Multicollinearity of variables in the Eurasian tree biomass database and generic allometry

Due to climate change, the forest biomass acquires exceptional ecological significance at the planetary level. To evaluate it, easy-to-implement methods are needed, one of which is the method of allometric models. Since biomass is related to some dendrometric indices of the tree, multifactorial allometric models began to be developed, however, their verification for multicollinearity was usually not carried out, and the models were often unstable. The paper attempts to optimize the structure of a multifactorial model of biomass under the condition of multicollinearity of independent variables based on the most comprehensive author's database of tree biomass to date. For this purpose, 11,170 model trees with measured aboveground biomass were selected from the database for 12 forest-forming genera, of which only 1,514 trees had their root biomass measured. The multicollinearity of the tree's age, DBH, and height is analyzed. Based on the calculation of the variance inflation coefficient (VIF), it was found that when these three variables being included in the model, give not correct result under the condition of multi-collinearity for 12 genera. When analyzing VIF for two prognostic variables in different combinations, the lowest VIF values (< 5) were found for age and DBH, but only for 6 genera, the most represented by the number of measurements. It would seem that a model that includes a combination of these two variables with a minimum VIF is optimal in its structure. However, the low multicollinearity of variables does not ensure the optimality of the model structure. Regression analysis of the relationship of biomass with these two variables showed that age in some cases is not statistically significant, and its contribution to explaining the biomass variability was less than 7%. This means that the optimal model structure can be established as a result of some compromise between low multicollinearity and high contribution of variables to the explanation of biomass variability. After all, generic allometric models have been calculated to estimate the aboveground, underground, and total biomass by DBH for 12 genera, which, when used locally, can produce biased results. Thus, the issue of optimizing the structure of the biomass model remains open, and in each case the researcher has to find a “golden mean” between selecting independent variables with minimal multicollinearity, on the one hand, and between including the largest number of statistically significant independent variables in the model, on the other. 

1. Armand D.L. Functional and correlation relations in physical geography. Proceedings of the All-Union Geographical Society, 1949, no. 1, pp. 81-94. (In Russ.)

2. Bedre R. Variance Inflation Factor (VIF) // Renesh Bedre’s Blog. URL: https://www.reneshbedre.com/blog/variance-inflation-factor.html (accessed Fabruary 19, 2025).

3. Chatterjee S., Simonoff J.S. Handbook of regression analysis. New York: Wiley, 2013. 252 p.

4. Chatterjee S., Hadi A.S., Price B. Regression analysis by example. New York: John Wiley and Sons, 2000. 359 p.

5. Draper N., Smith G. Applied regression analysis. Moscow: Statistics, 1973. 392 p. (In Russ.)

6. Dutcă I., Mather R., Ioraş F. Tree biomass allometry during the early growth of Norway spruce (Picea abies) varies between pure stands and mixtures with European beech (Fagus sylvatica). Canadian Journal of Forest Research, 2018, vol. 48, pp. 77– 84. DOI: 10.1139/cjfr-2017-0177.

7. Dutcă I., McRoberts R.E., Naesset E., Blujdea V. A practical measure for determining if diameter (D) and height (H) should be combined into D2H in allometric biomass models. Forestry: An International Journal of Forest Research, 2019, vol. 92, iss. 5, pp. 627-634. DOI: 10.1093/forestry/cpz041.

8. Falster D., Duursma R., Ishihara M., Barneche D.R., FitzJohn R.G., Varhammar A., Aiba M., Ando M., Anten N., Aspinwall M.J., Baltzer J.L., Baraloto C., Battaglia M., Battles J.J., Bond-Lamberty B., van Breugel M., Claveau Y., Dannoura M., Delagrange S., Domec J.-C., Fatemi F., Feng W., Gargaglione V., Goto Y., Hagihara A., Hall J.S., Hamilton S., Harja D., Hiura T., Holdaway R., Hutley L.S., Ichie T., Jokela E.J., Kantola A., Kelly J.W.G., Kenzo T., King D., Kloeppel B.D., Kohyama T., Komiyama A., Laclau J.-P., Lusk C.H., Maguire D.A., Le Maire G., Makela A., Markesteijn L., Marshall J., McCulloh K., Miyata I., Mokany K., Mori S., Myster R.W., Nagano M., Naidu S.L., Nouvellon Y., O'Grady A.P., O'Hara K.L., Ohtsuka T., Osada N., Osunkoya O.O., Peri P.L., Petritan A.M., Poorter L., Portsmuth A., Potvin C., Ransijn J., Reid D., Ribeiro S.C., Roberts S.D., Rodriguez R., Saldana-Acosta A., Santa-Regina I., Sasa K., Galia Selaya N., Sillett S.C., Sterck F., Takagi K., Tange T., Tanouchi H., Tissue D., Umehara T., Vadeboncoeur M.A., Valladares F., Vanninen P., Wang J.R., Wenk E., Williams R., de Aquino Ximenes F., Y.Atsushi, Yamada T., Yamakura T., Yanai R.D., York R.A. BAAD: a Biomass And Allometry Database for woody plants. Ecology, 2015, vol. 96, iss. 5, art. no. 1445.

9. Graham M.H. Confronting multicollinearity in ecological multiple regression. Ecology, 2003, vol. 84, pp. 2809–2815. DOI: 10.1890/02-3114.

10. Himmelblau D. Applied nonlinear programming. Moscow: Mir, 1975. 535 p. (In Russ.)

11. Ivakhnenko A.G. Self-learning recognition and automatic control systems. Kiev: Technika, 1969. 391 p. (In Russ.)

12. Ivakhnenko A.G. Method of group consideration of arguments in forecasting problems. Kiev: Institute of Cybernetics, 1977. 24 p. (In Russ.)

13. Jenkins J.C., Chojnacky D.C., Heath L.S., Birdsey R.A. Comprehensive database of diameter-based regressions for North American tree species / USDA Forest Service Northeastern Research Station. General Technical Report NE-319. 2004. 45 p.

14. Jucker T., Fischer F. J., Chave J., Coomes D.A., Caspersen J., Ali A., Loubota Panzou G.J., Feldpausch T.R., Falster D., Usoltsev V.A., Adu-Bredu S., Alves L.F., Aminpour M., Angoboy I.B., Anten N.P.R., Antin C., Askari Y., Muñoz R., Ayyappan N., Balvanera P., Banin L., Barbier N., Battles J.J., Beeckman H., Bocko Y.E., BondLamberty B., Bongers F., Bowers S., Brade T., van Breugel M., Chantrain A., Chaudhary R., Dai J., Dalponte M., Dimobe K., Domec J.C., Doucet J.L., Duursma R.A., Enríquez M., van Ewijk K.Y., Farfán-Rios W., Fayolle A., Forni E., Forrester D.I., Gilani H., Godlee J.L., Gourlet-Fleury S., Haeni M., Hall J.S., He J.-K., Hemp A., Hernández-Stefanoni J.L., Higgins S.I., Holdaway R.I., Hussain K., Hutley L.B., Ichie T., Iida Y., Jiang H.-S., Joshi P.R., Kaboli H., Kazempour Larsary M., Kenzo T., Kloeppel B.D., Kohyama T., Kunwar S., Kuyah S., Kvasnica J., Lin S., Lines E.R., Liu H., Lorimer C., Loumeto J.-J., Malhi Y., Marshall P.L., Mattsson E., Matula R., Meave J.A., Mensah S., Mi X., Momo S., Moncrieff G.R., Mora F., Nissanka S.P., O'Hara K.L., Pearce S., Pelissier R., Peri P.L., Ploton P., Poorter L., Pour M.J., Pourbabaei H., DupuyRada J.M., Ribeiro S.C., Ryan C., Sanaei A., Sanger J., Schlund M., Sellan G., Shenkin A., Sonké B., Sterck F.J., Svátek M., Takagi K., Trugman A.T., Ullah F., Vadeboncoeur M.A., Valipour A., Vanderwel M.C., Vovides A.G., Wang W., Wang L.-Q., Wirth C., Woods M., Xiang W., de Aquino Ximenes F., Xu Y., Yamada T., Zavala M.A. Tallo – a global tree allometry and crown architecture database. Global Change Biology, 2022, vol. 28, pp. 5254-5268. DOI: 10.1111/gcb.16302.

15. Liepa I.Ya. Dynamics of wood stocks: forecasting and ecology. Riga: Zinatne, 1980. 171 p. (In Russ.)

16. Lin J., Chen D., Yang S., Liao X. Precise aboveground biomass estimation of plantation forest trees using the novel allometric model and UAV-borne LiDAR. Frontiers in Forests and Global Change, 2023, vol. 6, art. no. 1166349. DOI: 10.3389/ffgc.2023.1166349.

17. Luo Y., Wang X., Ouyang Z., Lu F., Feng L., Tao J. A review of biomass equations for China's tree species. Earth System Science Data, 2020, vol. 12, iss. 1, pp. 21-40. DOI: 10.5194/essd-12-21-2020.

18. Marcoulides K.M., Raykov T. Evaluation of variance inflation factors in regression models using latent variable modeling methods. Educational and Psychological Measurement, 2019, vol. 79, iss. 5, pp. 874-882. DOI: 10.1177/0013164418817803.

19. McLone R.R. Mathematical modeling – the art of applying mathematics. Mathematical modeling. Moscow, 1979, pp. 9-20. (In Russ.)

20. Muukkonen P., Mäkipää R. Biomass equations for European trees: Addendum. Silva Fennica, 2006, vol. 40, iss. 4, pp. 763–773.

21. Nalimov V.V. Theory of experiment. Moscow: Nauka, 1971. 208 p. (In Russ.)

22. Paré D., Gertner G.Z., Bernier P.Y., Yanai R.D. Quantifying uncertainty in forest measurements and models: approaches and applications. Canadian Journal of Forest Research, 2016, vol. 46, p. 5. DOI: 10.1139/cjfr-2016-0029.

23. Picard N., Saint-André L., Henry M. Manual for building tree volume and biomass allometric equations: from field measurement to prediction. Food and Agricultural Organization of the United Nations, Rome, and Centre de Coopération Internationale en Recherche Agronomique pour le Développement, Montpellier, 2012. 215 p.

24. Poorter H., Jagodzinski A.M., Ruiz-Peinado R., Kuyah S., Luo Y., Oleksyn J., Usoltsev V.A., Buckley T.N., Reich P.B., Sack L. How does biomass allocation change with size and differ among species? An analysis for 1200 plant species from five continents. New Phytologist, 2015, vol. 208, iss. 3, pp. 736-749. DOI: 10.1111/nph.13571.

25. Repina E.G., Tsypin A.P., Zaichikova N.A., Shirnaeva S.Yu. Econometrics in the MS Excel tabular editor [Electronic resource]: practicum. Samara; Publishing House of the Samara State University of Economy, 2019. 1 CD-ROM. (In Russ.)

26. Satoo T. A synthesis of studies by the harvest method: primary production relations in the temperate deciduous forests of Japan. Ecological Studies: Аnalysis and synthesis. New York, 1970, vol. 1, pp. 55–72.

27. Sawadogo L., Savadogo P., Tiveau D., Dayamba S.D., Zida D., Nouvellet Y., Oden P.C., Guinko S. Allometric prediction of aboveground biomass of eleven woody tree species in the Sudanian savanna-woodland of West Africa. Journal of Forestry Research, 2010, vol. 21, pp. 475–481. DOI: 10.1007/s11676-010-0101-4.

28. Schmidhuber J. Deep learning in neural networks: An overview. Neural Networks, 2015, vol. 61, pp. 85–117. DOI: 10.1016/j.neunet.2014.09.003.

29. Snell O. Die Abhängigkeit des Hirngewichtes von dem Körpergewicht und den geistigen Fähigkeiten. Archiv für Psychiatrie und Nervenkrankheiten, 1892, h. 23, ss. 436-446.

30. Sprugel D.G. Correcting for bias in log-transformed allometric equations. Ecology, 1983, vol. 64, pp. 208–210.

31. Ter-Mikaelian M.T., Korzukhin M.D. Biomass equations for sixty-five North American tree species. Forest Ecology and Management, 1997, vol. 97, pp. 1–24. DOI: 10.1016/s0378-1127(97)00019-4.

32. Tretyakov N.V., Gorsky P.V., Samoilovich G.G. Forest Mensuration Handbook.

33. Tables for forest taxation. Moscow-Leningrad: Goslesbumizdat, 1952. 854 p. (In Russ.) Tyurin A.V., Naumenko I.M., Voropanov P.V. Forest auxiliary book. Moscow: Goslestekhizdat, 1945. 405 p. (In Russ.)

34. Usoltsev V.A. Growth and structure of forest biomass. Novosibirsk: Nauka, Siberian Publishing House, 1988. 253 p. (In Russ.)

35. Usoltsev V.A. Single-tree biomass data for remote sensing and ground measuring of Eurasian forests: digital version. 3rd ed., enl. Yekaterinburg; Botanical Garden of Ural Branch of RAS, 2023. 1 CD-ROM. (In Russ.)

36. Usoltsev V.A., Kanunnikova O.V., Platonov I.V. Investigation of errors in estimating the carbon pool of forests using allometric models. Modern problems of sustainable forest management, inventory and monitoring of forests: mat. of int. conf. St. Petersburg, 2006, pp. 363-370. (In Russ.)

37. Usoltsev V.A., Kolchin K.V., Noritsina Yu.V., Azarenok M.V., Bogoslovskaya O.A. Biases of generis species-specific allometric models in the local estimating the biomass of pine, cedar and fir trees. Eco-potential, 2017, vol. 18, no. 2, pp. 47-58. (In Russ.)

38. Usoltsev V.А., Kovyazin V.F., Tsepordey I.S., Chasovskikh V.P., Azarenok V.A. Foliage biomass of the forests of Eurasia: correction of empirical modeling methods. Izvestia Sankt-Peterburgskoj Lesotehniceskoj Akademii, 2020, iss. 232, pp. 50–78. DOI: 10.21266/2079-4304.2020.232.50-78. (In Russ.)

39. West G.B., Brown J.H., Enquist B.J. A general model for the origin of allometric scaling laws in biology. Science, 1997, vol. 276, pp. 122-126.

40. West G.B., Brown J.H., Enquist B.J. A general model for the structure and allometry of plant vascular system. Nature, 1999, vol. 400, pp. 664-667. DOI: 10.1038/23251.

41. Xu Q., Lei X., Zhang H. A novel method for approaching the compatibility of tree biomass estimation by multi-task neural networks. Forest Ecology and Management, 2022, vol. 508, art. no. 120011. DOI: 10.1016/j.foreco.2022.120011.

42. Zeng W. Developing tree biomass models for eight major tree species in China. Biomass Volume Estimation and Valorization for Energy. 2017, chapter 1, pp. 3-21. DOI: 10.5772/65664.

43. Zianis D., Mencuccini M. On simplifying allometric analyses of forest biomass. Forest Ecology and Management, 2004, vol. 187, pp. 311–332. DOI: 10.1016/j.foreco.2003.07.007.