United States Department of Agriculture

Forest Service

Northern Research Station

General Technical Report NRS-P-105

Moving from Status to Trends: Forest Inventory and Analysis (FIA) Symposium 2012

AbstractThese proceedings report invited presentations and contributions to the 2012 biennial Forest Inventory and Analysis (FIA) Symposium, which was hosted by the Research and Development branch of the U.S. Forest Service. As the only comprehensive and continuous census of the forests in the United States, FIA provides strategic information needed to evaluate sustainability of current forest management practices across all ownerships. Papers and abstracts included in the publication have been sorted into topic areas that match the sessions presented during the meeting. Symposium papers cover high priority and timely issue-based topics including climate change, wildlife, fire, bioenergy, geo-spatial extensions, monitoring over time, integrating remote sensing and GIS applications, statistical and related quantitative solutions to emerging needs, and many others.

Papers published in these proceedings were submitted by authors in electronic media. Some editing was done to ensure a consistent format. Authors are responsible for content and accuracy of their individual papers and the quality of illustrative materials and equations. The views and opinions expressed in this report are those of the presenters and authors and do not necessarily reflect the policies and opinions of the U.S. Department of Agriculture. The use of trade or firm names in this publication is for reader information and does not imply endorsement by the U.S. Department of Agriculture of any product or service.

Visit our homepage at: http://www.nrs.fs.fed.us/

Published by: For additional copies, contact:

USDA FOREST SERVICE USDA Forest Service11 CAMPUS BLVD., SUITE 200 Publications DistributionNEWTOWN SQUARE, PA 19073-3294 359 Main Road Delaware, OH 43015-8640November 2012 Fax: 740-368-0152

Manuscript received for publication August 2012

Cover design by Suzy Stephens, U.S. Forest Service

Moving from Status to Trends: Forest Inventory and Analysis Symposium 2012 361GTR-NRS-P-105

STRATeGIeS FoR ASSeSSING INTeR- AND INTRA-SPecIFIc vARIATIoN IN TRee BIoMASS IN THe INTeRIoR WeST

David l.R. Affleck, john M. Goodburn, and john D. Shaw1

Abstract.—W�ldf�re hazard m�t�gat�on and b�oenergy harvest�ng have emerged as forest management pr�or�t�es throughout the Inter�or West (IW) of the USA. Reg�onal forest �nventory and forecast�ng appl�cat�ons are therefore �ncreas�ngly focused on tree b�omass, �nclud�ng b�omass �n trad�t�onally non-merchantable components. Yet accurate b�omass equat�ons for the latter components are typ�cally lack�ng, even for major commerc�al spec�es. Th�s �s because prev�ous assessment efforts have focused on d�st�nct components, have used �ncons�stent methodolog�es, or have rel�ed on data w�th l�m�ted spat�al or b�ophys�cal extents. Here we rev�ew and contrast the current state of knowledge of a w�dely d�str�buted and commerc�ally �mportant spec�es, ponderosa p�ne (Pinus ponderosa), and a locally �mportant spec�es of the southwest, two-needle p�nyon (Pinus edulis). Ow�ng to h�gh levels of �nter- and �ntra-spec�f�c var�at�on �n b�omass allometry and to the h�gh costs of collect�ng b�omass data, an �mportant element of new data collect�on efforts w�ll be the development of standard�zed and accurate subsampl�ng procedures. We descr�be these subsampl�ng strateg�es and d�scuss potent�al adaptat�ons for two-needle p�nyon �n the southwest.

1 Assoc�ate Professor of B�ometr�cs (DLRA) and Assoc�ate Professor of S�lv�culture (JMG), Un�vers�ty of Montana, 32 Campus Dr�ve, M�ssoula, MT 59810; Analys�s Team Leader (JDS), U.S. Forest Serv�ce, Rocky Mounta�n Research Stat�on. DLRA �s correspond�ng author: to contact call 406-243-4186 or ema�l at dav�d.affleck@umontana.edu.

INTRoDucTIoNAcross the Inter�or West (IW) of the Un�ted States, the management of federally adm�n�stered forest lands �s �ncreas�ngly be�ng shaped by pol�c�es related to w�ldf�re hazard m�t�gat�on, b�oenergy development, and ecolog�cal restorat�on. At the same t�me, on pr�vate lands there �s he�ghtened �nterest �n regenerat�on and the growth dynam�cs of young stands. As such, there are grow�ng needs for rel�able �nventory �nformat�on on submerchantable trees and on forest b�omass pools such as fol�age, branchwood, and tops. These emerg�ng needs stand �n contrast to the cons�derable b�ometr�c knowledge concern�ng mature tree bole volume and taper that has been developed. Even for commerc�al

spec�es, few b�omass models su�table for reg�onal or subreg�onal analyses are ava�lable, and many of these are �ncomplete or �ncons�stent w�th one another.

In th�s paper we rev�ew the current state of knowledge concern�ng tree b�omass allometr�es, focus�ng on two IW spec�es: the w�dely d�str�buted ponderosa p�ne (Pinus ponderosa) and the locally �mportant southwestern two-needle p�nyon (Pinus edulis). We also cons�der the advantages and l�m�tat�ons of model-ass�sted subsampl�ng strateg�es that can be used to augment tree b�omass datasets and ref�ne reg�onal allometr�c models.

cuRReNT STATe oF KNoWleDGe coNceRNING BIoMASS AlloMeTRIeSPonderosa PinePonderosa p�ne �s found from northern Mex�co to southern Br�t�sh Columb�a, and from Cal�forn�a to

Moving from Status to Trends: Forest Inventory and Analysis Symposium 2012 362GTR-NRS-P-105

Nebraska. It �s a commerc�ally �mportant spec�es over much of the northern reaches of �ts range. It generally occup�es relat�vely dry s�tes but �s found over a broad cl�mat�c envelope: from the northern S�erras to the adjacent Great Bas�n �t grows on s�tes w�th annual prec�p�tat�on rang�ng from approx�mately 1800 mm/yr to 200 mm/yr (Callaway et al. 1994).

Most of our understand�ng of the b�omass allometr�es of ponderosa p�ne comes from stud�es conducted �n the northern port�on of the spec�es’ range. Yet even the relat�vely �ntens�ve f�eld work by, for example, Brown (1978) and Stand�sh et al. (1985), had l�m�ted geograph�c support relat�ve to the spec�es’ overall d�str�but�on. Moreover, past stud�es often employed d�fferent (and vary�ng) tree b�omass measurement protocols and part�t�oned totals �nto d�st�nct b�omass components. Though not �rresolvable, the latter �ssues do make �t challeng�ng to use prev�ously collected data or prev�ously developed models to �nform reg�onal or subreg�onal analyses. From smaller-scale stud�es such as Callaway et al. (1994) and Bouff�er et al. (2003), �t �s ev�dent that th�s spec�es’ b�omass allocat�on patterns and wood dens�ty vary w�th�n the reg�on ow�ng to env�ronmental and genet�c d�fferences; d�fferences �n data collect�on protocols compl�cate the detect�on and descr�pt�on of such var�at�ons.

Two-Needle PinyonIn the southwestern Un�ted States, two-needle p�nyon �s a locally �mportant spec�es �n terms of forest cover and secondary forest products. It �s a woodland spec�es rang�ng mostly over the states of Utah, Colorado, New Mex�co, and Ar�zona. Two-needle p�nyon �s found on warmer and dr�er s�tes than ponderosa p�ne. It grows �n both s�ngle-stem, excurrent forms (on better s�tes) and w�th spread�ng crowns and mult�ple stems (typ�cal form at lower elevat�ons).

Compared to ponderosa p�ne, l�ttle has been publ�shed on the b�omass allometr�es of two-needle p�nyon or on �ts genet�c var�at�on or wood dens�ty character�st�cs. Few b�omass stud�es have been conducted and these have been h�ghly local�zed �n terms of geograph�c

extent and have been based on small sample s�zes (�.e., 10 to 15 trees) (Darl�ng 1967, Gr�er et al. 1992).

BIoMASS SuBSAMPlINGMotivations and challengesF�eld assessment of tree b�omass �s an expens�ve undertak�ng �n the mounta�nous and remote reg�ons of the western Un�ted States. Bes�des access costs there are often d�ff�cult�es assoc�ated w�th the destruct�ve nature of the sampl�ng (e.g., �f large d.b.h. trees must be felled or �f standard log lengths cannot be ma�nta�ned). Mater�als process�ng costs are also s�gn�f�cant w�th defol�at�on of branches be�ng part�cularly t�me consum�ng (e.g., defol�at�on of f�rst-order branches can take more than one person-day for some con�ferous spec�es). At the same t�me, there �s a mult�tude of s�te and stand factors that can �nfluence the accumulat�on and apport�onment of b�omass, �nvest�gat�on of wh�ch necess�tates the accumulat�on of relat�vely large samples of trees for any g�ven spec�es. To collect a large sample �n the face of h�gh per-tree costs requ�res cons�derat�on of subsampl�ng strateg�es and potent�al sources of aux�l�ary (non-b�omass) �nformat�on.

The d�sadvantage of subsampl�ng for tree b�omass �s that w�th�n-tree sampl�ng errors become confounded w�th among-tree var�at�on. Put d�fferently, res�dual among-tree var�at�on �s �nflated by subsampl�ng error, mak�ng �t more d�ff�cult to �dent�fy and detect systemat�c var�at�on �n b�omass attr�butable to tree, stand, or s�te factors. Thus, effect�ve subsampl�ng strateg�es must allow the s�ze of the sample (�.e., the number of �ndependently selected trees) to �ncrease faster than tree-level uncerta�nty adds to the overall var�ance. Spec�f�cally, �f the subsampl�ng strategy �s unb�ased then the sample s�ze under subsampl�ng should be greater than the sample s�ze under full measurement by a factor of at least 1 + (σe/σt)

2, where σe �s the standard dev�at�on assoc�ated w�th the subsampling procedure and σt �s the cond�t�onal standard dev�at�on �n b�omass across trees w�th s�m�lar character�st�cs.

Moving from Status to Trends: Forest Inventory and Analysis Symposium 2012 363GTR-NRS-P-105

Development and ApplicationFundamental pr�nc�ples of tree allometry and past emp�r�cal work can be brought to bear to dev�se eff�c�ent and pract�cal tree b�omass subsampl�ng strateg�es. As one example, for a g�ven d.b.h. and he�ght much of the tree-to-tree var�at�on �n bole b�omass can be attr�buted to d�fferences �n bole taper. Th�s suggests us�ng prev�ous taper model�ng efforts to m�n�m�ze costs assoc�ated w�th b�omass measurements spec�f�c to the bole. Mult�ple taper stud�es have been conducted for ponderosa p�ne �n the northwest; part�cularly noteworthy are mult�-po�nt equat�ons such as those developed by Flewell�ng and Ernst (1996). Though parameter�zed for spec�f�c subreg�ons of the northwest, the latter equat�ons allow for f�ne cal�brat�on of pred�cted prof�les to actual tree taper v�a the �nput of mult�ple upper stem he�ght-d�ameter pa�rs (that are eas�ly measured on the downed tree). Po�nt-level pred�ct�ons (�ns�de- and outs�de-bark) from the taper prof�le, together w�th �ntegrated volumes, can then be used as control var�ates, allow�ng for accurate est�mat�on of bole b�omass components from a random or systemat�c sample of d�scs.

As a second example, numerous f�eld stud�es (e.g., Monserud and Marshall1999) have shown strong relat�onsh�ps between branch basal area and branch mass for many con�ferous spec�es. Less strong are relat�onsh�ps between, on the one hand, branch basal area and, on the other, branch fol�ar mass or branch woody b�omass. Nonetheless, such relat�onsh�ps form an �mportant bas�s for b�omass subsampl�ng strateg�es because branch basal areas are more eas�ly measured than branch masses (green or dry; whole or by component). Subsampl�ng procedures can draw on these relat�onsh�ps v�a unequal probab�l�ty des�gns (e.g., l�st sampl�ng or random�zed branch sampl�ng, RBS) and at the est�mat�on stage through the use of general�zed rat�o or regress�on est�mators. Such strateg�es obv�ate the need to measure full crown

we�ghts �n the f�eld, a task that �s both t�me-consum�ng and fraught w�th d�ff�cult�es due to the prevalence of broken and bur�ed branches. As part of an ongo�ng ponderosa p�ne b�omass study �n the northwest (Affleck and Turnqu�st 2012), RBS of f�rst-order branches has been used to cap�tal�ze on the strong quadrat�c relat�onsh�p between branch d�ameter and branch mass wh�le carry�ng out the crown sampl�ng �n a s�ngle pass up the stem. The use of general�zed regress�on est�mators w�th RBS �s also be�ng stud�ed as �n�t�al results have �nd�cated that a number of con�fer spec�es do not exh�b�t area-preserv�ng branch�ng at the f�rst-order level, a character�st�c �mpl�c�tly assumed �n RBS.

In develop�ng b�omass sampl�ng strateg�es for two-needle p�nyon, there �s cons�derably less mensurat�onal knowledge to draw on. Wh�le volume equat�ons have been developed for th�s spec�es, we are aware of no work on taper. Flex�ble taper equat�ons for other spec�es m�ght be subst�tuted and used for control var�ate sampl�ng, though the procedure would need to be mod�f�ed for mult�-stemmed growth forms. Inasmuch as the mult�-stemmed growth form �s common, �t �s ant�c�pated that crown subsampl�ng strateg�es based on branch scal�ng relat�onsh�ps w�ll be more useful for th�s spec�es. In part�cular, RBS strateg�es could be appl�ed �n d�fferent ways to est�mate crown character�st�cs on s�ngle stems or across mult�ple stems.

AcKNoWleDGMeNTSFund�ng for th�s research was prov�ded by a jo�nt venture agreement between the U.S. Forest Serv�ce and the Un�vers�ty of Montana (11-JV-11242305-120), by the Jo�nt F�re Sc�ence Program through a grant to DLRA and JMG (10-1-02-13), and by the Inland Northwest Growth and Y�eld Cooperat�ve.

Moving from Status to Trends: Forest Inventory and Analysis Symposium 2012 364GTR-NRS-P-105

lITeRATuRe cITeDAffleck, D.L.R.; Turnqu�st, B.R. 2012. Assessing

the accuracy of crown biomass equations for the major commercial species of the interior northwest: study plan and preliminary results. In: McW�ll�ams, W.; Roesch, F.A., eds. Mon�tor�ng Across Borders: 2010 Jo�nt Meet�ng of the Forest Inventory and Analys�s (FIA) Sympos�um and the Southern Mensurat�on�sts. Gen. Tech. Rep. SRS-157. Ashev�lle, NC: U.S. Department of Agr�culture, Forest Serv�ce, Southern Research Stat�on: 247-254.

Bouff�er, L.A.; Gartner, B.L.; Domec, J.-C. 2003. Wood density and hydraulic properties of ponderosa pine from the Willamette Valley vs. the Cascade Mountains. Wood and F�ber Sc�ence. 35: 217-233.

Brown, J.K. 1978. Weight and density of crowns of Rocky Mountain conifers. Res. Pap. INT-197. Ogden, UT: U.S. Department of Agr�culture, Forest Serv�ce, Intermounta�n Forest and Range Exper�ment Stat�on. 56 p.

Callaway, R.M.; DeLuc�a, E.H.; Schles�nger, W.H. 1994. Biomass allocation of montane and desert ponderosa pine: an analogy for response to climate change. Ecology. 75: 1474-1481.

Darl�ng, M.L.S. 1967. Structure and productivity of a pinyon-juniper woodland in Northern Arizona. Durham, NC: Duke Un�vers�ty. Ph.D. d�ssertat�on.

Flewell�ng, J.W.; Ernst, R.L. 1996. Stem profile estimation - east-side species. Unpubl�shed report to the Inland Northwest Growth and Y�eld Cooperat�ve. On f�le at the U.S. Department of Agr�culture, Forest Serv�ce, Forest Management Serv�ce Center, Fort Coll�ns, CO.

Gr�er, H.L.; Ell�ott, K.C.; McCullough, D.G. 1992. Biomass distribution and productivity of Pinus edulis-Juniperus monosperma woodlands of north-central Arizona. Forest Ecology and Management. 50: 331-350.

Monserud, R.A.; Marshall, J.D. 1999. Allometric crown relations in three northern Idaho conifer species. Canad�an Journal of Forest Research. 29: 521-535.

Stand�sh, J.T.; Mann�ng, G.H.; Demaerschalk, J.P. 1985. Development of biomass equations for British Columbia tree species. Informat�on Report BC-X-264. V�ctor�a, BC: Canad�an Forestry Serv�ce, Pac�f�c Forest Research Centre.

The content of this paper reflects the views of the author(s), who are responsible for the facts and accuracy of the information presented herein.

Printed on recycled paper

Morin, Randall S.; Liknes, Greg C., comps. 2012. Moving from Status to Trends: Forest Inventory and Analysis (FIA) Symposium 2012; 2012 December 4-6; Baltimore, MD. Gen. Tech. Rep. NRS-P-105. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. [CD-ROM]. 478 p.

These proceedings report invited presentations and contributions to the 2012 biennial Forest Inventory and Analysis (FIA) Symposium, which was hosted by the Research and Development branch of the U.S. Forest Service. As the only comprehensive and continuous census of the forests in the United States, FIA provides strategic information needed to evaluate sustainability of current forest management practices across all ownerships. Papers and abstracts included in the publication have been sorted into topic areas that match the sessions presented during the meeting. Symposium papers cover high priority and timely issue-based topics including climate change, wildlife, fire, bioenergy, geo-spatial extensions, monitoring over time, integrating remote sensing and GIS applications, statistical and related quantitative solutions to emerging needs, and many others.

KEY WORDS: statistics, estimation, sampling, modeling, remote sensing, forest health, data integrity, environmental monitoring, cover estimation, international forest monitoring

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, age, disability, and where applicable, sex, marital status, familial status, parental status, religion, sexual orientation, genetic information, political beliefs, reprisal, or because all or part of an individual’s income is derived from any public assistance program. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternate means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at 202-720-2600 (voice and TDD). To file a complaint of discrimination, write to USDA, Director, Office of Civil Rights, 1400 Independence Avenue, S.W., Washington, DC 20250-9410, or call 800-795-3272 (voice) or 202-720-6382 (TDD). USDA is an equal opportunity provider and employer.

Northern Research Station

www.nrs.fs.fed.us