AND THE

3^rd ANNUAL FOREST INVENTORY & ANALYSIS SYMPOSIUM

GRAND TRAVERSE RESORT, TRAVERSE CITY, MICHIGAN

October 17, 18 & 19, 2001

* * * * * ABSTRACTS FOR ORAL & POSTER PAPERS * * * * *

(Document Modified, Monday, February 06, 2006)

* * * * * ORAL PAPERS * * * * *

Using auxiliary information to estimate stand tables, Mike Clutter, University of Georgia, Athens, GA and Ed Green, Rutgers University, New Brunswick , NJ

Abstract: We present a novel estimator for stand tables when few sample points / plots are available. The estimator is compared to the usual point sampling estimator and a precision-weighted composite estimator. Both the composite estimator and the pseudo-Bayes appear superior to the standard point sampling estimator in simulation studies. A variance estimator for the pseudo-Bayes estimator is presented that allows users to compare standard point sampling variance with a variance estimate that includes the auxiliary information.

A comparison of several techniques for imputing tree level data, southern FIA, David Gartner, Southern Research Station, Charleston, SC and Gregory Reams, Southern Research Station, Asheville, NC

Abstract: As Forest Inventory and Analysis (FIA) changes from periodic surveys to the multipanel annual survey, new methods of updating data will be needed. Several methods of updating plot level information such as using growth models, multiple imputation, and the ratio estimator have been discussed previously. However, imputing information at the plot level may be unwieldy for updating more detailed data such as the diameter distribution and species composition tables. Several methods for updating these more detailed data tables such as imputing whole tree lists from donor plots, imputing all the individual trees for each plot, and imputing plot level attributes such as total volume or basal area and then disaggregating the growth to the individual trees will be compared. The data from the last periodic survey of Georgia and Georgia’s first two panels will be used to compare the different imputation methods.

Evaluating Landsat-based stratification for timber and non-timber attributes on California’s North Coast, Antti Kaartinen, Jeremy Fried and Paul Dunham, Pacific Northwest Research Station , Portland, OR

ABSTRACT: Three Landsat TM-based GIS layers were evaluated as alternatives to conventional, manual, photo interpretation-based stratification of FIA field plots. Estimates for timberland area, timber volume, volume of down wood, and area of three different shrub cover classes were calculated for California’s North Coast Survey Unit, and were compared on the basis of standard errors of the estimates, conformance to FIA accuracy standards, and the gain in precision achieved by stratification relative to simple random sampling and to conventional photo interpretation. Some remote sensing approaches were found to be far less costly than the conventional method, with very little sacrifice in precision.

Effects of registration errors between remotely sensed and ground data in forest surveys, Paul Patterson and Michael S. Williams, Rocky Mountain Research Station, Fort Collins, CO

Abstract: The estimation of area by land cover type is a key component of most large-scale forest inventories. Historically, these estimates were derived from a large sample of points, taken from aerial photos, followed by a smaller sample of ground points, which were used to correct for errors in the classification of the aerial photo points. Naturally, there has been much interest in replacing aerial photography with satellite imagery because of cost reductions and the fact that the satellite imagery is a census of the land base, rather than a sample. One problem with using satellite imagery is the registration errors between a pixel and a plot or point on the ground. The estimators and modes of inference can differ substantially depending on whether a plot or a point on the ground are measured. Two terms, tessellated and point paradigm, are used to differentiate between the two approaches when the ground data consists of plots or points, respectively. The effects of registration errors are compared using the two paradigms for estimating the area of land by cover type. The effects of registration errors on the expected value and variance of the forest area estimator under both the tessellated and point paradigm were studied using a simulation study, where a percentage of the samples had a random one or two pixel registration error. The simulation study shows that registration errors increased the variance of the estimator of forest area from 4% to 434%. The estimator of forest area under both the tessellated and point paradigm exhibited no detectable bias. In the presence of registration errors, the estimated variance under the tessellated paradigm tended to over-estimate the true variance with the achieved coverage rate for a nominal 80% confidence interval ranging from about 81-86%. For sample sizes of fewer than 100 ground points, the estimated variance under the point paradigm tended to under-estimate both the 80% confidence interval and the true variance, regardless of whether there were registration errors or not. Further testing showed that sample sizes of between 100 and 250 ground points were needed before the estimator of the variance under the point paradigm converged to within 5% of the true variance of the estimator of forest area. Thus, we conclude that registration errors can drastically increase the variance of area estimators and confidence intervals based on the estimated variance will not achieve their nominal coverage rates. The forest area estimator under the tessellated paradigm was clearly more robust than the area estimator under the point paradigm.

Annual forest inventory estimates based on the moving average, Francis Roesch, James Steinman and Michael Thompson, Southern Research Station, Asheville, NC

Abstract: Three interpretations of the simple moving average estimator, as applied to the USDA Forest Service’s annual forest inventory design, are presented. A corresponding approach to composite estimation over arbitrarily defined land areas and time intervals is given for each interpretation, under the assumption that the investigator is armed with only the spatial/temporal matrix of moving average estimates. The advantages and practical limitations of each interpretation are discussed.

Developing Forest 5 and the next generation of regional forest growth simulation models, Christopher Schwalm and Alan Ek, University of Minnesota, St. Paul, MN

Abstract: Tree level models driven by boundary conditions expected to change under global warning are largely nonexistent. Forest v5.1, a numerical model with a daily timestep that predicts the growth of deciduous and coniferous species, was designed to remedy this deficit for the Great Lakes Region of North America. Model drivers mimic natural controls on plant growth and model initialization requires only typical field plot data. Forest v5.1 predicts the carbon, nutrient and water cycle as these influence tree growth with particular emphasis on light interception and assimilation. Design concepts and developments to date are described.

* * * * * POSTER PAPERS * * * * *

A comparison of stratification effectiveness between then national land cover dataset and photo-interpretation in western Oregon, Paul Dunham, Dale Weyermann, and David Azuma, Pacific Northwest Research Station, Portland, OR

Abstract: Stratifications developed from National Land Cover Data (NLCD) and from photo-interpretation (P-I) were tested for effectiveness in reducing sampling error associated with estimates of timberland area and volume from FIA plots in western Oregon. Strata were created from NLCD through the aggregation of cover classes and the creation of ‘edge’ strata by re-classifying pixels at class boundaries. Strata were created from aerial photography by interpretation of a sample grid for relevant attributes. NLCD-based stratifications are less costly than P-I and sacrifice little precision on inventory estimates. Neither P-I nor NLCD stratifications achieved the FIA target of 10%sampling error/billion ft³ of volume but both met the 3%sampling error/million acres of timberland standard.

Report of the 1999 and 2000 annual inventories of Maine's forests, Kenneth Laustsen, Maine Forest Service, Augusta, ME

Abstract: “In summary, the continuous approach to inventory will yield information that fluctuates from year to year and that may cause concern or lack of confidence in some end users.” (Andrew J. R. Gillespie, “Pros and Cons of Continuous Forest Inventory: Customer Perspectives.” Presented at the “Integrated Tools for Natural Resource Inventories in the 21^st Century” Conference, August 16 – 19, 1998. Boise, Idaho.) That comment captures the essence of this poster presentation.

To date, Maine has separately issued two annual inventory reports:

1. Report of the 1999 Annual Inventory of Maine’s Forests - October 24,2000; is based on an initial Panel 1 systematic sample of 646 plots and

2. Second Annual Inventory Report on Maine’s Forests - September 6, 2001; which is based on an intensified Panel 1 sample of 684 plots and a Panel 2 sample of 687 plots.

The fluctuations have been minimal: with 2 additional forest type groups being sampled; and several additional acreage and volume estimates becoming significantly different from the previous 1995 periodic inventory due to the power of the combined dataset of 1999’s Panel 1 and 2000’s Panel 2.

A first look at measurement error on FIA plots using blind checks in the PNW, Susanna Melson, Pacific Northwest Research Station, Portland, OR

Abstract: The PNW FIA program has initiated a blind plot protocol on 15 plots per state per year where a full, duplicate set of measurements are collected by a crew having limited knowledge of the data obtained during the first visit. Comparison of such data permits quantification of many aspects of plot measurement variation. Preliminary analysis from the first two seasons of blind plots is presented for variation in condition class fractions, understory cover, species richness, down wood counts, and selected tree variables such as dbh, height, crown ratio, crown class and damage.

Inventory and assessment of mossy forest of Mt. Amuyao for biodiversity resources conservation and management, Precila Gonzales-Salcedo and Feliciano Calora, Benguet State University, La Trinidad, Benguet, Philippines

Abstract: Floral inventory and assessment of the northern slope of Mt. Amuyao, Barlig, Mountain Province was conducted along altitudinal gradients from 1,600 to 2,600 masl. Circular plot method was used in this study wherein each plot was divided into four quadrants with reference to four cardinal directions. Three sample plots were constructed for each elevation starting from 1,600 up to 2,600 masl with a 200-m contour interval. With the use of Holdridge Life Zone Model (HLZM), three zones of different vegetation and floristic composition were identified. The HLZM vegetative classification was based on the principle of the relationship between the physiognomy and the vegetation cover of the environment. Four subzones were also categorized within the whole study site based from the association of dominant and co-dominant tree species. The classification of subzones was based on the computed importance values (I.V.) of all the tree species identified in all elevations. In this study, alpha diversity refers to diversity within each zone while gamma diversity pertains to the diversity of the whole northern slope of Mt. Amuyao. Diversity levels were determined using Shannon-Weiner Index (H'). Inter-and-intra zonal comparisons were made using Statistical Analysis System (SAS), SPSS and Similarity Index (SI). Survey questionnaires revealed that utilization of diverse bioresources in the study site is different for each zone. Community utilization of diverse bioresources were determined as well as the conservation status of each plant in the northern slope of Mt. Amuyao. A total of 280 vascular plant species belonging to 180 genera and 84 families were recorded from actual floristic surveys and vegetation analysis of the northern slope of Mt. Amuyao. Of the 280 species, 9 are large trees (3.21%); 25 medium sized trees (8.93%); 50 small trees (17.86%); 27 shrubs (9.64%); 53 herbs (18.93%); 18 epiphytes (6.43%); 68 ferns (24.28%); 12 grasses (4.29%); 16 vines (5.71%); 1 liana (0.36%); and 1 climbing bamboo (0.36%). Gamma diversity of this sector of the mountain is H' = 4.33. The northern slope of Mt. Amuyao is divided into three zones and four subzones, as follows: Zone 1 - Tropical Moist Forest (1,600 - 1,800 masl) has a total of 64 species distributed in 59 genera and 37 families. Of the 64 species, 20 are trees, 5 shrubs, 14 herbs, 3 epiphytes, 10 ferns, 9 grasses and 3 vines. This zone has an alpha diversity of H' = 2.65. Subzone A - Pinus forest (1,600 - 1,800 masl) is located at this zone. Zone 2 - Tropical Premontane Wet Forest (1,801 - 2,400 masl) has a total of 202 species distributed in 130 genera and 74 families. Of the 202 species, 56 are trees, 25 shrubs, 42 herbs, 16 epiphytes, 45 ferns, 4 grasses, 13 vines, and 1 parasitic plant. It has an alpha diversity of H' = 4.61. Subzone B - Lithocarpus-Decaspermum forest (1,801 - 2,200 masl) is located in this zone together with Subzone C - Drimys-Phyllocladus forest (2,201 - 2,400 masl). Zone 3 - Tropical Montane Rain Forest (2,401 - 2,701 masl) has a total of 74 species distributed in 61 genera and 44 families. Of the 74 species, 34 are trees, 6 shrubs, 10 herbs, 5 epiphytes, 15 ferns, 2 vines and 2 parasitic plants. This zone has an alpha diversity of H' = 3.12. Subzone D - Lithocarpus-Dacrycarpus forest is found within this zone at elevations of 2,401 to 2,701 masl. The conservation status of the280 vascular plant species are as follows: 84 species are endemic; 3 rare; 95 common and widespread. Most of these species has multiple uses. According to their uses and functional roles, the 280 species with multiple uses can be classified as followed: 21 are medicinal; 13 timbers; 105 ornamental; 122 landscaping; 29 domestic uses and the rest is of ecological importance. Among the three Zones, Zone 2 has the highest species diversity (H' = 4.61), followed by Zone 3 (H' = 3.12) and lastly, Zone 1 (H' = 2.65). A total of 115 Finnalig terms consisting of Finnalig names for vascular plants and household words are recorded here for the first time. Agricultural production is the main livelihood of the people in the area. Rice is the main crop planted in rice terraces at the foot of Mt. Amuyao located at Barangay Macalana. Indigenous people have traditional calendar that identifies each season based from natural phenomena like fruiting and flowering season, planting season, summer, appearance of migratory birds and others. The calendar can be used to determine which particular plant species are being harvested based from the season. During summer, forfor (Pinus kesiya Royle ex Gordon; Pinaceae) is being harvested for several purposes. This forest commodity is an indispensable product in their daily lives. This is being used as fuelwood, lumber for construction purposes and even source of cash during off season specially during summer. Minor forest products like sopor (Drimys piperata Hook f.; Winteraceae) are also harvested for several uses. Fikor (Schizostachyum diffusum (Blanco) Merr.; Gramineae) is widely harvested during chor-am, the harvesting season. Wild animals are also gathered from the forest during the hunting season which starts from October to February. Pine trees, moss, orchids, and other tree species with economic values are gathered for profit. Traditional beliefs practiced by the natives in extraction of forest products are also identified. Some of the identified beliefs could help in the conservation of the resources. However, it was revealed by the study that cultural and indigenous practices of the people in the area are eroding due to modernization. The community has no reforestation or rehabilitation project in the area due to the belief that trees can perpetually reproduce on its own and that there is no need for to replanting. Utilization of bioresources in Mt. Amuyao is still sustainable however, with the present economic condition in the area, it is necessary that socio-economic aspect of the natives be given immediate action by the government. The socio-economic status of the indigenous tribes in the study site affects the utilization of bioresources and therefore should be given urgent attention concomitant with the government's implementation of natural resources planning and management. It is predicted that if no action will be done, denudation and depletion of these diverse bioresources of Mt. Amuyao will occur in the next decade or so. (Abstract Received February 02, 2006)