If we can understand factors governing the transition from earlywood to latewood then this will help enormously with our models of wood quality, including MFA.  Latewood, as you know, generally has lower MFA than earlywood.  In a recent study with Matt Waghorn, a postgraduate student, we attempted to model environmental factors and needle water potential impacts on MFA and air-dry density in a designed experiment with four different drought treatments. Measurements were from Silviscan at 2 mm intervals along radii with samples taken at 30 cm above ground level from juvenile radiata pines over one growing season, and we were looking at two-weekly periods of growth within the radii.  While our models brought all the drought treatment effects together (model residuals were not related to treatments), it was clear that they did not explain the transition from earlywood to latewood, which was a very interesting result.  The MFA model included PAR, needle water potential, and relative humidity as independent variables, plus two interactions.

Thanks Euan.  Is that work published?  I don't suppose you can post the equation here?  It would be interesting to code it into our eCambium model and see how it compares with what we currently have.  The model predicts needle water potential and we have PAR and RH.

The following is part of the abstract of a paper we published on this topic. Perhaps it can be of interest, but with another wood and approach.

The microfibril angle of the S2 layer of the secondary wall was measured in fibres taken from eight-year-old trees of 11 Eucalyptus clones growing in four sites in Brazil. The overall mean microfibril angle measured was 8.88. While the angle seemed to decrease slightly from pith to bark in a non-linear fashion, this decrease was not statistically significant. The microfibril angle showed statistically significant differences between sites and between clones, although the broad sense heritability of the microfibril angel was low (h2s0.293). Various theories attempt to explain the microfibril orientation, but we conclude that environmental stresses play an important role in defining the angle in Eucalyptus wood.

Lima, J. T.; Breese, M. C.; Cahalan, C. M. 2004. Variation in microfibril angle in Eucalyptus clones. Holzforshung 16: 160-166.

Hi Geoffrey. We recently completed a study on variation of MOE over different sites. There were quite a number of factors related to MOE (environmental and others). We did measure MFA over a sub-sample of the trees and together with density it explained MOE in lumber fairly well. The MOE results therefore might give you some clues - see attachment. Results on MFA is still in press but i can send you a copy if you want. An interesting relationship was that between N:K ratio from leaf analysis and MOE. We suspect that it might be related to earlywood latewood switch.

Article The prediction of the flexural lumber properties from standi...

Hi Geoff

The effect of environment on MFA tends to be fairly weak and more related to growth rate than anything else. There are some studies that looked for this and found no effect on MFA. The factors controlling MFA at the cellular level are likely to be complex and are poorly understood. Keep in mind the variation in MFA between individual tracheids is much larger than variation in the average MFA between rings or trees. One of my colleagues Bernadette Nanayakkara is using microcores to quantify tracheid development across growth seasons. Unfortunately they are not measuring MFA but this approach might be a useful one. I think it will be important in such a study to have information on the history of individual trees - was there compression wood in the preceeding rings etc. Regarding earlywood latewood transitions I have an impression that there is a strong genetic influence both for MFA and density but not sure anyone really looked at this. This is based on the strong differences between species in growth ring appearance (ignoring growth rate which is more likely strongly environmentally effected. Plenty of scope for interesting studies on this topic. Presence of reaction wood is likely to be the dominant environmental effect and is often ignored in studies, paricularly mild compression wood which is often not obvious without microscopy.

Environmental factors can exert great influence on different wood characteristics, but the genetic factor should be the most important. If environmental factors modify the width of the growth rings, this can affect the average fibril angle, but the variation can not be affected.

Interesting responses but I was hoping we might look at more quantitative ways to describe the radial and subannual variation in MFA along the lines that Euan mentioned.  While genetic factors are important, genetic variation acts through physiological mechanisms.  I agree MFA variation is linked to EW / LW differences which are probably related to rates of cell enlargement and division, which in turn are related to water status and photosynthate availability.  Rupert Wimmer and I looked at daily variation in E. globulus in a 1992 paper.  Annual averages are going to be affected by both the proportion of EW/LW and the absolute values of MFA.  

MFA is wildly varying within and between fibres as Lloyd mentions. I  found that in radiata seedlings I looked at in the early 90's.  So can we synthesise these into some form of systematic logic of physiologically meaningful relationships that can be quantified?  I think at this stage leaving CW aside??

dont know if directly relevant but we found a rather counter intuitive increase in density with irrigation in E,nitens in Chile. Presumably wood production continued into the summer but stopped in the controls...if MFA changes with the season then I guess this trait would be affected by this environmental manipulation via a modification of phenology

Rod, a lot of growth rate and density relationships are phenologically influenced.  I tried to diagram this in a 2002 paper in Dendrochronlogia which I will try to attach here.  hence a lot of our modelling relationships are trying to separate growth-rate mediated effects from direct effects.  

The question still remains, on a daily basis, what processes are happening in the cambium that may be driving MFA values?  At this stage we can only address averages among a population of cells.  

BTW more than happy for folk to critique the logic of this flowchart.  I know it falls short at many points.

C.B. very interested in the paper you mentioned on MFA.

Geoff

In 2005 we did a study where we measured MFA of all three layers of the secondary wall in juvenile and mature wood. Only the S2 varies within tree, the others are constant. So whatever your determining factors are they only act at a particular stage of cell wall formation. I guess also that S2 formation probably takes more than 1 day. I beleive there is probably some degree of spontaneous self assembly in this process and this is weakly deflected by some unknown factor(s).

Geoff,

For the model of MFA in degrees all variables were transformed using scaled power transformations. The lambda coefficients were as follows:

Variable (units)             value

nwp       (KPa)                 4.45

rh           (%)                     0.48

MFA       (deg)                  2.71

par         (uMol/m2/s)    -0.51

The model below describes MFA development at two weekly intervals for the second year's growth of juvenile radiata pine in a polyhouse in Rotorua:

Variable   Coefficient      Standard error  P-value

Intercept   330197.100  53114.520         0.000

par                     -78.900        19.820         0.000

nwp         1038899.900 300646.660       0.001

rh             -184152.200  30301.870         0.000

par:rh                 44.800         11.340         0.000

nwp:rh     -584108.50 171698.170         0.001

Cheers,

Euan

Thanks for this Euan.  I'll discuss with Dave Drew about coding this into our model and see how it predicts against SilviScan data.