Question about incorporating telemetry data

74 views

Skip to first unread message

drd11

unread,

Aug 19, 2020, 3:20:18 PM8/19/20

to secr

My capture data consist of deer pellets genetically identified to individual. Unfortunately, we have very few recaptures of individuals but we do have telemetry data.

I have successfully run models where I used the telemetry data to estimate sigma and then declare sigma to be a fixed parameter.

Now I want toincorporate the telemetry data into the capthist object under type="independent" study design (we never captured pellets from our radio-collared deer). I'm thinking this should incorporate some uncertainty about sigma using the telemetry data rather than setting sigma to be a fixed parameter (correct?).

I have 4 sessions (years) with 1 occasion in each session. I appear to have successfully created the capthist object,

### Create trap object

trCH <- read.capthist(fnCH,"detector.txt", detector="count", verify=F)

summary(trCH)

I created the telemetry object, and get no error message but it appears there is only 1 animal in session 1, when my data contains telemetry data from 6 animals...

> ### Create telemetry object

> telem <- read.table(fntelem,col.names = c("session","uniqdeer","occasion","utm.easting","utm.northing","sex"))

> head(telem)

session uniqdeer occasion utm.easting utm.northing sex

1 1 8159_8160M2015 1 274642.8 4506815 M

2 1 8159_8160M2015 1 274871.4 4506519 M

3 1 8159_8160M2015 1 274487.8 4507548 M

4 1 8159_8160M2015 1 273530.1 4507194 M

5 1 8159_8160M2015 1 274435.6 4506969 M

6 1 8159_8160M2015 1 275044.4 4506739 M

> teCH <- read.telemetry(fntelem, verify=T)

No errors found :-)

> summary(teCH)

$`1`

Object class capthist

Detector type telemetry

Telemetry type independent

Counts by occasion

1 Total

n 1 1

u 1 1

f 1 1

M(t+1) 1 1

losses 0 0

detections 101 101

detectors visited 0 0

detectors used 0 0

Empty histories : 1

1 telemetered animals, 0 detected

101-101 locations per animal, mean = 101, sd = NA

Individual covariates

M:1

$`2`

Object class capthist

Detector type telemetry

Telemetry type independent

Counts by occasion

1 Total

n 2 2

u 2 2

f 2 2

M(t+1) 2 2

losses 0 0

detections 205 205

detectors visited 0 0

detectors used 0 0

Empty histories : 2

2 telemetered animals, 0 detected

102-103 locations per animal, mean = 102.5, sd = 0.71

Individual covariates

M:2

$`3`

Object class capthist

Detector type telemetry

Telemetry type independent

Counts by occasion

1 Total

n 2 2

u 2 2

f 2 2

M(t+1) 2 2

losses 0 0

detections 205 205

detectors visited 0 0

detectors used 0 0

Empty histories : 2

2 telemetered animals, 0 detected

102-103 locations per animal, mean = 102.5, sd = 0.71

Individual covariates

M:2

$`4`

Object class capthist

Detector type telemetry

Telemetry type independent

Counts by occasion

1 Total

n 1 1

u 1 1

f 1 1

M(t+1) 1 1

losses 0 0

detections 507 507

detectors visited 0 0

detectors used 0 0

Empty histories : 1

1 telemetered animals, 0 detected

507-507 locations per animal, mean = 507, sd = NA

Individual covariates

M:1

> sigmatelem <- RPSV(teCH, CC=TRUE)

> sigmatelem

$`1`

[1] 859.2692

$`2`

[1] 565.1725

$`3`

[1] 823.8823

$`4`

[1] 316.4994

and combined the two using addTelemetry

> ### Join telemetry and trapping objects

> CHI <- addTelemetry(trCH, teCH, type = 'independent')

No errors found :-)

Warning messages:

1: In (function (detectionCH, telemetryCH, type = c("concurrent", "dependent", :

covariates in telemetryCH do not match detectionCH so covariates discarded

2: In (function (detectionCH, telemetryCH, type = c("concurrent", "dependent", :

covariates in telemetryCH do not match detectionCH so covariates discarded

3: In (function (detectionCH, telemetryCH, type = c("concurrent", "dependent", :

covariates in telemetryCH do not match detectionCH so covariates discarded

4: In (function (detectionCH, telemetryCH, type = c("concurrent", "dependent", :

covariates in telemetryCH do not match detectionCH so covariates discarded

> summary(CHI)

$`1`

Object class capthist

Detector type count, telemetry

Telemetry type independent

Detector number 200

Average spacing 100 m

x-range 267457 278608 m

y-range 4503596 4517656 m

Usage range by occasion

1 2

min 0 0

max 1 1

Counts by occasion

1 2 Total

n 18 1 19

u 18 1 19

f 19 0 19

M(t+1) 18 19 19

losses 0 0 0

detections 21 101 122

detectors visited 19 0 19

detectors used 200 0 200

Empty histories : 1

1 telemetered animals, 0 detected

101-101 locations per animal, mean = 101, sd = NA

$`2`

Object class capthist

Detector type count, telemetry

Telemetry type independent

Detector number 200

Average spacing 100 m

x-range 267457 278608 m

y-range 4503596 4517656 m

Usage range by occasion

1 2

min 0 0

max 1 1

Counts by occasion

1 2 Total

n 25 2 27

u 25 2 27

f 27 0 27

M(t+1) 25 27 27

losses 0 0 0

detections 28 205 233

detectors visited 24 0 24

detectors used 200 0 200

Empty histories : 2

2 telemetered animals, 0 detected

102-103 locations per animal, mean = 102.5, sd = 0.71

$`3`

Object class capthist

Detector type count, telemetry

Telemetry type independent

Detector number 200

Average spacing 100 m

x-range 267457 278608 m

y-range 4503596 4517656 m

Usage range by occasion

1 2

min 0 0

max 1 1

Counts by occasion

1 2 Total

n 12 2 14

u 12 2 14

f 14 0 14

M(t+1) 12 14 14

losses 0 0 0

detections 13 205 218

detectors visited 11 0 11

detectors used 200 0 200

Empty histories : 2

2 telemetered animals, 0 detected

102-103 locations per animal, mean = 102.5, sd = 0.71

$`4`

Object class capthist

Detector type count, telemetry

Telemetry type independent

Detector number 200

Average spacing 100 m

x-range 267457 278608 m

y-range 4503596 4517656 m

Usage range by occasion

1 2

min 0 0

max 1 1

Counts by occasion

1 2 Total

n 43 1 44

u 43 1 44

f 44 0 44

M(t+1) 43 44 44

losses 0 0 0

detections 52 507 559

detectors visited 37 0 37

detectors used 200 0 200

Empty histories : 1

1 telemetered animals, 0 detected

507-507 locations per animal, mean = 507, sd = NA

I attached a plot of the mask, the captures, and telemetry data

but when I run the model the estimation routine does not vary from the starting values provided (or even when letting secr guess the starting values). It runs for about 40 iterations and just stops and always gets "-Inf" for the log-likelihood estimate.

Here's the secr.fit model statement. Each mask is the same for each session

Sigma <- secr.fit(capthist=CHI, model=list(D~session, g0~1, sigma~1),

start=c(-5,-.35,.33,-3.4,-3,6), detectfn="HHN",

ncores=6, mask=c("mask1","mask2","mask3","mask4"),

method="Newton-Raphson", verify=F)

Any mistakes being made? Is the problem with reading in the telemetry data?Suggestions for options to try? Is this approach not going to work if there are very few (if any) captures at different stations?

Thanks,

Duane

Mask capture telemetry objects.pdf

Murray Efford

unread,

Aug 27, 2020, 7:42:05 PM8/27/20

to secr

Hi there

Sorry this has been lying around for a while - my telemetry aversion shows, and I'm rusty on it.

Did you use the 'data' argument of read.telemetry? It seems to work for me with the attached data file and my working version of secr:

Murray

testdf <- read.table('teCHcapt.txt')
teCH2 <- read.telemetry(data = testdf, covnames = 'sex')
plot(teCH2, tracks = TRUE)
summary(teCH2)
# Object class       capthist 
# Detector type      telemetry 
# Telemetry type     independent 
# 
# Counts by occasion 
# 1 Total
# n                  12    12
# u                  12    12
# f                  12    12
# M(t+1)             12    12
# losses              0     0
# detections        120   120
# detectors visited   0     0
# detectors used      0     0
# 
# Empty histories :  12 
# 12 telemetered animals, 0 detected
# 10-10 locations per animal, mean =  10, sd = 0 
# 
# Individual covariates
# sex  
# F:7  
# M:5