Results Lessons Learned in Developing New Stand Level

  • Slides: 20
Download presentation
Results & Lessons Learned in Developing New Stand Level Growth & Yield Models for

Results & Lessons Learned in Developing New Stand Level Growth & Yield Models for MN, WI & MI from FIA Data Alan Ek and David Wilson* Department of Forest Resources, University of Minnesota, aek@umn. edu Presented October 8, 2019, Duluth, MN Forest Resources Association Meeting _______________________ *David Wilson is now with MN DNR David. C. Wilson@state. mn. us 1

Abstract • G&Y model development has always been selective in data…thus results incorporate important

Abstract • G&Y model development has always been selective in data…thus results incorporate important assumptions on origin, management and disturbance (mortality) of plot data… that affect inference in model usage. • Recent modeling results using FIA data indicate these assumptions are VERY important as they affect local to forest wide model usage, e. g. , in harvest scheduling. • Here we provide guidance on choices for these assumptions and new models for their implementation. 2

Outline 1. 2. 3. 4. 5. History of G&Y modeling for the Lake States

Outline 1. 2. 3. 4. 5. History of G&Y modeling for the Lake States Advantages & challenges with FIA G&Y data Examining FIA data…growth, disturbance and decline New models, assumptions and applications Lessons learned 3

1. History of G&Y Modeling for the Lake States • Normal & empirical yield

1. History of G&Y Modeling for the Lake States • Normal & empirical yield tables & models – (Brown & Gevorkiantz 1934), Walters & Ek, 1993) • Stand table projection – (FACCS ~ Wilson, Domke & Ek 2010) • Stand level models based on growth data – (Red pine…Buckman 1962; Lundgren 1977) • Individual tree based models (USDA FS FVS) • Data selection choices – undisturbed or managed plots…typically the case (clean data) – any plots… including FIA…(Walters & Ek 1993) – recognizing plot disturbance…(Wilson, Morin, Frelich & Ek 2019) 4

2. Advantages & challenges with FIA G&Y data and documented disturbance • Wide representation…a

2. Advantages & challenges with FIA G&Y data and documented disturbance • Wide representation…a local to state to regional sample covering a wide range of growing conditions • Unbalanced samples by age, site, etc. • Realistic yet noisy data in terms of stand histories, composition, treatments and disturbances 5

Disturbance and mortality defined • Disturbance in FIA…windstorms, insects & disease, wildfire, animal damage,

Disturbance and mortality defined • Disturbance in FIA…windstorms, insects & disease, wildfire, animal damage, human disturbance • Regular mortality…due to competition between trees • Irregular mortality…. due to disturbance and can be large! 6

3. Examining FIA data - Aspen Five year growth trajectories in volume per acre

3. Examining FIA data - Aspen Five year growth trajectories in volume per acre (ft 3) by stand age in MN. N = 1, 166 re-measured plots sampled by FIA (2014) Blue line is data mean (approximate). 7

Things to look for… • Plots with very high growth rates • Plots that

Things to look for… • Plots with very high growth rates • Plots that are delayed or slow in growth • Ups and down of trajectories (ingrowth and mortality) and cumulative growth • Plots/stands don’t live forever! • Effect of site index and relative density • And…models that can describe the component processes and results 8

4. New models: GLFPS Great Lakes Forest Projection System (Based on USDA FS FIA

4. New models: GLFPS Great Lakes Forest Projection System (Based on USDA FS FIA data 1998 -2017, MN, WI, MI) • A new stand level model assembled in a spreadsheet • Encompasses 16 major forest types in the region • System submodels (for per acre values) – – – – Delay = expected time from bare ground to Dbh N = number of live trees 1” and larger B = basal area and d. B = basal area growth R = relative density (as % of Stand Density Index (SDI) maximum D = quadratic mean diameter (inches) H = dominant/codominant total height (feet) V = volumes cubic feet (gross & net), plus biomass, & carbon ZEOs = modification to basal area survival 9

See Wilson & Ek 2019…for details of (GLFPS) Model Stand Model Designation Form* 1

See Wilson & Ek 2019…for details of (GLFPS) Model Stand Model Designation Form* 1 Number of trees per acre Dbh >. 95 (in) N t = 2 Quadratic mean Dbh (in) D = 3 Relative density R = 4 Basal area growth per acre (ft 2) d. B = 5 Basal area yield per acre (ft 2) B = 6 Average total height of dominant / codominant trees (ft) H = 7 Volume gross Dbh > 4. 95 (in) from 1. 0 ft stump to 4 in or larger top dob (ft 3) V = b + b 0. 5+b 0 H/((1 -(1 -b 0 H))(N/N 0)b 1) b 0+b 1 D+b 2 N+b 3 S+b 4 A+b 5 DN b 0(B/A)(1 -(B/(Bmax)))(S/Smean) b 0 R 0. 5 S 0. 5 ((A)/(b 1+(A))) 0 Plus ZEOS adjustments 1 BH b 2 10

GLFPS projections and interpretation (Aspen forest type, basal area (B) growth; FIA data, MN,

GLFPS projections and interpretation (Aspen forest type, basal area (B) growth; FIA data, MN, WI, MI) • storiesid and slow • B growth projection (dark green) regular mortality only • B yield projection (light green) regular & irregular mortality • B growth projection (orange) with ZEOs modification (i. e. , with regular and irregular mortality) 11

GLFPS projections with ZEOs & interpretation (Example aspen forest type) with A = stand

GLFPS projections with ZEOs & interpretation (Example aspen forest type) with A = stand age, TPA = trees per acre and N > t trees per acre and showing asymptotes for t = 1, 5, and 48 tree shown, respectively, as thick black, red, and orange vertical lines. User can specify ZEO limit for N 12

GLFPS spreadsheet 13

GLFPS spreadsheet 13

GLFPS spreadsheet enlarged GLFPS also has options for thinning and expected level of irregular

GLFPS spreadsheet enlarged GLFPS also has options for thinning and expected level of irregular mortality 14

GLFPS draft growth & yield curves from model fits Aspen n=6294 Site indices 50,

GLFPS draft growth & yield curves from model fits Aspen n=6294 Site indices 50, 70 & 90, dark to light green ← 32 cords/ac ← 25 cords/ac 15

5. Lessons Learned 1. FIA data can be effective in developing new G&Y models.

5. Lessons Learned 1. FIA data can be effective in developing new G&Y models. 2. Recognizing growth patterns, disturbance and mortality are key to system/model development and interpretation of applications. * Applies to all forms of models. ** 3. Applications need to consider assumptions, E. g. , 1. 2. Single stand projection with only regular mortality Forest wide projection with disturbance--regular & irregular mortality 4. Study up on disturbance, incorporate it in estimation depending on model formulation and application. 5. More remeasurements on the way! ____________________ *Concept of growth & yield model compatibility is complicated! ** Including individual tree based models 16

GLFPS projections and interpretation (Aspen forest type, basal area (B) growth; FIA data, MN,

GLFPS projections and interpretation (Aspen forest type, basal area (B) growth; FIA data, MN, WI, MI) Where are you? • storiesid and slow • B growth projection (dark green) regular mortality only • B yield projection (light green) regular & irregular mortality • B growth projection (orange) with ZEOs modification (i. e. , with regular and irregular mortality) 17

Questions? Acknowledgements: Funding for this project was provided by the Minnesota Forest Interagency Information

Questions? Acknowledgements: Funding for this project was provided by the Minnesota Forest Interagency Information Cooperative (IIC), a state, county and University partnership for developing, promoting and sharing forest resources data, models, analysis and tools. 18

References Bechtold, W. A. and P. L. Patterson. [Editors] 2005. The enhanced forest inventory

References Bechtold, W. A. and P. L. Patterson. [Editors] 2005. The enhanced forest inventory and analysis program - national sampling design and estimation procedures. USDA For. Serv. GTR SRS-80. 85 p. Brown, R. M. and S. R. Gevorkiantz. 1934 Revised. Volume, yield, and stand tables for tree species in the Lake States. Univ. of Minnesota Agric. Exp. Sta. Tech. Bull. 39. 208 p. Buckman, R. E. 1962. Growth and yield of red pine in Minnesota. USDA Forest Serv. Tech. Bull. 1272. 50 p. Burkhart, H. E. , and M. Tom. 2012. Modeling forest trees and stands. Springer, NY. 457 p. Dixon, G. E. Essential FVS: 2017 Revised. A user’s guide to the forest vegetation simulator. Internal Rep. Forest Collins, CO. USDA For. Serv. Forest Management Service Center. 226 p. Lundgren, A. L. 1981. The effect of initial number of trees per acre and thinning densities on timber yields from red pine plantations in the Lake States. USDA For. Serv. RP NC-193. 25 p. Walters, D. K. , Ek, A. R. 1993. Whole stand yield and density equations for fourteen forest types in Minnesota. North. J. Appl. For. 10: 75 -85. Wilson, D. C. , G. M. Domke and A. R. Ek. 2014 Forest Age Class Change Simulator (FACCS): A spreadsheetbased model for estimation of forest change and biomass availability. Staff Paper Series Number 228, Department of Forest Resources, University of Minnesota. St. Paul, MN 19 p plus 2 appendices Wilson, D. C. and A. R. Ek. 2017. Imputing plant community classification forest inventory plots. Ecological Indicators. 80: 327 -338. Wilson, D. C. and A. R. Ek. 2017. Evaluating old age forests in Minnesota Forestry Research Notes 303. St. Paul, MN: Department of Forest Resources, University of Minnesota. 5 p. Wilson, D. C. and A. R. Ek. 2019. Whole Stand Growth and Yield Models for Major Forest Types in the Upper Great Lakes Region. Staff Paper Series 254. St. Paul, MN: Department of Forest Resources, University of Minnesota. 32 p plus appendices. https: //www. forestry. umn. edu/our-department/publications 19

References-continued Wilson, D. C. , R. S. Morin, L. E. Frelich and A. R.

References-continued Wilson, D. C. , R. S. Morin, L. E. Frelich and A. R. Ek. 2019. Monitoring disturbance intervals in forests: a case study of increasing forest disturbance in Minnesota. Annals of Forest Science. In Press. Wilson, D. C. and A. R. Ek. 2018. Evidence of rapid forest change in Minnesota Forestry Research Notes 305. St. Paul, MN: Department of Forest Resources, University of Minnesota. 8 p. Wilson, D. C. , J. M. Zobel and A. R. Ek. 2018. Forest Cover Type and Productivity as Related to Physiography. Minnesota Forestry Research Notes 306. St. Paul, MN: Department of Forest Resources, University of Minnesota. 3 p. plus appendix. Wilson, D. C. and A. R. Ek. 2019. Hardwood stand modeling using the Forest Vegetation Simulator (FVS). Minnesota Forestry Research Notes 307. St. Paul, MN: Department of Forest Resources, University of Minnesota. 5 p. plus appendix. Wilson, D. C. and A. R. Ek. 2019. Stand Volume, Biomass and Carbon Equations for the Upper Great Lakes Region. Minnesota Forestry Research Notes 308. St. Paul, MN: Department of Forest Resources, University of Minnesota. 9 p. Zobel, J. M. , Ek, A. R. , and O’Hara, T. J. 2014. Description and implementation of a single cohort and lifespan yield and mortality model forest stands in Minnesota Forestry Research Notes No. 298. St. Paul, MN: Department of Forest Resources, University of Minnesota. Zobel, J. M. , A. R. Ek and T. J. O’Hara. 2015. Quantifying the Opportunity Cost of Extended Rotation Forestry with Cohort Yield Metrics in Minnesota. Forest Science. DOI: http: //dx. doi. org. ezp 2. lib. umn. edu/10. 5849/forsci. 14 -215 20