I show here preliminary results on the forest dynamics base on two censuses of the four 1ha plots. Growth rate (mm/year) is on average 2.3051851, ranging [1.9292218, 2.7758197. Mortality rate (% dead per year) is on averaged 1.6659041, ranging [1.1487083, 1.8780032. which is within usual range of values observed in tropical forests [1-2 %].

Two plots were established in order to overlap with 40-40 m plots established by Mark Fogiel in 1994. We can therefore compare the statistics between the two censuses. However, only a subset of the Fogiel plot labels were found, meaning a certain number of tree either died or lost their labels.

I compared the growth rates for the Fogiel plot 16 that is within our 1ha plot number 1. We found data for 31 individuals (51 in the original plot, so 20 trees are missing or dead). The mean growth rate is lower compared to what we observe for our plots 1.2.

In addition, with Elsa, we also quickly remeasured (we may have missed some trees) on 40-40 m Fogiel plot in a forest patch on the Bouamir rocher (plot 14). The mean growth rate is 1.4 i.e. also two-fold lower than those observed in our 1ha plots. The estimated mortality rate for this plot is 3.7 % which is higher than the range of values observed in our 1ha-plots.

Below is a boxplot of all individual growth rate. I apply filters to trim potentially misleading values. Are excluded: growth rate > 20 mm/y and when the second census diameters are below the diameter of the first census minus twice the standard deviation of a linear model (following forestGeo method). This means that some negative growth are not excluded because they are considered to be possible given the inaccuracy of measurements (or the height at which measurments are taken).

My hypothesis is that the lower mean growth rate (between 1994 and 2019 (plot 16) or 2022 (plot 14)) is biased by missed trees that lost their labels. I assume trees that grow faster would tend to loose their labels faster.

Hence, we must relocalize trees still alive that lost their labels, otherwise we will underestimate growths. When a map of the relative position of each tree is available, this is possible. If not, this becomes very complicated. It depends of the stem density, but in most cases, relocating trees without the map that would be very time-consuming.

All tree growth rate between first and second census.