Take the 2-minute tour ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

If I want to predict litterfall, what data do I need to collect, and what statistical model might I use?

For example, I might use the following coding to record phenological stage every other day, or perhaps once per week:

|value| description
| 0   | all leaves fallen
| 0.5 | most leaves fallen
| 1   | no more green in canopy half of leaves have fallen
| 2   | most leaves yellow or red few leaves have fallen
| 3   | noticeable reddening or yellowing, green still present
| 4   | summer condition

Once I have these data, I can also collect weather data. Now, say I want to predict the day on which each of the transitions ($4\rightarrow 3$, $3\rightarrow 2$, etc) occur.

Many studies use a temperature metric of growing degree days $GDD = (T_{max}+T_{min})/2-T_{base}$, but I have also seen chiling days and photoperiod used to predict changes in phenological stage. I would like to develop a function (statistical model) $f$ that would allow me to predict a date of state change from environemental variables, such as $$D_{4\rightarrow 3} = f(GDD)$$

My questions:

  • what controls litterfall? Is it photoperiod, temperature, other?
  • do the controls vary by species?
  • is there a "standard" approach to modelling senescence?
share|improve this question
there was a related question on stats.SE (about modeling bud burst) stats.stackexchange.com/q/9797/1381 –  David Mar 6 '13 at 3:37
You might check out the R package chillR by Eike Luedeling, and his various papers documenting its use . –  Oreotrephes Aug 2 '13 at 14:35
add comment

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.