Increment modeling

Metadata for growth/scales:

FMC (num) circuli number identifying First marine circulus

FMC.spacing (num) FMC intercirculi spacing

smolt.incr (num) smolt increment (focus to last circuli before FMC)

marine.growth (num) marine growth = FMC to end

total.growth (num) entire scale from 1st circuli to the end

smolt.circ (num) number of smolt circuli

marine.circ (num) number of marine circuli

total.circ (num) number of total circuli

FS.incr (num) first summer at sea growth (FMC to FSM)

FW.incr (num) first winter at sea growth (first circuli after FSM to M1)

PS.incr (num) post-smolt growth (FMC to M1)

FSM.spacing (num) first summer maximum intercirculi spacing

M1.spacing (num) M1 intercirculi spacing

FS.circ (num) First summer, number of circuli

FW.circ (num) First winter, number of circuli

PS.circ (num) Post-smolt, number of circuli

SS.incr (num) Second summer at sea growth (first circuli after M1 to SSM (second summer max))

SW.incr (num) Second winter at sea growth (first circuli after SSM to M2)

Y2.incr (num) Year 2 growth (first circuli after M1 to M2)

SSM.spacing (num) Second summer maximum intercirculi spacing

M2.spacing (num) M2 intercirculi spacing

SS.circ (num) Second summer, number of circuli

SW.circ (num) Second winter, number of circuli

Y2.circ (num) Year-two number of circuli

Model creation and selection for growth increment data

#Combine "ME" and "Maine" levels
Carlin_full_growth$State[Carlin_full_growth$State == "ME"] = "Maine"

Check distribution of dependent variables (dv).

Create function to loop through all dv for set of models. Expvars include cohort year, sea age, state of recapture (location), season of recapture, length at recapture, and weight at recapture, plus relevant interactions (length and weight never included in same model due to covariance)

Run makemodels function to create all possible models for the four DV abd return the AIC model selection tables.

makemodels(dv)

Write top models for each of the DVs and store in list. Create function to make summary and anova table for each of the top models.

Run summarizetopmodels to produce summary tables.

summarizetopmodels(topmods)

Example candidate models and AIC model selection table for smolt increment.

smolt.aictab

## 
## Model selection based on AICc:
## 
##                      K   AICc Delta_AICc AICcWt Cum.Wt      LL
## yearxagestatexmolen 22 451.85       0.00   0.50   0.50 -203.52
## yearxagestatemolen  15 453.12       1.26   0.26   0.76 -211.37
## yearxagestatexmowt  22 455.04       3.19   0.10   0.86 -205.12
## yearagestatelen     12 455.64       3.78   0.07   0.94 -215.70
## yearagestatemolen   14 456.90       5.05   0.04   0.98 -214.28
## yearxagestatemowt   15 459.36       7.51   0.01   0.99 -214.49
## yearagestatewt      12 459.57       7.72   0.01   1.00 -217.67
## yearagestatemowt    14 462.96      11.11   0.00   1.00 -217.31
## yearagelen           5 471.28      19.42   0.00   1.00 -230.62
## yearagemolen         7 473.20      21.35   0.00   1.00 -229.56
## yearagestate        12 488.53      36.68   0.00   1.00 -232.17
## yearagestatemo      13 493.54      41.68   0.00   1.00 -233.65
## yearage              4 510.61      58.76   0.00   1.00 -251.29

formattable(smolt.modsum)

formattable(FS.modsum)

First Winter Increment

Top model: FW.incr ~ SeaAge + RecaptureYear + State + RecaptureSeason + capLen

formattable(FW.modsum)

Postsmolt Increment

Top model: PS.incr ~ SeaAge + RecaptureYear + State + capLen

formattable(PS.modsum)