Stands and associated lists of tree characteristics may be established using either real or simulated data.
Real tree data
Real tree data may be entered interactively or imported from external files.
Simulated trees for stands aged 25 years or less
The Establishment/regeneration dialog tab for stands aged 25 or less is designed to use data collected in various monitoring surveys such as establishment and performance surveys. The current performance survey is completed by a stand age of 14 years; however, spruce may be too small at this age to indicate productivity. We are allowing the capacity to enter survey data as late as age 25. We have not analyzed crop-tree vs. all-tree relationships nor distribution shapes at age 25; there may be some simulation problems at these later ages, however the mean and SD should be simulated closely. Additional data origin type labels are provided for naturally suckering, seeding or planted trees. With the exception of performance survey crop tree data, all inputs and data handling under the data origin type labels are the same; they are primarily provided for the user's convenience (but see crop-tree vs. all-tree data below).
An important note on establishing a tree-list: Since there is tremendous variation in the early growth and survival of white spruce, we recommend that users initialize their spruce and mixedwood stands at a later age, (e.g. Alberta performance survey age of 7-14 years) when early dynamics due to stock handling, frost/chinook occurrence, and grass/shrub competition have sorted themselves out. MGM's early growth functions for white spruce are based primarily on the Alberta SDS juvenile PSP data. Spruce growth and mortality trajectories follow the average of this data, but only account for SI and tree competition; not the more important drivers noted above and their interactions with tree competition. The "planted" and "natural seed" origin types with initialization at stand age <7 years are therefore of limited use for spruce stands and mixedwoods.
Pure deciduous dynamics are simpler; the "natural suckering" origin type and initialization at stand age 2 (after most suckering has occurred) should result in reliable projections. However, in many provinces, an establishment survey is completed at 3-5 years stand age, and may be more useful for initializing the treelist.
Additional work is underway to provide initial estimates of white spruce tree size under various edatope-silviculture combinations. (These features are currently disabled).
Crop-tree vs. All-tree data (Note: this option is not applicable to the current Reforestation Standards of Alberta (RSA). 2015)
The "orange book" regeneration survey protocol for Alberta (ASRD 2007) requires that only one "crop-tree" be measured for height in each 10 m2 regeneration plot. The crop tree must pass minimum height and free-to-grow requirements. Frequently, though certainly not always, it is the tallest deciduous or acceptable conifer tree that is measured. Mean height of these crop trees is then reported in block summaries of typically ≥64 regeneration plots (ASRD 2007). Saskatchewan and Manitoba have the same regen plot size and similar crop tree selection. The justification for selecting the taller trees at this stage is that there is some evidence these are most likely to persist in the stand at harvest (Tiarks et al. 1999). However, many jurisdictions and companies put in enhanced surveys with some plots that sample all trees. In order to deal with these two types of data, MGM has two options for initializing a tree-list from performance survey data, selected by the Data types "PerfSurveyAlltrees" and "PerfSurveyCroptrees".
If the former is selected, the tree-list is simulated from the mean and standard deviation as given (av ht, SD ht). However, "PerfSurveyCroptrees" is selected, the mean and SD are corrected for measuring the crop trees only, a type of extreme value selection. An adjustment is made to estimate the all-tree average and SD before simulating the tree-list. This adjustment is based on an analysis of Mixedwood Management Association (MWMA) member companies' enhanced performance surveys (2003) from 49 spruce and spruce-dominated mixedwood cutblocks scattered across Alberta.
Unfortunately we have no all-tree information from establishment survey data to build similar crop-tree measurement corrections for pure deciduous stands. We hope to obtain such data in the future. Please contact us if you are interested or have such data. We are also seeking DC and pine mixedwood strata enhanced performance survey datasets.
What is important to you: if you have crop-tree data collected under a protocol similar to Alberta (10 m2 plots, roughly the tallest tree per species group measured), use the "PerfSurveyCropTree" origin type and your block (or block sub-strata*) average and SD of crop tree height average and SD. If you have larger regen plots, or other data sources, we recommend you determine the all-tree average and SD by your own methods first, then enter these under the data origin "PerfSurveyAlltrees".
*Large areas within cutblocks holding widely different size, productivity, ecosite, density or composition should be simulated as separate tree-lists in MGM. Resulting yield curves can later be combined on an area-weighted basis.
Estimating Average and Standard Deviation of DBH
Although not required by regeneration survey protocols, the average and standard deviation of DBH may be available from additional data collected in the survey. If it is not and DBH measures are left blank, they will be estimated from regressions on average block height from the MWMA 2005 enhanced survey dataset. Additionally, when your trees are close to 1.3m tall, a significant portion will not have DBH, creating a data truncation issue. MGM will override any DBH average and SD you add in this case, estimating a value that delivers the correct DBH distribution for trees above 1.3m. In some cases (especially when average height <1.3m), the average DBH will be negative. This makes sense if you think it through. An information box will pop up to inform you if MGM estimates DBH for you. If all trees are <1.3m tall, DBH measures will be set to zero since no DBH values need to be simulated.
Tree-list size distributions
A bivariate Weibull distribution is used to simulate tree-lists. This allows near normal distributions of height and dbh to be simulated (typical of aspen) as well as various skewed distributions, from log-normal to exponential shapes (found in white spruce with significant natural ingress). A "max ht" value is requested in the dialog since skewed and exponential distributions can result in some anomalously high values. Typically the max ht should be the average height + 2 standard deviations, although it may need to be slightly higher if SD is large relative to the average. A warning message will pop up when the average, SD and max are incompatible.
Multiple cohorts (different species or samples) of trees may be simulated in one Establish or Regen Event to create complex stand structure. Each cohort is simulated independently.
Simulated trees for stands older than 25 years
These tree-lists are based on stand characteristics that must be provided by the user. Older PSP data suggest cohorts of one species have a bivariate normal frequency distribution for diameter and height, so this is what is used here. Parameters (mean and standard deviation for DBH, mean and standard deviation for height, and correlation between height and DBH) for the each tree cohort must be provided by the analyst although a basic lookup table is provided to give preliminary parameter values. These lookup tables are based on simple averages of available plot data. Multiple cohorts of trees may be simulated to create complex stand structure.