Beginning with our earlier database and synthesis of individual-tree biomass equations for North America (Jenkins et al. 2003; Jenkins et al. 2004), we will update with current literature on new equations, add equations including tree height (which were excluded in previous work); and add shrub biomass equations. All equations for predicting tree biomass from dbh and height will be assembled into an easily-searchable spreadsheet-style database. Because the literature on tree biomass equations includes so many species and size gaps, as well as some overlap, direct comparison of actual equations would not help land managers to choose from among diverse approaches and equations. Instead, a meta-analysis of all equations will be conducted to develop new 'generalized biomass equations' by using predicted values from equations (within dbh and height ranges of original data). Generalized equations will be developed for species groups by using statistical graphic and clustering techniques. The Forest Inventory and Analysis (FIA) database will be used to establish initial species groups for maximum numbers of biomass equations. Actual tree distribution data will be examined for numbers of species within diameter classes. Results from above will be combined for final determination of number of equations that can be devised for species groups on a region-by-region basis for application by land managers. Quantitative analyses will be used where possible but final judgment will likely rely heavily upon professional judgment and simple graphical comparison. Concurrent with analysis above, a special test will be done to examine the need for height in final equations produced from the meta-analysis. Final decision on height inclusion will be supported by a literature summary on height use in biomass equations. Two sets of parameters for final equations will be estimated from meta-analysis regression: (1) from generated data (within dbh and height range of original developmental data) and (2) from FIA data. These parallel results will be examined in ensuing analyses to determine final parameters for numbers of equations needed to estimate biomass for all U.S. tree species in all regions. Variances for equation-predicted values will be estimated by a bootstrap resampling method and used to calculate confidence intervals. For example, equation parameters will be repeatedly refit by leaving out a group of data each time to produce many estimates of predicted values for given predictor variables which in turn can be used to estimate a variance of predicted values. Since biomass estimates are typically given in dry weight units for consistency, dry weight units will be used. However, green weight conversions for practical application will also be devised. If enough data are available, seasonal green weight equations may also be developed. Steps above will be repeated (where possible) for estimating shrub biomass. However, there will be many fewer shrub equations in the literature. Equations estimating shrub biomass from shrub cover measurements will be sought. Lastly, results will be examined for further study of biomass estimation by using allometric scaling theory. Study will be submitted to a professional journal for peer review and publication.