7.09
14	19	28.65	7.88
15	21	31.50	7.08
17	17	32.46	9.62
20	14	34.38	7.82
21	19	36.30	9.75
22	24	38.19	9.54
26	16	47.12	8.77
29	19	46.96	8.67
37	21	44.39	8.34
42	20	41.07	9.10

3.30. Biomass Expansion Factor (BEF): The BEF of different tree age groups was estimated as the ratio of the biomass to the volume, resulting in a dimensional variable and expressed in Mg m^− 3. The BEF of guava varied with age of the trees (Table 3). It was found that at 9th year the BEF ranged from 0.426–1.291 with a mean of 0.778 Mg m^− 3. Gradually with age the data indicated a decreasing trend and attained a steady state. At 22nd year the BEF ranged from 0.234–0.632 with a mean of 0.286 Mg m^− 3. Beyond this there was fluctuation in the trend.

Table 3

Biomass expansion factor of grafted guava.
Age	BEF(Mg m^-3)
Age	Range	Mean	SD
9	0.426–1.291	0.778	0.564
10	0.507–1.107	0.796	0.559
11	0.386–0.913	0.589	0.648
14	0.441–0.871	0.495	0.436
15	0.282–0.763	0.407	0.392
17	0.210–0.590	0.364	0.322
20	0.222–0.578	0.381	0.243
21	0.201–0.534	0.302	0.242
22	0.234–0.632	0.286	0.254
26	0.232–0.636	0.271	0.273
29	0.210–0.590	0.285	0.230
37	0.188–0.702	0.296	0.224
42	0.104–0.865	0.316	0.289

3.40. Relationship between AGB and BGB: The best estimates of BGB is obtained by destructive methods(Ganeshamurthy et al.2016., Rupa et al. 2022). Through excavation it was found that the root is 0.29 times that of AGB. This conversion factor of 0.29 obtained in this study was used to calculate the BGB(Table 2). With sample trees it was observed that major part of the root biomass (80 ± 5per cent) accumulated in the first 0.75 m from the tree stumps.

3.50. Biomass estimation: Preliminary scatter plot was used to examine the data set. In this study two forms of models viz., multiple linear regression model and power model (yi = a(X)b) were used to estimate the tree biomass where y = biomass of tree and “a” and “b” are scaling factors. The predicted biomass obtained by these two models is presented in Figs. 2 and 3. Both the equations were statistically significant (p < 0.05) for both scaling parameters, “a” and “b”. The MLR and Power models are presented below:

MLR model : AGB = 40.51–0.78 PBG + 5.56(PBGxNPB) R² = 0.941

Power model : AGB = 11.52(PBGxNPB)^0.839 R² = 0.940

Based on the R² values both MLR model and the Power model fitted equally well for estimation of above ground biomass of orchard guava trees.

Biomass estimation was also made using mango and sapota power models developed by Ganeshamurthy et al.(2016., Rupa et al. 2022) to see if a common equation can be used to predict tree biomass of all fruit species. The results are presented in Fig. 4.

This comparison showed that both mango and sapota specific equations grossly under estimated the guava tree biomass by about 40 to 50 per cent.

Absence a reliable method for nondestructive estimation of tree biomass in orchard plants is responsible for poor data base on carbon sequestration by fruit orchards (Ganeshamurthy et al. 2016, Rupa et al. 2022). The authors have earlier developed allometric equations for nondestructive estimation of tree biomass in orchard mango and sapota(Ganeshamurthy et al, 2016, Rupa et al. 2022). In this study the equation developed for guava orchards is reported.

In guava orchards like in other fruit crops the stem and primary branches accounted for the largest proportion of the total aboveground biomass by weight. The stem wood below primary branching in orchard fruit trees are very low because all the orchard trees are allowed to branch at a very low height from the ground or from the graft union which is kept just above the ground while planting the seedlings. The major wood is accumulated in primary branches. Therefore the biomass of both main stem below branching and primary branches was combined for calculating its distribution. Ganeshamurthy et al(2016) reported similar distribution in grafted mango trees and in guava trees (Rupa et al. 2022). The biomass distribution in guava changed with age of the trees. The stem and primary branches accounted for 44.14 per cent in young trees (9 years age) and showed an increasing trend with age. It reached about 47 per cent after 17 years. This followed a declining trend in the proportion of secondary branch wood with age and attained stability at about 46 per cent. The secondary branch wood slightly declined with age in young trees (9–10 years) and became stable after 17 years of age. This is perhaps due to increasing wood density of stem and primary branches with age and also by and large constant density of secondary branch wood. The other reason for this might be the practice of pruning the secondary branches for ease of management.

4.10. Relationship between Tree Age and Allometric Parameters: With tree age as independent variable the changes in PBG x NPB with time was predicted by applying several equations to select an appropriate growth model. Mori(2001) proposed logarithmic and nonlinear exponential equations for predicting the time scale changes. Therefore we first tested this as this equation showed a good prediction in other environments. The logarithmic regression model was therefore applied to predict PBG X NPB from age:

Y = a ln(X) – b.

The results of the best predictive growth models is presented in Figs. 1. Using this equation we could predict the relationship between tree age and the identified tree allometric parameter PBG x NPB. This showed that allometric parameter was significantly related with age of the trees(r = 0.858).

4.20. Biomass Expansion Factor (BEF): Biomass Expansion Factor is commonly used in selviculture to directly estimate the merchantable biomass (t/ha). This helps in trade to know the dry weight of the merchantable volume of the growing stock and to estimate the size of the non-merchantable components. In this study the purpose of calculating BEF is different from selviculture. Here this was needed as a complement of growth models that do not include biomass predictions, but to reduce the uncertainty associated with the use of BEFs for biomass estimation. It was reported by Ganeshamurthy et al(2016) that Initially the BEF is very high followed by a decreasing phase and finally a steady phase. In this study we missed out this initial phase of very high value between first year to 9th year. Here the data started with 9th year when the tree was in economic bearing. It was found that at 9th year the BEF ranged from 0.426–1.291 with a mean of 0.778 Mg m^− 3(Table 3). Gradually with age the data indicated a decreasing trend and attained a steady state. At 22nd year the BEF ranged from 0.234-0.632with a mean of 0.286 Mg m^− 3. Beyond this there was fluctuation in the trend. This is because as the tree grew up the canopy volume varied with the extent of pruning. In guava unlike mango the pruning is a general practice and specifically the outer pruning is minimal because of feasibility of cutting the branches which have overgrown leading to the excess spread of canopy both east-west and north-south direction. The BEF by and large attained stability beyond 22 years. Such similar observations in other species were made by several authors (Ganeshamurthy et al.2016, Rupa et al. 2022., Tobin et al.2007). Unlike forest trees grafted fruit trees are subjected to various canopy management practices leading to fluctuations in BEF values. Hence the BEF is not a reliable data in case of grafted fruit trees. Nevertheless these reports support the findings concerning resource allocation during the growth process.

4.30. Relationship between AGB and BGB: The best estimates of BGB is obtained by destructive methods(Ganeshamurthy et al.2016., Rupa et al. 2022). BGB is estimated for several purposes. But in this study the purpose was for carbon storage estimation. Exploratory studies conducted on several fruit trees including guava showed that the values matched with those of values recorded by us for mango(Ganeshamurthy et al 2016). Therefore in this study the conversion factor of 0.29 obtained in this study and that reported by Ganeshamurthy et al.(2016) for mango was used in guava as well(Table 2). With sample trees we observed that major part of the root biomass (80 ± 5per cent) accumulated in the first 0.75 m from the tree stumps. The error associated with the chosen excavation area in the drip circle of the trees was considered to be relatively small as the measurements made in this study suggests that root biomass stock would appear to reduce exponentially with the distance from the tree stump. IPCC GPG(Sanesi et al.2013) reported the mean default value of R as 0.32 with a range of 0.24–0.50 for trees with aboveground biomass stock of 50–150 tones dry weight ha¹. Our measured values fell within the range reported in the literature(Cairns et al.1997., Sanesi et al.2013., Penman et al. 2004).

4.40. Biomass estimation: Allometric equations were developed for mango and sapota orchard trees by Ganeshamurthy et al.(2016, Rupa et al. 2022). With this experience two forms of models viz., power model (yi = a(X)b) and multiple linear regression model were used in guava where y = biomass of tree and “a” and “b” are scaling factors. Since results indicated suitability of both MLR and power models a comparison was made with these two models and presented in Figs. 2 and 3. Both equations were statistically significant (p < 0.05) for both scaling parameters, “a” and “b”. Based on the R² values both MLR model and the Power model fitted equally well for estimation of above ground biomass of orchard guava trees. There was a good agreement between the observed and the predicted biomass using both the equations. Published information shows that most equations relate tree biomass to diameter or diameter coupled with height. A review of equations developed for 65 species by Zianis and Mencuccini(2004) showed that in most cases tree diameter is the most commonly used single metric for tree allometry. These equations mostly deal with forest species and addressed selviculture issues and specifically the timber part. Very few addressed the mono-cropped tropical fruit trees like guava from the perspective of CS. However, their application to orchard/grafted fruit trees is problematic for two reasons: first, the complicated branching of grafted plants starting from just above the ground/graft union and second, the DBH a very common parameter used in published allometric equations of either mango or related species is not possible to measure in orchard/grafted trees because of the branching beginning just above ground level/ above the graft union. Hence the equation developed specifically for the grafted plants in this study will be of immense use in working out the biomass of orchards and to work out the CS of guava.

One of the purposes of this work is to see whether a common allometric equation can be used for all grafted fruit trees just like common multiple species equation in forestry using DBH as the allometric parameter. Therefore a comparison was made between the AGB predicted with these two models with that of mango and sapota equations developed by Ganeshamurthy et al.(2016) and is presented in Fig. 4. This comparison showed that both mango and sapota specific equations grossly under estimated the tree biomass by about 40 to 50 per cent. The deviations are accounted mainly to the tree canopy management. In guava pruning is done regularly to get seasonal crops such as summer, monsoon and winter crops. The same is not true in mango and sapota. Further the compactness of guava canopy, wood density and crop regulation are not similar to that of either mango or sapota. Therefore mango or sapota equation is not suitable for estimating the guava tree biomass. Since the purpose here was to see if mango/sapota equation can predict guava tree biomass the results clearly showed that mango or sapota equation cannot predicts guava tree biomass and hence for estimating guava tree biomass only guava tree specific allometric equation developed in this study is suitable.

Grafted guava trees of commercial orchards require independent allometric equation for estimation of tree biomass. The equations were hence developed using parameters other than DBH. The root to shoot ration also differed from those reported for forest trees. The BEF by and large attained stability beyond 16 years. The equations developed using PBG x NPB fitted the data well, and was statistically significant. There was a good agreement between the observed and the predicted biomass using both MLR and power model equations. Further our purpose was to see if mango equation can predict guava tree biomass. The results of this study confirmed that mango/sapota equation does not predicts guava tree biomass and hence mango/sapota tree allometric equation cannot be used for estimating guava tree biomass.

Funding source: Indian Council of Agricultural Research, India

Financial Support: The financial support received from ICAR- Emeritus Scientist Programme is gratefully acknowledged.

Conflicts of Interest: The authors declare that there is no conflict of interest in this work

Ethical Standards: This article does not contain any studies involving animals performed by any of the authors

Cairns, M.A,, Brown, S., Helmer, E.H & Baumgardner, G.A.(1997) Root biomass allocation in the world’s upland forests. Oecologia. 111, 1–11.
Ganeshamurthy,A.N., Ravindra,V., Venugopalan, R., Malarvizhi, M & Bhat, R.M. (2016) Biomass Distribution and Development of Allometric Equations for Non-Destructive Estimation of Carbon Sequestration in Grafted Mango Trees. Journal of Agricultural Science 8, 201-211.
Ketterings, Q.M., Coe, R., van Noordwijk. M., Ambagau, Y & Palm, C.A. (2001) Reducing uncertainty in the use of allometric biomass equations for predicting above-ground tree biomass in mixed secondary forests. Forest Ecology and Management . 120, 199-209.
Mori, S.M.(2001) Equations for predicting diameter, height, crown width and leaf area of San Joaquin Valley street trees. Journal of Arboriculture. 27, 306–317.
Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L., Miwa, K., Ngara, T., Tanabe, K & Wagner, F, (eds). (2003) IPCC good practice guidance for land use, land use change and forestry. Institute for Global Environmental Strategies (IGES), Hayama, Kanagawa, Japan.
Rupa, T.R., Ganeshamurthy,A.N., Karusala Alivelu., Rajendran, S., Sateesha, A.& Aruna B. (2022) How much a guava orchard sequesters carbon inits life time underseasonally dry tropical savanna climateon an Alfisol?(Agrochemica Vol. 66, No. 4 (2022).
Sanesi, G., Lafortezza, R., Colangelo, G., Marziliano, P.A & Davies, C. (2013) Root system investigation in sclerophyllous vegetation: an over-view. Italian Journal of Agronomy.. 8e17, 121-126.
Seber, G.A.F., & Wild, C,J.(1989) Nonlinear Regression. John Wiley and Sons, New York.
Tobin, B & Nieuwenhuis M.(2007) Biomass expansion factors for Sitka spruce (Piceasitchensis (Bong). Carr.) in Ireland. European Journal of Forest Research. 126, 189-196.
Zianis, D & Mencuccini, M.(2004) On simplifying allometric analyses of forest biomass. Forest Ecology and Management . 187, 311–332.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

A Biomass Estimation Model for Nondestructive Estimation of Guava Tree Biomass

Status:

Version 1

Abstract

Figures

1.0 Introduction

2.0 Materials And Method

3.0 Results

4.0 Discussion

5.0. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1