TITLE: An email about estimating productivity
DATE: 2023-05-24
AUTHOR: John L. Godlee
====================================================================


I had a student ask me something along the lines of:

 How do I calculate woody productivity in a woodland plot? I have
seen others calculate the difference in standing biomass between
censuses as net primary productivity. Is this right?

In our lab group we use a lot of repeat inventory tree growth data,
from plots in forests, savannas, and woodlands, to understand rates
of biomass dynamics and land-atmosphere carbon exchange. I wrote
quite a long email in response, so I thought I would share it here:

 There are lots of subtly different approaches to quantifying
woody "productivity". The literature is not always totally precise
in their definitions of productivity, which can make it all a bit
confusing.

 Generally, Net Primary Productivity (NPP) is defined as the
atmospheric carbon uptake by vegetation after subtracting the CO2
emitted via plant respiration, i.e. GPP minus autotrophic
respiration, where GPP is Gross Primary Productivity, the rate of
atmospheric carbon uptake by plants over a given time period.

 Crucially, NPP doesn't account for carbon losses from the system
via the mortality and subsequent decomposition of woody stems, i.e.
heterotrophic respiration, or loss of carbon from woody stems due
to disturbances like fire.

 So, to measure NPP you can't simply calculate the difference in
living woody biomass between censuses. This calculation of living
woody biomass change would be more readily termed Net Biome
Productivity (NBP).

 Instead, to measure NPP you would have to sum all the woody
growth that has occurred within the census interval, including
woody stems which have died within the census interval. This is not
straightforward, as in our fire-prone system in Bicuar sometimes
trees sometimes die and burn away within the census interval.
Feasibly, trees may also germinate, grow to sufficient size to be
measured, and die within a single census interval. There are ways
to estimate this unmeasured productivity, e.g. Talbot et al.
(2014), and Kohyama et al. (2019).

 The next thing to think about then, is whether you actually want
to measure NPP (the rate of woody growth in living stems), or NBP
(the rate of ecosystem-level woody biomass change, including
mortality losses). I will leave you to consider this.

 I have written R code to estimate both NPP and calculate NBP,
using the methods in Kohyama et al. (2019). Let me know if you
would like me to send it over.

 [2014]: https://doi.org/10.1016/j.foreco.2014.02.021
 [2019]: https://doi.org/10.1016/j.foreco.2018.11.010
 [NPP and calculate NBP]: https://github.com/johngodlee/ProdVital