GotelliLab / EcoSimR

Repository for EcoSimR, by Gotelli, N.J. , Hart E. M. and A.M. Ellison. 2014. EcoSimR 0.1.0
http://ecosimr.org
Other
27 stars 10 forks source link

Should I use co-occurrence analysis? #77

Open MittyAdai opened 2 years ago

MittyAdai commented 2 years ago

Hello! I am a student and also a beginner in ecology field and R, My goal is to check whether two species are spatially segregated. Not sure if i should use co-occurrence analysis in EcoSimR, Anyway, I did run the model but I had a hard time interpreting it. Hope someone could give me a hand to check if I'm doing it right! (p.s. I'm not a English native speaker, please don't mind if I made some grammatical mistakes)

My data contains two species in 64 sites, and I referred to the examples of following website: https://cran.microsoft.com/snapshot/2017-04-21/web/packages/EcoSimR/vignettes/CoOccurrenceVignette.html#caveats

Here are my code and results:

Create a Null model REAA <- cooc_null_model(speciesData=dataREAA, algo="sim10", suppressProg=TRUE, algoOpts=list(rowWeights=(1:2),colWeights=(1:64))) summary(REAA) plot(REAA, type="hist")

Summary Time Stamp: Fri Dec 3 20:55:29 2021 Reproducible: FALSE Number of Replications: 1000 Elapsed Time: 0.49 secs Metric: c_score Algorithm: sim10 Observed Index: 138 Mean Of Simulated Index: 144.8 Variance Of Simulated Index: 2330.1 Lower 95% (1-tail): 72 Upper 95% (1-tail): 230 Lower 95% (2-tail): 60 Upper 95% (2-tail): 250 Lower-tail P = 0.469 Upper-tail P = 0.547 Observed metric > 453 simulated metrics Observed metric < 531 simulated metrics Observed metric = 16 simulated metrics Standardized Effect Size (SES): -0.14091

If the observed c-score has no significant difference with simulated c-score, does it mean the two species is just randomized distributed (not segregated or aggregated)?

Also, the result is somehow a little bit weird if I used the sim9:

Time Stamp: Fri Dec 3 22:15:23 2021 Reproducible:
Number of Replications:
Elapsed Time: 1.3 secs Metric: c_score Algorithm: sim9 Observed Index: 138 Mean Of Simulated Index: 138 Variance Of Simulated Index: 0 Lower 95% (1-tail): 138 Upper 95% (1-tail): 138 Lower 95% (2-tail): 138 Upper 95% (2-tail): 138 Lower-tail P = 1 Upper-tail P = 1 Observed metric > 0 simulated metrics Observed metric < 0 simulated metrics Observed metric = 1000 simulated metrics Standardized Effect Size (SES): NaN

How should I explain this situation?

Hope someone could give me some guides, Thanks a lot! Mitty