Is There a Need for Site Productivity Functions for Short-Rotation Woody Crop Plantings?

Mark Downing, Ph.D., Economist; Gerald A. Tuskan, Ph.D., Forest Geneticist
Biofuels Feedstock Development Program, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6352 U.S.A.

From the Proceedings, Second Biomass Conference of the Americas: Energy, Environment, Agriculture, and Industry; pages 207-215. Meeting held August 21-24, 1995, Portland, Oregon; published by National Renewable Energy Laboratory, Golden, Colorado.

For over a decade, researchers have used small-scale research plots to assist development and selection of high yielding, pest-resistant clones of fast-growing hardwoods such as hybrid poplar (Populus spp.). Substantial advances have been made in the techniques and criteria for screening species and selecting clones. Data from these research plots indicate that the ultimate performance of selected clones is dependent upon variable factors in the environment. Until now, researchers could only determine the suitability of a given site for such clones, not the actual yield potential of the site. Recently in the north central United States, several clones were planted on larger-than-research-scale plots on private land recontracted under the Conservation Reserve Program (CRP). The historical database could not provide a framework which would allow producers to predict the yield potential of a particular clone on a specific site. Through a systematic combination of clonal trials on experimental research-scale plots and operational plantings on 50 to 100 acre agricultural-scale field plots, it may be possible to develop yield functions or site quality equations which would predict biomass yields at rotation for selected clones. Such estimates will (1) reduce the probability of planting failure, (2) allow maximum expression of the genetic potential of selected superior clones, and thus (3) facilitate accurate economic planning for both the producer and conversion facility manager.

Through financial cooperation and integrated research a hybrid poplar (Populus spp.) plantation network was established in 1987 in the upper north-central United States. There have been several research objectives related to the establishment and monitoring of this network over the last 8 years. The most important may have been the screening and identification of hybrid poplar clones for operational use in Minnesota, North Dakota, South Dakota, and Wisconsin. Through the screening process, researchers have worked to develop accurate tools with which to estimate yield potential of selected clones on these particular sites. In addition to clonal screening, these plantations have been used to characterize pest resistance in the studied clones and to categorize environmental factors that influence growth.

It is known that two major environmental factors that determine yield potential are the edaphic and climatic characteristics of the selected sites. Although many crop management factors like weed control affect mortality and eventual productivity at rotation age, the soil and climatic factors are generally a matter of initial site selection rather than site modification. Thus, site selection becomes an important contributor to final plant productivity. This paper addresses three aspects of site selection and characterization related to short-rotation woody culture (SRWC) production. The first, consists of a review of historical data and literature dealing with soil and climate factors related to SRWC. The second aspect of this paper outlines potential application of site productivity functions from a hybrid poplar producer and hybrid poplar consumer perspective. And the third, presents a framework for the development of site productivity functions based on estimates of the number of acres, test sites, and clones to test on each site required to create robust prediction equations.

Because of the importance of the north-central U.S. to the development of biomass energy crops in the U.S. and because of the prominent role hybrid poplar play in this development, this region will be used in this paper as a model for the development of similar productivity functions in other regions and for other species of SRWC.

Several authors over the past 10 years have published criteria for site selection for hybrid poplars (Heilmann et al. 1991, Arbor Day Foundation 1993, Hansen et al. 1993). Site selection criteria for eastern cottonwood (Populus deltoides Bart.) were developed nearly 35 years ago (Broadfoot 1960). All of these criteria describe the need for moderate textured, well-drained soils with moderately high fertility, 3-8% organic matter and soil pH between 5.6 and 7.8. High water availability through either high water holding capacity or high water table are also required, though sites with prolonged flooding or standing water should be avoided. These criteria are designed mainly to eliminate sites that (1) do not provide minimal site requirements for soil texture, fertility, or organic matter content or (2) exceed threshold characteristics such as pH or slope. Although these criteria are important, they do not establish a relationship between site "quality" and ultimate hybrid poplar productivity. Such a relationship would (1) decrease production risk and supply uncertainty for producers and consumers, respectively, (2) ensure reliable productivity and yield from selected clones, (3) assist in determination of number of sites needed of a particular "quality" to attain an overall production level and (4) provide a reliable basis with which to determine alternative economic returns to different scales of production. The need for a more refined quantitative description of this relationship is the basis of this paper.

There is little literature and data available describing the relationship of hybrid poplar productivity to soils and climate within any SRWC plantations. Hansen et al. (1995), reporting on 130 acres of demonstration plantations in the north-central U.S., states that average productivity in 7- and 8-year-old hybrid poplar plantations equaled 3.3 dry tons per acre per year (TAY), ranged up to 4.6 TAY for the best clone at the best site, and surpassed 8.0 TAY on the best plot. Figure 1 demonstrates the effect of site and clonal selection on biomass production for 6-year-old plantations (Hansen et al. 1993). Although there was little difference in production within sites among clones, there was a substantial difference in production between sites, illustrating the importance of and opportunity for site selection to attain consistently high yields in future plantings (Hansen et al. 1993). Hansen et al. (1993) did report that water availability was a major factor in determining biomass production on a given site.

Figure 1. Effect of site and clonal selection on biomass production (6-year old plantations). From Hansen et al., 1993.

Hansen et al. (1993) also showed no definitive results to suggest that a particular clone would be recommended at one site or another. A regression equation: total tree dry weight=-1.67 * 0.23 (Diameter at Breast Height2), accounted for 98% of the total variation in biomass productivity and did not significantly vary among clones. Similar results were obtained by Tuskan and Rensema (1990) where total tree dry weight for four hybrid poplar clones grown in North Dakota could be reliably predicted by a single regression equation. Hansen et al. (1995) did note that the regression coefficients did vary significantly by site, supporting the hypothesis that there is a relationship between site quality and hybrid poplar productivity.

Berguson (1992) examined the north-central plantation network with the objectives to characterize soils within the network and to investigate the relationship between plantation productivity to soil and climate characteristics. Twenty-three sampling sites were chosen in Minnesota, Wisconsin, North Dakota, and South Dakota. Complete soil profiles were described according to Soil Conservation Service (SCS) standards. Soil bulk densities, organic matter content, and portion of gravel, sand, silt and clay were determined. Complete chemical analyses were performed. Particle size distribution and organic matter content were used to estimate the available water capacity by regression equation. Climate information was acquired from the National Oceanic and Atmospheric Administration (NOAA) data obtained nearest each planting site. Monthly average high and low temperatures, total annual rainfall and cooling degree days were included in this database.

The results of this assessment indicated that although soil characteristics varied widely across the network there was no relationship between any of the measured soil characteristics and poplar productivity. Berguson (1992) did note that growth, stand age, and soil characteristics were confounded among sampled sites. This confounding may have obscured the effect of soil water holding capacity, total available moisture or soil texture on tree biomass. After accounting for stand age, average cooling degree days was found to be positively correlated with mean tree dry weight. The R2 value indicated that 91 percent of the total variation in tree dry weight was explained by differences in averaging cooling days among sites. Berguson concluded that plantation productivity relationships with soil and climate characteristics would become stronger and more evident over time as plantation stands matured. Berguson recommended that the analysis could be improved by (1) obtaining daily temperature and growing degree days for all sites, (2) improving the precision of precipitation estimates, (3) using exponential weighing of soil properties, (4) collecting tree growth measurements at other sites, and (5) adding total runoff input by analyzing topography of land surrounding the sites.

Berguson (1994), through funding provided by the Electric Power Research Institute (EPRI), evaluated the land suitability of soils and climate for poplar production within the north central U.S. As an indicator of site quality, Berguson relied on the United States Department of Agriculture Soil Conservation Service (USDA-SCS) land capability classes to predict growth. The analyses were done with U.S. Forest Service Forest Inventory and Analyses (FIA) unit boundaries. Division of some of the units was necessary to provide a more detailed evaluation of soil suitability within the units. From the data presented in this paper, and as noted by Graham and Downing (1995), it is apparent that there is no relationship between SCS land capability class and poplar productivity. Heimlich (1994) confirms this as well. Furthermore, information on land rent, as provided by the Conservation Reserve Program (CRP), was not related to hybrid poplar productivity (see Table 8, Berguson 1994). Berguson does conclude that MN-AB MN-NP and MI-WUP units (Figure 2) have less potential production due to lower quality soils and low percentage of agricultural land. MN-P2 and MN-P3, and WI-SE units are considered highly productive in general but any drought-risk areas may pose difficulty in making site specific clonal selections (see Figure 2).

Figure 2. Forest Inventory and Analysis (FIA) units and subunits used in analysis (From Berguson 1994).

Reliable site productivity functions (that is, equations that would predict final mean annual dry weight at rotation based on edaphic and climatic characteristics of the selected site) would allow individual landowners to assess the potential biomass crops would offer as an alternative crop for their available land. This type of information would lead to more accurate projections of economic return from these lands. In addition, a consumer in a market for SRWC wood could more accurately plan annual feedstock acquisitions based on site productivity functions. For example, if a conversion facility requires 500 tons per day and the plant operates for 200 days per year, then 10,000 tons per year are needed. This would require harvesting 2000 acres per year if productivity is projected to be 5 TAY. If these plantations were managed on a 10 year rotation, the facility would need to rely on a 20,000 acre "fibershed" base. Alternatively, if sites could be identified that yield 7.5 TAY, then the total number of acres committed to the conversion facility could be reduced to ca. 13,000 acres. Accurate site productivity functions would permit conversion facility managers (1) to select the most productive sites within a given supply radius or (2) to project biomass yields from available land if site selection was not possible. Thus, both the landowner and conversion facility manager would benefit from reliable site productivity functions.

The ability to provide site productivity functions and thus site indices to predict final biomass yields on larger scales can only be accomplished through continued refinement and collection of all possible environmental determinants of growth. Although previous studies were unable to detect relationships between direct (e.g. soil attributes) or indirect (e.g. land rent values) site characteristics and measured productivity, Berguson and Hansen both noted the limited number of sites available for conducting such studies.

There are three areas for continued and new research. Accomplishing these goals over the next 5-10 years would (1) reduce the probability of planting failure, (2) allow maximum expression of the genetic potential of selected superior clones, and thus (3) facilitate accurate economic planning for both the producer and conversion facility manager.

First, measurement of the current hybrid poplar plantation network must continue through age 12 (currently age 9) in order to completely understand the actual yield potential at these sites, as well as verify the developed biomass regression equation. Information from these established sites will provide 1) measurements of mean annual increment which had not culminated on the study sites, 2) defined optimal rotation ages and yields of intensively cultured hybrid poplar, and 3) a clear relationship between foliar N and plot yield based on a significant decline in foliar N observed in the last year of study.

Second, in the spring of 1994, 3 clones of hybrid poplar were planted on 1000 acres of CRP land near Alexandria, MN. Intensive agricultural production budget information is currently being tracked for each year of production through harvest. This information is the only production-scale budget information being assembled on hybrid poplar. Production budget data such as these will provide economic information at the farm or producer scale. It will also provide sufficient information needed to establish pricing information for a potentially developing market for energy or increased hardwood market in the pulp and paper industry (Downing et al. 1994).

The most definitive data currently being assembled and tracked over time are the relationships of 28 soils on 14 CRP sites across the 1000 acres mentioned above and the specific number of plants and identified clones of hybrid poplar planted. This information will provide valuable yield information on sites where clones and soil types as well as SCS soil capability classes are mapped with scaled precision (Downing et al. 1994). This information will be augmented by data collected from a second 1000 acre planting established in May 1995. The growth and performance data on these sites will need to continue to be tracked through one full rotation. While the information will contribute to the data base needed to develop site productivity functions, it is limited to one climatic zone.

Finally, to permit robust statistical testing of the relationship between site characteristics and biomass productivity we propose the establishment of a regional site quality characterization study for the north-central U.S. This study would involve 4-8 test sites within each of three suitable subregions noted by Berguson (1995). At each site four genetically-diverse, highly-productive, selected hybrid poplar clones should be planted in 12 x 12 clonal blocks replicated four times per site. Establishment and maintenance procedures should be based on standard protocols outlined by Hansen et al. (1993). The entire study should be replicated in two successive years to account for variable climatic conditions. Finally, the study should be maintained for the full rotation length of 8-10 years. Annual incremental productivity should be based on the regression equations developed by Hansen et al. (1995), with the final productivity values based on destructive sampling and direct measures of dry weight. Weather stations should be established at each location and a complete assessment of soil properties should be conducted before and after completion of the study. This information should be combined with all additional data available on environmental factors and growth rates from other regional experimental, pre-operational, and commercial plantations to create a reliable site productivity function for the north central U.S.

Accurate site productivity functions would have utility to both potential producers and consumers of SRWC biomass. The establishment of systematically selected research plots for the development of site productivity functions will allow (1) a robust examination of the relationship between site characteristics and biomass productivity, and if justified, (2) the development of such functions. Data used in the correlation analyses would be collected at each site, as opposed to near available source, to assure accuracy. Furthermore, existing research plots and pre-operational plots in the region will be leveraged against the new plots to help validate the tested relationships.

Through funding provided by the U.S. Department of Energy's (DOE) Biofuels Feedstock Development Program (BFDP) and the Legislative Commission on Minnesota Resources (LCMR) from 1987 through 1991, the United States Forest Service (USFS) and University of Minnesota at Crookston established a hybrid poplar plantation network.

Arbor Day Foundation. 1993. Trees for fuelwood: A step towards energy diversity. The Arbor Day Institute. 80 pp.

Berguson, W. 1994. Evaluation of Land Suitability and Production Economics of Hybrid Poplar. Final Report, Natural Resources Research Institute, University of Minnesota, Duluth, Minnesota 55811. 33 pp.

Berguson, W. 1992. The Relationship of Hybrid Poplar Productivity to Soils and Climate Within the USFS/UM Plantation Network. Final Report. Natural Resources Research Institute, University of Minnesota, Duluth, Minnesota 55811. 24 pp.

Broadfoot, W.M. 1960. Field guide for evaluating cottonwood sites. U.S.D.A. Forest Service Occ. Paper 178. 6 pp.

Downing, M., R. Pierce and T. Kroll. 1994. 1994 Establishment Cost Data: Minnesota Wood Energy Scale-up Project. ORNL TM-12914. Oak Ridge National Laboratory, Oak Ridge, TN.

Graham, R. L. and Mark Downing. 1995. The Potential Supply and Cost of Biomass from Energy Crops in the Tennessee Valley Authority Region. ORNL Report #6858, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6352.

Hansen, E., D. Netzer and D. Tolsted. 1993. Guidelines for establishing poplar plantations in the north-central U.S. U.S.D.A.-Forest Service Res. Note NC-363. 6 pp.

Hansen, E., M. Ostry, D. Netzer and D. Tolsted. 1993. Short Rotation Woody Crops Trials for Energy Production in North Central United States. Annual Report. North Central Forestry Experiment Station, Forest Sciences Laboratory, 1831 Highway 169 East, Grand Rapids, Minnesota 55744. 16 pp.

Hansen, E., D. Netzer, M. Ostry, D. Tolsted and K. Ward. 1995. Short Rotation Woody Crop Trials for Energy Production in North Central United States. Final Report. North Central Forestry Experiment Station, Forest Sciences Laboratory, St. Paul, Minnesota. 41 pp.

Heilmann, P., R.F. Stettler, D.P. Hanley and W. Carkner. 1990. High yielding poplar plantations in the Pacific Northwest. Washington State University publication, PNW356. 32 pp.

Heimlich, Ralph E. 1989. "Productivity and Erodibility of U.S. Cropland." United States Department of Agriculture - Economic Research Service Agricultural Economics Report #604, January 1989.