emchristensen/ants_plants_lda

Dynamics of rodent, ant, and plant communities at Portal, and how they relate

ants_plants_lda's Issues

ant traits

need a list of traits for all the ant species we have -- granivore, omnivore, etc

timeseries analysis

the main question is how the community changes over time:

• do a basic ED between time steps of the community

species presence in colony vs bait data

I'm noticing that different species are recorded in the colony vs bait data. As in species presence/absence doesn't line up for a given plot/year combo (or even for the site as a whole).

If we want to compare species composition between years, should we maybe combine the two data types to get a "master" p/a estimate? @beecycles

ordinations

use some ordination method (PCA/RDA/NMDS) to visualize the community changing through time -- is it moving in a direction or randomly jumping around?

number of topics: plants

Using parLDA from LDATS on annual plant data chooses 32 topics for winter annuals and 29 for summer annuals. This is more topics than there are years of observations.

Bagchi 2017 methods

Some very valid qualms were raised about doing a straight Bray-Curtis dissimilarity timeseries analysis a la Bagchi et al 2017.
Idea from Morgan/Ethan: do a pseudo-wavelet thing, calculate dissimilarity between all pairs of points one year apart and plot. Same for all pairs of points 2 years apart, and so on. If a shift is really major on a short and long time scale, it will come out as a large dissimilarity on multiple measures.