This work aimed to model the growth and yield of Eucalyptus stands located in northern Brazil, at the individual tree level, by using artificial neural networks (ANNs). Data from permanent plots were used for training the neural networks to predict tree height and diameter as well as mortality probability. Once trained, the networks were evaluated using an independent data set. The first group was composed of 33 plots (11 in each productive capacity class) and was used for artificial neural network training. In five measurements, this group totaled 8,735 cases (measurements of individual trees), as each plot had 53 trees on average throughout this evaluation. The second group was composed of 30 plots (10 in each productive capacity class) and was used for model validation. This group totaled 7,756 cases. Were tested different network architectures Multilayer Perceptron (MLP). Results revealed an underestimation bias for number of surviving trees. However, estimates of diameter, height, and volume per hectare were found to be accurate. This indicates that artificial neural networks are a viable alternative to the traditional growth and yield modeling approach in the forestry sector. 1. Introduction Individual tree models are constituted by a set of submodels that estimate diameter and height growth as well as mortality probability through tree- and stand-related variables and through competition data [1–3]. According to Munro [4], these models may be categorized according to how they consider the competition among trees, as represented by distance-dependent and distance-independent models. Since Newnham [5], many studies have been conducted worldwide in an attempt to improve growth models at the individual tree level and the relevant submodels. Methods to estimate parameters as well as different explanatory variables have been evaluated in an attempt to produce accurate, unbiased estimates of diameter and height growth and also tree mortality [6, 7]. In Brazil, few studies have been conducted that used these types of growth model for Eucalyptus [8, 9], a genus whose planted area intended to supply the processing industry amounts to 4.75 million hectares [10]. Some models have been developed and fitted for some species, including canjerana (Cabralea canjerana), black cinnamon (Nectandra megapotamica), cedar (Cedrela fissilis), and araucaria (Araucaria angustifolia), in natural forest conditions, yet without computing all the submodels that compose an individual tree model [11–14]. Typically, estimates of equation parameters for the models are derived from
References
[1]
J. L. Clutter, J. C. Fortson, L. V. Pienaar, G. H. Brister, and R. L. Bailey, Timber Management: A Quantitative Approach, John Wiley & Sons, New York, NY, USA, 1983.
[2]
L. S. Davis and K. N. Johnson, Forest Management, McGraw-Hill, New York, NY, USA, 1987.
[3]
J. K. Vanclay, Modeling Forest Growth and Yield: Aplications to Mixed Tropical Forest, CAB International, Wallingford, UK, 1994.
[4]
D. D. Munro, “Forest growth models—a prognosis,” in Growth Models for Tree and Stand Simulation, J. Fries, Ed., pp. 1–21, Royal College of Forestry, Stockholm, Sweden, 1974.
[5]
R. M. Newnham, The development of a stand model for Douglas-fir [Ph.D. thesis], University of British Columbia, Canada, 1964.
[6]
P. Soares and M. Tomé, “A distance dependent diameter growth model for first rotation eucalyptus plantation in Portugal,” in Empirical and Process—Bases Models for Forest Tree and Stand Growth Simulation, A. Amaro and M. Tomé, Eds., pp. 267–270, Salamandra, 1997.
[7]
F. Crescente-Campo, P. Soares, M. Tomé, and U. Diéguez-Aranda, “Modeling noncatastrophic individual tree mortality for Pinus radiate plantations in northwestern Spain,” Forest Ecology and Management, vol. 257, no. 6, pp. 1542–1550, 2010.
[8]
B. R. Mendes, N. Calegario, C. E. S. Volpato, and A. A. Melo, “Desenvolvimento de modelos de crescimento de árvores individuais fundamentado em equa？？es diferenciais,” Cerne, vol. 12, pp. 254–263, 2006.
[9]
F. B. Martins, Modelagem de crescimento em nível de árvore individual para plantios comerciais de eucaliptos [Ph.D. thesis], Universidade Federal de Vi？osa, Brazil, 2011.
[10]
ABRAF-Associa？？o Brasileira de Florestas Plantadas, “Anuário estatístico da ABRAF: ano base 2010. Brasília,” 2011, http://www.abraflor.org.br/estatisticas/ABRAF11/ABRAF11-BR.pdf.
[11]
M. A. Durlo, “Rela？？es morfométricas para Cabralea canjerana (Well.) Mart,” Ciencia Florestal, vol. 11, no. 1, pp. 141–149, 2001.
[12]
J. B. Della Flora, M. A. Durlo, and P. Soathelf, “Modelo de crescimento para árvores singulares—Nectandra megapotamica (Spreng.) Mez,” Ciencia Florestal, vol. 14, no. 1, pp. 165–177, 2004.
[13]
M. A. Durlo, F. J. Sutili, and L. Denardi, “Modelagem da copa de Cedrela fissilis Vellozo,” Ciencia Florestal, vol. 14, no. 2, pp. 79–89, 2004.
[14]
T. Chassot, Modelos de crescimento em diametro de árvores individuais de Araucaria angustifólia (Bertol.) Kuntze na floresta ombrófila mista [Ph.D. thesis], Universidade Federal de Santa Maria, Brazil, 2009.
[15]
J. C. C. Campos and H. G. Leite, Mensura？？o Florestal: Perguntas e Respostas, Universidade Federal de Vi？osa, Vi？osa, Brazil, 2009.
[16]
P. Miehle, M. Battaglia, P. J. Sands et al., “A comparison of four process-based models and a statistical regression model to predict growth of Eucalyptus globulus plantations,” Ecological Modelling, vol. 220, no. 5, pp. 734–746, 2009.
[17]
D. Merkl and H. Hasenauer, “Using neural networks to predict individual tree mortality,” in Proceedings of the Int’l Conference on Engineering Applications of Neural Networks, pp. 10–12, Gibraltar, UK, 1998.
[18]
M. Weingartner, D. Merkl, and H. Hasenauer, “Improving tree mortality predictions of Norway Spruce stands with neural networks,” in Proceedings of the Symposiun on Integration in Environmental Information Systems, Austria, 2000.
[19]
H. Hasenauer, D. Merkl, and M. Weingartner, “Estimating tree mortality of Norway spruce stands with neural networks,” Advances in Environmental Research, vol. 5, no. 4, pp. 405–414, 2001.
[20]
M. J. Diamantopoulou, “Artificial neural networks as an alternative tool in pine bark volume estimation,” Computers and Electronics in Agriculture, vol. 48, no. 3, pp. 235–244, 2005.
[21]
E. G？rgens, Estima？？o do volume de árvores utilizando redes neurais artificiais [Ph.D. thesis], Universidade Federal de Vi？osa, Brazil, 2006.
[22]
M. L. M. Silva, D. H. B. Binoti, J. M. Gleriani, and H. G. Leite, “Ajuste do modelo de Schumacher e Hall e aplica？？es de redes neurais artificiais para estimar volumes de árvores de eucalipto,” árvore, vol. 33, no. 6, pp. 1133–1139, 2009.
[23]
J. M. Paruelo and F. Tomasel, “Prediction of functional characteristics of ecosystems: a comparison of artificial neural networks and regression models,” Ecological Modelling, vol. 98, no. 2-3, pp. 173–186, 1997.
[24]
M. Gevrey, I. Dimopoulos, and S. Lek, “Review and comparison of methods to study the contribution of variables in artificial neural network models,” Ecological Modelling, vol. 160, no. 3, pp. 249–264, 2003.
[25]
H. G. Leite, M. L. M. da Silva, D. H. B. Binoti, L. Fardin, and F. H. Takizawa, “Estimation of inside-bark diameter and heartwood diameter for Tectona grandis Linn. trees using artificial neural networks,” European Journal of Forest Research, vol. 130, no. 2, pp. 263–269, 2011.
[26]
A. P. Braga, Carvalho, A. C. P. L. F, and T. B. Ludemir, “Redes neurais artificiais,” in Sistemas Inteligentes, S. O. Rezende, Ed., pp. 141–168, Manole, Barueri, Brazil, 2003.
[27]
A. K. Jain, J. Mao, and K. M. Mohiuddin, “Artificial neural networks: a tutorial,” Computer, vol. 29, no. 3, pp. 31–44, 1996.
[28]
S. Haykin, Redes Neurais: Princípios e Prática, Bookman, Porto Alegre, Brazil, 2001.
[29]
J. M. Barreto, Introdu？？o às Redes Neurais Artificiais, Universidade Federal de Santa Catarina, Florianópolis, Brazil, 2002.
[30]
B. T. Guan and G. Gertner, “Using a parallel distributed processing system to model individual tree mortality,” Forensic Science, vol. 37, pp. 871–885, 1991.
[31]
B. T. Guan and G. Gertner, “Modeling red pine tree survival with an artificial neural network,” Forensic Science, vol. 37, pp. 1429–1440, 1991.
[32]
D. L. Schomoudt, P. Li, and A. L. Abbot, “Machine vision using artificial neural networks with local 3d neighborhoods,” Computers and Electronics in Agriculture, vol. 97, pp. 101–119, 1997.
[33]
Q. B. Zhang, R. I. Hebda, Q. J. Zhang, and R. I. Alfaro, “Modeling tree-ring growth responses to climatic variables using artificial neural networks,” Forest Science, vol. 46, no. 2, pp. 229–239, 2000.
[34]
C. Liu, L. Zhang, C. J. Davis, D. S. Solomon, T. B. Brann, and L. E. Caldwell, “Comparison of neural networks and statistical methods in classification of ecological habitats using FIA data,” Forest Science, vol. 49, no. 4, pp. 619–631, 2003.
[35]
S. A. Corne, S. J. Carver, W. E. Kunin, J. J. Lennon, and A. W. S. van Hees, “Predicting forest attributes in Southeast Alaska using artificial neural networks,” Forest Science, vol. 50, no. 2, pp. 259–276, 2004.
[36]
R. ？z？elik, M. J. Diamantopoulou, J. R. Brooks, and H. V. Wiant Jr., “Estimating tree bole volume using artificial neural network models for four species in Turkey,” Journal of Environmental Management, vol. 91, no. 3, pp. 742–753, 2010.
[37]
R. A. Demolinari, Crescimento de povoamentos de eucalipto n？o-desbastados [Ph.D. thesis], Universidade Federal de Vi？osa, Vi？osa, Brazil, 2006.
[38]
M. G. Silva, Produtividade, idade e qualidade da madeira de Eucalyptus destinada à produ？？o de polpa celulósica branqueada [Ph.D. thesis], Universidade de S？o Paulo, S？o Paulo, Brazil, 2011.
[39]
T. D. Keister and G. R. Tidwell, “Competition ratio dynamics for improved mortality estimates in simulated growth of forests stands,” Forest Science, vol. 21, pp. 46–51, 1975.
[40]
G. R. Glover and J. N. Hool, “A basal area ratio predictor of loblolly pine plantation mortality,” Forest Science, vol. 25, pp. 275–282, 1979.
[41]
R. C. de Miranda, J. C. C. Campos, F. de Paula Neto, and L. M. de Oliveira, “Predi？？o da mortalidade regular para eucalipto,” árvore, vol. 13, pp. 152–173, 1989.
[42]
S. A. Machado, A. E. Tonon, A. F. Filho, and E. B. Oliveira, “Comportamento da mortalidade natural em bracatingais nativos em diferentes densidades iniciais e classes de sítio,” Ciencia Florestal, vol. 12, pp. 41–50, 2002.
[43]
R. Maestri, C. R. Sanquetta, and J. C. Arce, “Modelagem do crescimento de povoamentos de Eucalyptus grandisatravés de processos de difus？o,” Floresta, vol. 33, pp. 169–182, 2003.
[44]
L. M. B. Rossi, H. S. Koehler, C. R. Sanquetta, and J. E. Arce, “Modelagem da mortalidade em florestas naturais,” Floresta, vol. 37, pp. 275–291, 2007.
[45]
D. Zhao, B. Borders, M. Wang, and M. Kane, “Modeling mortality of second-rotation loblolly pine plantations in the Piedmont/Upper coastal plain and lower coastal plain of the southern United States,” Forest Ecology and Management, vol. 252, no. 1–3, pp. 132–143, 2007.
[46]
A. R. Stage, “Prognosis model for stand development,” USDA Forest Service Research Papers INT-137, USDA, Washington, DC, USA, 1973.
R. P. Lippmann, “An introduction to computing with neural nets,” IEEE ASSP Magazine, vol. 4, no. 2, pp. 4–22, 1987.
[49]
A. P. Braga, Carvalho, A. P. L. F, and T. B. Ludemir, Redes Neurais Artificiais: Teoria e Aplica？？es, Rio de Janeiro, Rio de Janeiro, Brazil, 2000.
[50]
R. G. D. Steel and J. H. Torrie, Principles and Procedures of Statistics, McGraw-Hill, New York, NY, USA, 1960.
[51]
P. A. Murphy and H. S. Sternitzke, Growth and Yield Estimation for Loblolly Pine in the West Gulf, Southern Forest Experiment Station, New Orleans, La, USA, 1979.
[52]
J. Siipilehto, “A comparison of two parameter prediction methods for stand structure in Finland,” Silva Fennica, vol. 34, no. 4, pp. 331–349, 2000.
[53]
A. Monty, P. Lejeune, and J. Rondeux, “Individual distance-independent girth increment model for Douglas-fir in southern Belgium,” Ecological Modelling, vol. 212, no. 3-4, pp. 472–479, 2008.
[54]
H. Pretzsch, P. Biber, and J. ？ursky, “The single tree-based stand simulator SILVA: construction, application and evaluation,” Forest Ecology and Management, vol. 162, no. 1, pp. 3–21, 2002.
[55]
F. A. Graybill, Theory and Application of Linear Model, Duxbury, North Scituate, Mass, USA, 1976.
[56]
D. A. Hamilton Jr., “A logistic model of mortality in thinned and unthinned mixed conifer stands of northern Idaho,” Forest Science, vol. 32, no. 4, pp. 989–1000, 1986.
[57]
I. E. Bella, “A new competition model for individual tree,” Forest Science, vol. 17, pp. 364–372, 1971.
[58]
R. F. Daniels, “Simple competition indices and their correlation with annual loblolly pine tree growth,” Forest Science, vol. 22, pp. 454–456, 1976.
[59]
H. Hasenauer and R. A. Monserud, “A crown ratio model for Austrian Forests,” Forest Ecology and Management, vol. 84, no. 1–3, pp. 49–60, 1996.
[60]
S. Vospernik, R. A. Monserud, and H. Sterba, “Do individual-tree growth models correctly represent height: diameter ratios of Norway spruce and Scots pine?” Forest Ecology and Management, vol. 260, no. 10, pp. 1735–1753, 2010.
[61]
H. Hasenauer, “Princípios para a modelagem de ecossistemas florestais,” Revista Ciência & Ambiente, vol. 20, pp. 53–69, 2000.
[62]
Q. V. Cao, “Predictions of individual-tree and whole-stand attributes for loblolly pine plantations,” Forest Ecology and Management, vol. 236, no. 2-3, pp. 342–347, 2006.
[63]
M. Tome and H. E. Burkhart, “Distance-dependent competition measures for predicting growth of individual trees,” Forest Science, vol. 35, no. 3, pp. 816–831, 1989.
[64]
H. Hasenauer, R. A. Monserud, and T. G. Gregoire, “Using simultaneous regression techniques with individual-tree growth models,” Forest Science, vol. 44, no. 1, pp. 87–95, 1998.
[65]
M. S. González, RíO, M. del, I. Ca？ellas, and G. Montero, “Distance independent tree diameter growth model for cork oak stands,” Forest Ecology and Management, vol. 225, no. 1–3, pp. 262–270, 2006.
[66]
K. Andreassen and S. M. Tomter, “Basal area growth models for individual trees of Norway spruce, Scots pine, birch and other broadleaves in Norway,” Forest Ecology and Management, vol. 180, no. 1–3, pp. 11–24, 2003.
[67]
C. H. Peng and X. Wen, Recent Applications of Artificial Neural Networks in Forest Resource Management: An Overview, American Association for Artificial Intelligence Technical, Menlo Park, Calif, USA, 1999.
[68]
E. Dogan, B. Sengorur, and R. Koklu, “Modeling biological oxygen demand of the Melen River in Turkey using an artificial neural network technique,” Journal of Environmental Management, vol. 90, no. 2, pp. 1229–1235, 2009.
[69]
P. W. West, “Simulation of diameter growth and mortality in regrowth eucalypt forest of southern Tasmania,” Forensic Science, vol. 27, pp. 603–616, 1981.
[70]
D. W. Hann and C. H. Wang, Mortality Equations for Individual Trees in the Mixed-Conifer Zone of Southwest Oregon, vol. 67 of Research Bulletin, Forest Research Laboratory, Oregon, Wash, USA, 1990.
[71]
R. A. Monserud and H. Sterba, “Modeling individual tree mortality for Austrian forest species,” Forest Ecology and Management, vol. 113, no. 2-3, pp. 109–123, 1999.
[72]
F. Crecente-Campo, P. Soares, M. Tomé, and U. Diéguez-Aranda, “Modelling annual individual-tree growth and mortality of Scots pine with data obtained at irregular measurement intervals and containing missing observations,” Forest Ecology and Management, vol. 260, no. 11, pp. 1965–1974, 2010.
[73]
M. Palahí and T. Pukkala, “Optimising the management of Scots pine (Pinus sylvestris L.) stands in Spain based on individual-tree models,” Annals of Forest Science, vol. 60, no. 2, pp. 105–114, 2003.
[74]
F. C. C. Uzoh and W. W. Oliver, “Individual tree height increment model for managed even-aged stands of ponderosa pine throughout the western United States using linear mixed effects models,” Forest Ecology and Management, vol. 221, no. 1–3, pp. 147–154, 2006.
