Hypothesis 4: Leaf traits affect drought response

[mcgregorian1]

*H4(?)- Larger trees suffer more because larger trees tend to be species with more drought-sensitive traits.**

Note: I'm not sure if we want to get into traits in this paper. I think traits deserve to be a separate, stand-alone paper, both in terms of scientific significance and complexity. However, we could always include TLP (and ring porosity?) here and then dive into traits more thoroughly in anther paper. P4a- TLP predicts drought response Preliminary result : True (#6).*

P4b- TLP is lower (larger negative) in smaller/ understory trees Next step:

• [x] test this.

P4c- Inclusion of TLP in model eliminates (or significantly reduces) effect of height (and elev). Preliminary results : False. Model including TLP + dbh much better than just TLP, and TLP + dbh + elev is slightly better than TLP+dbh (#6)). Next step:

• [ ] test whether strength of height term is reduced by including TLP.

Originally posted by @teixeirak's writing in issue #7

[mcgregorian1]

P4b

I'm not sure how to test this because all species have the same TLP regardless of canopy position.

[teixeirak]

Test it on a community level (with no random species effect): Is there a correlation between dbh and TLP? Or alternatively, if you bin stems into size classes and calculate average TLP for each size class, do smaller size classes have lower (more negative)?

[mcgregorian1]

Doing a quick analysis shows that subcanopy have a higher range of tlp values

This is seen across the different main drought years as well.

Running a model with resistance value as the response variable, and dbh/tlp as fixed effects, I get the following:

[teixeirak]

Great, this is in line with what I expected. Could you please do the same with ring porosity?

It appears that we'll be able to reject the hypothesis that large trees suffer more because large size classes are dominated by more drought sensitive species. Good! I never thought that made sense.

[mcgregorian1]

For ring porosity, trees with "ring" values are almost the same by canopy position. The biggest difference is in diffuse (the NAs were subset out for the linear model below. These are the Pinus individuals).

Running a linear model with dbh and rp gives me the following:

Just to be transparent, the above linear model was run with data for all years, but only run as a linear model (so each value of rp was repeated 4 times for the different years). I'm not entirely sure it makes a difference, but I ran that again including year as a fixed effect and got the following:

[teixeirak]

Thanks, looks good.

[mcgregorian1]

P4c

P4c- Inclusion of TLP in model eliminates (or significantly reduces) effect of height (and elev). Preliminary results : False. Model including TLP + dbh much better than just TLP, and TLP + dbh + elev is slightly better than TLP+dbh (#6)). Next step:

test whether strength of height term is reduced by including TLP.

To follow up on this, I did a model comparison with dbh, height, tlp, and elev (plus year and random). Interestingly, height and tlp do the best at prediction (R2 is 0.16). Below is the chart, plus the coefficients.

[teixeirak]

I just noticed that year is being treated as a continuous variable. Please make it discreet, as we're not testing for a directional shift through time.

[mcgregorian1]

Having changed from 4 to 3 years (by averaging 1964-1966 into just one), and making "year" be discreet, I do get different results in which model is best. If anything, it shows a better case for elevation and dbh. However, the R2 has dropped to 0.12.

[teixeirak]

I think we're done with this issue.