: (GLMM) (pseudo replication) ( ) ( ) & Markov Chain Monte Carlo (MCMC)? /30

Size: px

Start display at page:

Download ": (GLMM) (pseudo replication) ( ) ( ) & Markov Chain Monte Carlo (MCMC)? /30"

みさえかつもと
5 years ago
Views:

PlotNet 6 ( ) 2006-01-19 TOEF(1998 2004), AM, growth6 DBH growth (mm) 1998 1999 2000 2001 2002 2003 2004 10 20 30 40

1 PlotNet 6 ( ) TOEF( ), AM, growth6 DBH growth (mm) DBH (cm) : kubo@ees.hokudai.ac.jp kubo/show/2006/plotnet/ /30

2 : (GLMM) (pseudo replication) ( ) ( ) & Markov Chain Monte Carlo (MCMC)? /30

3 ( ) TOEF( ), AM, growth6 DBH growth (mm) DBH (cm) ( ) ( ) /30

4 ( ) ( ) : : (TOEF) : /30

5 : 6 6? year growth6 TOEF( ), AM, growth6 DBH (cm) DBH growth (mm) (5 ) /30

6 ? (1) TOEF( ), AM, growth6 TOEF( ), AP, growth6 DBH growth (mm) DBH growth (mm) DBH (cm) DBH (cm) TOEF( ), OJ, growth6 DBH growth (mm) ( ) : β 1 log(dbh) + β 2 log(dbh) DBH (cm) /30

7 ? (2) year growth year growth year growth /30

8 ( )? /30

9 : Temperature VPD Temperature Precipitation PPFD : 6-9 ( ) VPD PPFD : ( ) ( ) ( ) /30

10 y1 A A "#%$# W y1 & B y2 B W y2 C C!! & #%$(') & : *,+(-. #%/ /30

11 ? (3) DBH growth (mm) TOEF( ), AM, growth DBH (cm) ; ( ) /30

12 : = (fixed effects) (random effects) TOEF( ), AM, growth6 DBH growth (mm) DBH (cm) fixed effects ; Poisson random effects ; 1 Gamma (mixed model) Poission Gamma /30

13 : ( ) L(β j, γ k, s {G i }) = i T r=0 [ R(r s) y Y ] P (G 6,y r, λ i,y ) dr, random effects R(r s) ( 1 s 2 Gamma ) fixed effects P (G 6,y λ i,y ) ( λ i,y Poisson ) P (G 6,y λ i,y ) = λg6,y i,y exp( λ i,y ), G 6,y! λ i,y exp( ) λ i,y = exp( X j β j x i,j,y + X k γ k w i,k,y ), /30

14 : R Random effects s fixed effects Akaike s Information Criteria (AIC) /30

15 : (Acer mono) DBH growth (mm) TOEF( ), AM, growth6 ( ): DBH (cm) temp.ps: best=16.4/width=4.6 ppfd: factor=0.02 vpd: factor=0.05 ( ): temp.mb: center=14.2/slope=1.0 rain: rbest=0.9/time= density posterior AM /30

16 : (Acer amoenum) DBH growth (mm) TOEF( ), AP, growth6 ( ): DBH (cm) temp.ps: best=16.4/width=5.8 ppfd: factor=0.05 vpd: factor=0.05 ( ): temp.mb: center=14.8/slope=2.0 rain: rbest=0.7/time= density posterior AP /30

17 : (Ostrya japonica) DBH growth (mm) TOEF( ), OJ, growth6 ( ): DBH (cm) temp.ps: best=16.0/width=5.8 ppfd: factor=0.04 vpd: factor=0.05 ( ): temp.mb: center=14.2/slope=2.0 rain: rbest=0.7/time= density posterior OJ /30

18 : TOEF( ), AM, growth6 DBH growth (mm) DBH growth (mm) DBH (cm) TOEF( ), AP, growth DBH (cm) ( ) /30

19 - - ( ) /30

20 {open, close} : 22 : 5m : ,,,, CN,, : 1 ( ) , 7, /30

21 :? :????? :? logistic p: ( 1 ) q: ( 1 ) p = exp(β CD + β LD L)/Z, L : 0, 1. q = exp(β CL + β LL L)/Z, Z = 1 + exp(β CD + β LD L) + exp(β CL + β LL L) /30

22 : : 1 ( )!?? /30

23 Nest : - - : β x = β Hyperspecies x (Hyperspecies) (Species) (Individual) + β Species x + β Individual x /30

24 : Markov Chain Monte Carlo (MCMC) Gibbs /30

25 : β CD = β Hyperspecies CD + β Species CD + β Individual CD /30

26 : 0.31 (close), 0.09 (open), /30

( )? p: ( 1 ) A: ( ; ) L: N: ( % ) p = 1 1 + exp( (β C

27 ( )? p: ( 1 ) A: ( ; ) L: N: ( % ) p = exp( (β C + (exp(β A ) + exp(β L )L)A + β N N)) /30

28 : /30

29 : average life (years) Caj Clj Ms Sb St EjSl Rt Pe Sp Cs Pn Vo Sg Pt Cc Ia Qs Na Pj Ir La nitrogen (%, spc mean) /30

30 : Random effects ( ) plot data? MCMC TOEF( ), AM, growth6 DBH growth (mm) DBH (cm) /30

12/1 ( ) GLM, R MCMC, WinBUGS 12/2 ( ) WinBUGS WinBUGS 12/2 ( ) : 12/3 ( ) :? ( :51 ) 2/ 71

2010-12-02 (2010 12 02 10 :51 ) 1/ 71 GCOE 2010-12-02 WinBUGS kubo@ees.hokudai.ac.jp http://goo.gl/bukrb 12/1 ( ) GLM, R MCMC, WinBUGS 12/2 ( ) WinBUGS WinBUGS 12/2 ( ) : 12/3 ( ) :? 2010-12-02 (2010 12