Regional mixed-effects height–diameter models for loblolly pine (Pinus taeda L.) plantations

European Journal of Forest Research (Impact Factor: 2.1). 04/2007; 126(2):253-262. DOI: 10.1007/s10342-006-0141-7

ABSTRACT A height–diameter mixed-effects model was developed for loblolly pine (Pinus taeda L.) plantations in the southeastern US. Data were obtained from a region-wide thinning study established by the Loblolly
Pine Growth and Yield Research Cooperative at Virginia Tech. The height–diameter model was based on an allometric function,
which was linearized to include both fixed- and random-effects parameters. A test of regional-specific fixed-effects parameters
indicated that separate equations were needed to estimate total tree heights in the Piedmont and Coastal Plain physiographic
regions. The effect of sample size on the ability to estimate random-effects parameters in a new plot was analyzed. For both
regions, an increase in the number of sample trees decreased the bias when the equation was applied to independent data. This
investigation showed that the use of a calibrated response using one sample tree per plot makes the inclusion of additional
predictor variables (e.g., stand density) unnecessary. A numerical example demonstrates the methodology used to predict random
effects parameters, and thus, to estimate plot specific height–diameter relationships.

Download full-text


Available from: Harold E. Burkhart, May 17, 2014
82 Reads
  • Source
    • "In a mixed-effects model, both fixed and random parameters are simultaneously estimated, which allows variability for given phenomena among various factor levels to be modeled (Lindstrom and Bates 1990). This characteristic makes mixed-effects models more efficient when prediction for a new individual is required and prior information is available (Lappi and Bailey 1988, Gregoire et al. 1995, Garber and Maguire 2003, Leites and Robinson 2004, Trincado et al. 2007). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Quantification of forest biomass is important for practical forestry and for scientific purposes. It is fundamental to develop generic individual-tree biomass models suitable for large-scale forest biomass estimation. However, compatibility of forest biomass estimates at different scales may become a problem. We developed generic individual-tree biomass models using a mixed-effects modeling approach based on aboveground biomass data of Masson pine (Pinus massoniana Lamb.) from nine provinces in southern China. Mixed-effects modeling could provide an effective approach to solving the compatibility of forest biomass estimates at different scales. A simple allometric function requiring diameter at breast height was used as a base model to construct generic individual-tree mixed-effects biomass models. Two factors of tree origin (natural and planted forests) and geographic region (nine provinces or three subregions) were included as random effect factors in the models. The results showed that the mixed-effects model not only provided more accurate estimates, but also possessed good universality compared with the population average model. We, therefore, recommend the mixed-effects model 17 to estimate national and regional-scale biomass for Masson pine in southern China. The mixed-effects modeling approach is versatile and can also be applied to construct generic individual-tree models for other tree species and variables.
    Southern Forests: a Journal of Forest Science 02/2014; 76(1):47-56. DOI:10.2989/20702620.2013.870389 · 0.90 Impact Factor
  • Source
    • "To account for growing condition differences among stands for the same species, in addition to D, these models contain measures of site quality and/or stand density, among others. Studies have shown that a calibrated mixedeffects H–D model often produces better predictions than a region-wide H–D model containing stand-level regressors (Trincado et al. 2007, Temesgen et al. 2008, Huang et al. 2009). Mixed-effects models fit using plot-level data can also be calibrated at the plot level during operational inventories, providing more localised H–D relationships within a stand. "
    [Show abstract] [Hide abstract]
    ABSTRACT: To demonstrate the utility of mixed-effects height–diameter models when conducting forest inventories, mixed-effects height–diameter models are presented for several commercially and ecologically important conifers in the inland Northwest of the USA. After obtaining height–diameter measurements from a plot/stand of interest, these mixed-effects models can be calibrated to produce localised individual tree height estimates. An example of model calibration is provided. Based on model calibration results using independent data, the use of three to five trees for a particular species from a plot to calibrate the model will likely provide a reasonable compromise between predictive ability and field sampling times. If calibrated at the stand level, three trees could be used but larger sample sizes of 10 or 15 for an individual species may be reasonable. Similar mixed-effects height–diameter models could also be developed for species in the Southern Hemisphere.
    Southern Forests: a Journal of Forest Science 02/2014; 76(1):1-9. DOI:10.2989/20702620.2013.870396 · 0.90 Impact Factor
  • Source
    • "Details of the prediction of random effects parameters in the forestry context are provided e.g. by Calama and Montero (2004), for nonlinear mixed-effects models, and by e.g. Trincado et al. (2007), for linear mixed-effects models. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Artificial neural network methods appear to be a reliable alternative to traditional methods of tree height prediction in even-aged stands. However, this has not been demonstrated for uneven-aged forests. Two back-propagation artificial neural networks were constructed, and their performance in estimating the height of pure uneven-aged stands of common beech (Fagus sylvatica L.) in northwestern Spain was compared with that of the models most commonly used to estimate tree height (nonlinear calibrated local and generalized mixed-effects models and generalized fixed-effects models). All approaches produced accurate results, reducing the root mean squared error by more than 22% relative to basic nonlinear regression. Nonetheless, considering practical use of the models, the traditional approaches require measurement of several trees for calculation of stand-specific variables (generalized models) and for model calibration (mixed-effects models). Back-propagation artificial neural networks require less sampling effort because no height measurements are required for their implementation. However, this technique was not the best height predictor, because of the high degree of variability in site quality between stands. In this case, the local mixed-effects models yielded the best results and provided the best balance between the accuracy of the model and sampling effort.
    Forest Ecology and Management 11/2013; 307:63-73. DOI:10.1016/j.foreco.2013.07.014 · 2.66 Impact Factor
Show more