Forests are experiencing an environment that changes much faster than in at least the past several hundred years. In addition, the abiotic factors determining forest dynamics vary depending on its location. Forest modeling thus faces the new challenge of supporting forest management in the context of environmental change. This review has focused on three types of models that are used in forest management: empirical, process-based and hybrid models. Recent approaches might lead to the applicability of empirical models under changing environmental conditions, such as (i) the dynamic state-space approach, or (ii) the development of productivity-environment relationships. 25 process-based models in use in Europe were analyzed in terms of its structure, inputs and outputs having in mind a forest management perspective. Two paths for hybrid modeling were distinguished: (i) coupling of EMs and PBMs by developing signal-transfer environment-productivity functions; (ii) hybrid models with causal structure including both empirical and mechanistic components. Several gaps of knowledge were identified for the three types of models reviewed. The strengths and weaknesses of the three model types differ and all are likely to remain in use. There is a trade-off between how little data the models need for calibration and simulation purposes, and the variety of input-output relationships that they can quantify. PBMs are the most versatile, with a wide range of environmental conditions and output variables they can account for. However, PBMs require more data making them less applicable whenever data for calibration are scarce. The EMs, on the other hand, are easier to run as they require much less prior information, but their simplicity makes them less reliable in the context of environmental changes. These different deficiencies of PBMs and EMs suggest that hybrid models may be a good compromise, but a more extensive testing of these models is required in practice.
The objective of this study was to develop a system of compatible equations to estimate eucalyptus ( Eucalyptus globulus Labill.) tree aboveground biomass and biomass of tree components for forest biomass prediction across regional boundaries. Data came from 441 trees sampled on several sites (99 and 14 plots in planted and coppice regenerated stands, respectively) representative of the eucalyptus expansion area in Portugal. The system of equations, simultaneously fitted using seemingly unrelated regression, was based on the allometric model for the biomass of stem wood, stem bark, leaves, and branches. Total aboveground biomass was expressed as the sum of the biomass of the respective tree components. The study allowed the following conclusions: (i) there is a significant increase in the predictive ability of the models that include height (stem components) or crown length (crown components) as an additional predictor to diameter at 1.30 m; (ii) there is a clear effect of the stage of development of the stand on tree allometry, with a decreasing pattern of the allometric constants; (iii) no effect of stand density, site index or climate on tree allometry was found; and (iv) for practical purposes, the same system of equations can be used for planted and coppice regenerated stands.
1 December 2006 in Canadian Journal of Forest Research
The work described in this paper aimed to establish a single set of parameter values for a process-based model (3-PG), applicable to Eucalyptus globulus Labill. in Portugal. Initial testing was done with parameter values from Australia using climate, biometric, and soils data from an irrigation and fertilization trial and a spacing trial. The model provided good estimates for stem mass and basal area, poor estimates for leaf mass, and reasonable estimates for volume. The fit between simulated and observed values was then improved by tuning parameter values to produce a final set. The calibrated model was tested, and performed well, against data from permanent sample plots (PSPs) at different locations across Portugal. Volume and basal area predictions made by 3-PG for PSPs were then compared with predictions made by the empirical model in use for E. globulus plantations in Portugal. Differences were negligible. Model outputs with the Australian parameter set and the optimum set for Portugal indicated that partitioning of carbohydrates (net primary productivity) was very different in E. globulus grown in Portugal and Australia. The study has confirmed the potential of this process-based model as a practical tool to support forest management decision-making.
1 May 2003 in Forestry: An International Journal of Forest Research
Although the speed of growth and adaptability of the north‐west American conifer Douglas‐fir has been recognized in Portugal, it represents only ∼0.1 per cent of the total forest cover. This small area is spread across the mountainous areas of the north and centre of the country. This study models Douglas‐fir productivity based on site factors and estimates the potential area for Douglas‐fir in Portugal. Soil, climate and topographic data were collected on 39 plots across the range of sites where Douglas‐fir grows in Portugal. The analysis of the data followed three steps: (1) selection of site factors related to the variation in the site index; (2) choice of candidate models; and (3) consideration of the best model to explain and predict the site index from site factor variables. The best multiple linear regression model explained 90 per cent of site index variation but included variables not readily available in the field. A model using digitized site data explained 54 per cent of the variation in the site index and mapped the areas with potential for Douglas‐fir growth in Portugal. It is estimated that a potential Douglas‐fir planting area of 250 000 ha exists where trees will exceed 17 m dominant height at age 30 years. This would correspond to 8 per cent of the existing Portuguese forest area. The best sites for Douglas‐fir growth are located along north coastal to central regions at altitudes between 500 m and 1000 m with moisture deficit (precipitation minus evapotranspiration) above 1000 mm. Areas with acceptable sites for Douglas‐fir growth are located in the north and centre of Portugal at 700–1000 m elevation and have a moisture deficit above 400 mm.
1 May 2003 in Forestry: An International Journal of Forest Research
The first plantations of the North‐West American conifer Douglas‐fir were established in Portugal in 1904. Investigations into growth and yield patterns were started in 1969. Since then others have carried out further studies. This study includes data from the previous studies and covers the whole range of site conditions where Douglas‐fir grows in Portugal. Dominant height data from 87 stem analysis trees from 12 sites were used to model dominant height growth of Douglas‐fir using the McDill–Amateis, Chapman–Richards and Lundqvist–Korf growth functions. The Chapman–Richards and Lundqvist–Korf growth functions were used in their integral and difference forms. For the evaluation of the candidate growth models' performance, three steps were adopted: (1) all the candidate growth equations were fitted with the data available from stem analysis from 87 trees, eliminating growth equations with non‐logical and non‐biological consistency and poor statistical properties; (2) the remaining growth equations were cross‐validated using two sub‐samples of the stem analysis data and selection of the growth equations with the best statistical results; (3) validation, using the whole data set from stem analysis for fitting growth models and the stand data for calculation of the prediction errors, was carried out. Out of the nine models evaluated five were rejected in the first step and two in the second step. The two best models had similar results in the third step and were compared with previous Douglas‐fir site index curves. The McDill–Amateis function performed best.
1 July 2013 in Canadian Journal of Forest Research
Simple, operational tools are required for forest managers to quantify the effects of soil fertility on tree growth and ecosystem sustainability leading to precise, sustainable forest management. The simplified process-based 3-PG model (Landsberg, J.J., and Waring, R.H., For. Ecol. Manage. 95: 209–228, 1997) provides a useful framework for operational prediction of forest growth. However, no simple objective method for relating the effects of available soil nutrients to the model fertility parameter fertility rating (FR) is yet available. The present study aimed to compare several general modeling approaches for the estimation of FR values from soil relative nutrient contents (RNCs) to maximum nonlimiting values, measured in the whole soil profile, at continuous inventory plots of Eucalyptus globulus Labill. in several locations under different parent materials in northwestern Spain. The modeling approaches tested provided good predictions of FR values from RNCs. In particular, using the minimum value of the most significant RNCs showed considerable potential for modeling FR values and plantation growth responses to them. This modeling approach showed promise to be further tested as a generally applicable strategy for estimating the effect of soil nutrients on forest plantations growth.