dealing with fire severity

[wcornwell]

FESM atm

[wcornwell]

one way to do this. this is only the after checklists and just the distribution of fire severities where the bird was observed.

[coreytcallaghan]

Ah, so you are thinking a 'separate' mini-analysis, but yet still opens up the door for some discussion, etc...?

[wcornwell]

yup

[coreytcallaghan]

I'm indifferent. Seems like that would play to the 'get it out' approach well. :)

[wcornwell]

i think some of the "increasing species" are just playing around the edges of the fire and in low severity areas.

[coreytcallaghan]

Yup, was thinking that using this as an aid in interpreting the 'main' results would be good then in that instance.

[Josh-Lee1]

Very cool! That all sounds good to me. Do we need to make one with before checklists to observe change?

[wcornwell]

Yeah I know what you mean--did the birds move? or were they just in areas that burned severely?--but I'm a bit confused how to do that analysis. Any clever ideas welcome.

[coreytcallaghan]

Where is the average severity in area around checklist data? I want to add it to this dataset: data_df_with_date<-read.csv("processed_data/ebird_data_with_fire_dates.csv")

[wcornwell]

data_df<-read_csv("processed_data/ebird_data_with_fire_dates.csv")

final_sl_dat <- data_df %>%
  #  bind_rows(lists_without) %>%
  dplyr::filter(!is.na(day_of_fire)) %>%
  mutate(DURATION.MINUTES=as.numeric(DURATION.MINUTES),
         EFFORT.DISTANCE.KM=as.numeric(EFFORT.DISTANCE.KM)) %>%
  mutate(MONTH=month(OBSERVATION.DATE)) %>%
  mutate(before.after=ifelse(OBSERVATION.DATE>day_of_fire, "After", "Before")) %>%
  filter(before.after=="After")

fesm1000<-raster("Data/Raw/fesm_geotiff/fesm1000.tif")
b<-SpatialPointsDataFrame(cbind(final_sl_dat$LONGITUDE,final_sl_dat$LATITUDE), final_sl_dat, match.ID = FALSE,
                          proj4string = CRS("+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs"))
bb<-spTransform(b,crs(fesm1000))
final_sl_dat$sev<-extract(fesm1000,bb)

[coreytcallaghan]

Don't we just want to model this somehow:

Note that geom_jitter was used.

[wcornwell]

yes somehow.

It's not a nice neat linear model problem. fire severity for the before observations (ie severity of the future fire) is quite confusing.

[wcornwell]

could be a before/after * severity interaction, maybe?

[wcornwell]

possible figure:

any thoughts?

[coreytcallaghan]

I think this is solid. It will definitely help with writing the discussion a bit. And is probably the focal point for a whole paragraph. We can make an interactive version and host it on github and link to it, too. So people can see what species are which....

[Josh-Lee1]

Very cool! Is the trendline for all species? And will we have something like this for all traits?

On Thu., 10 Sep. 2020, 10:24 Corey Callaghan, notifications@github.com wrote:

I think this is solid. It will definitely help with writing the discussion a bit. And is probably the focal point for a whole paragraph. We can make an interactive version and host it on github and link to it, too. So people can see what species are which....

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Josh-Lee1/eBird-Fire-Index/issues/15#issuecomment-689894630, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOENO3TOSFL54BWMAPBAGA3SFAMEXANCNFSM4QP7FDRA .

[coreytcallaghan]

I don't think it is necessary for all traits, but tailor it to what we want to write our discussion around, in my opinion.

@wcornwell can you just paste the df underlying this plot into an RDS somewhere and let me know where. I'll try making an interactive figure and putting it in this repository.

[wcornwell]

library(readr)
library(plotly)
oo<-read_csv("processed_data/severity_data_and_traits.csv")
z<-ggplot(oo,aes(x=median.sev,y=-1*Estimate,label=COMMON.NAME,))+
  geom_point(aes(col=feeding_specialisation))+geom_smooth(method="lm")+ylab("modeled response to fire")+
  geom_hline(yintercept = 0,linetype="dashed")+xlab("Median fire severity in post-fire observations")+
  theme_classic()

ggplotly(z)

[coreytcallaghan]

Hey @Josh-Lee1 can you go into your settings and flick github pages on?

Should look something like this:

[Josh-Lee1]

Ok done. Let me know if I clicked the wrong button.

[coreytcallaghan]

Sweet - this link should be live now and sharable with anyone: https://josh-lee1.github.io/eBird-Fire-Index/interactive_figure.html

[Josh-Lee1]

Cool! Thanks Corey.

On Thu., 10 Sep. 2020, 11:02 Corey Callaghan, notifications@github.com wrote:

Sweet - this link should be live now and sharable with anyone: https://josh-lee1.github.io/eBird-Fire-Index/interactive_figure.html

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Josh-Lee1/eBird-Fire-Index/issues/15#issuecomment-689906619, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOENO3Q4ZPYZKCG6WBVVZT3SFAQQ3ANCNFSM4QP7FDRA .

[coreytcallaghan]

We can pretty it up once we have an idea of what the final figure will look like.

[Josh-Lee1]

Hey @wcornwell Just want to clarify that you have called anything with 2 or more feeding guilds a generalist and specialists only have 1? Thanks