retentionmort: Estimating User-Based Tagging Mortality and Tag Shedding in...

In retmort: Estimate User-Based Tagging Mortality and Tag Loss in Mark-Recapture Studies

Estimating User-Based Tagging Mortality and Tag Shedding in Field Mark-Recapture Studies

Description

This model estimates the percent loss in tagged animals at large for field-based recapture studies based on a linear decrease in survival and tag retention (including lost tags and missidentified tags) for five weeks per tagging cohort based on laboratory retention/survival studies. The retentionmort() function can be used following a recapture field study to estimate user-based tag loss in animals at large. The model is changed by linear regression coefficients of weekly tag loss rate, weekly mortality rate, and their respective intercepts. The coefficients used can be selected from the currently included list using the err input or be customized. This function is also capable of working with a cofactor with two conditions (e.g. class1 individuals and large individuals) to improve resolution for more specified studies.

Usage

retentionmort(
  nT,
  n_c1 = nT,
  TaL,
  c,
  R,
  err = 2,
  m_mort_c1 = NA,
  b_mort_c1 = NA,
  m_ret_c1 = NA,
  b_ret_c1 = NA,
  m_mort_c2 = NA,
  b_mort_c2 = NA,
  m_ret_c2 = NA,
  b_ret_c2 = NA
)

Arguments

nT	A vector of the number of tagged individuals for each tagging effort.
n_c1	(optional) A vector of the number of tagged individuals in one of two categorical variables. If this is not being used then n_c1 will be equal to nT.
TaL	A vector of the cumulative number of tagged individuals following each effort.
c	One value for the total number of tagging efforts.This value must be greater than or equal to 6.
R	A vector of the number of recaptured individuals per effort.
err	A value (between 1 and 26) that represents the weekly mortality rate and weekly tag loss rate from a preloaded case study listed in the metadata. Alternatively, model coefficients can be manually included using a combination of the preceding parameters. While the preloaded data are based on weekly time stamps, customized model coefficients can reflect any time period specified and the projection will predict loss at 5 times the time interval. 1 = Large (> 61mm TL) and class1 (< 61mm TL) Mummichogs tagged with VIE in caudal peduncle (avg + 95% CI) - McCutcheon et al. in prep 2 = Large (> 61mm TL) and class1 (< 61mm TL) Mummichogs tagged with VIE in caudal peduncle (avg) - McCutcheon et al. in prep 3 = Large (> 61mm TL) and class1 (< 61mm TL) Mummichogs tagged with VIE in caudal peduncle (avg - 95% CI) - McCutcheon et al. in prep 4 = American Eel elvers (80 – 149 mm TL) tagged with 2 VIE tags in anterior, posterior, central of body - Eissenhauer et al. 2024 https://doi.org/10.1002/nafm.11016 5 = Mummichogs (45 - 82 mm TL) tagged with 8mm PIT tags in abdominal cavity - Kimball & Mace 2020 https://doi.org/10.1007/s12237-019-00657-4 6 = Mummichogs (45 - 82 mm TL) tagged with 12mm PIT tags in abdominal cavity - Kimball & Mace 2020 https://doi.org/10.1007/s12237-019-00657-4 7 = Pinfish (45 - 82 mm TL) tagged with 8mm or 12mm PIT tags in abdominal cavity - Kimball & Mace 2020 https://doi.org/10.1007/s12237-019-00657-4 8 = Cichlids (29 - 59 mm TL) tagged with VIE in various locations on body - Jungwirth et al. 2019 https://doi.org/10.1007/s00265-019-2659-y 9 = River Shiners (36 - 49 mm TL) tagged with VIE using anesthesia in various locations - Moore & Brewer 2021 https://doi.org/10.1002/nafm.10607 10 = River Shiners (50 - 56 mm TL) tagged with 8 mm PIT using anesthesia in various locations - Moore & Brewer 2021 https://doi.org/10.1002/nafm.10607 11 = River Shiners (40 - 51 mm TL) tagged with VIE using no anesthesia in various locations - Moore & Brewer 2021 https://doi.org/10.1002/nafm.10607 12 = River Shiners (50 - 55 mm TL) tagged with 8 mm PIT using no anesthesia in various locations - Moore & Brewer 2021 https://doi.org/10.1002/nafm.10607 13 = Delta Smelt (> 70 mm FL) tagged with injected acoustic tag - Wilder et al. 2016 https://doi.org/10.1080/02755947.2016.1198287 14 = Delta Smelt (> 70 mm FL) surgically tagged with acoustic tag - Wilder et al. 2016 https://doi.org/10.1080/02755947.2016.1198287 15 = Rohu Carp tagged with floy tags under dorsal fin - Hadiuzzaman et al. 2015 https://www.researchgate.net/publication/289460932_Feasibility_study_of_using_floy_tag_and_visible_implant_fluorescent_elastomer_marker_in_major_carps 16 = Silver Carp tagged with floy tags under dorsal fin - Hadiuzzaman et al. 2015 https://www.researchgate.net/publication/289460932_Feasibility_study_of_using_floy_tag_and_visible_implant_fluorescent_elastomer_marker_in_major_carps 17 = Black Bullhead (mean TL = 153.3 mm) tagged with VIE near dorsal fin - Schumann et al. 2013 https://benthamopen.com/contents/pdf/TOFISHSJ/TOFISHSJ-6-41.pdf 18 = Bluegill (mean TL = 75.8 mm) tagged with VIE near dorsal fin - Schumann et al. 2013 https://benthamopen.com/contents/pdf/TOFISHSJ/TOFISHSJ-6-41.pdf 19 = Channel Catfish (mean TL = 127.9 mm) tagged with VIE near dorsal fin - Schumann et al. 2013 https://benthamopen.com/contents/pdf/TOFISHSJ/TOFISHSJ-6-41.pdf 20 = Juvenile Burbot (88 - 144 mm TL) tagged with coded wire tag on snout, periocular region, nape, pectoral fin base, dorsal fin base, and anal fin base - Ashton et al. 2013 https://doi.org/10.1080/02755947.2014.882458 21 = Delta Smelt adults (45 - 77 mm FL) and juveniles (20 - 40 mm FL) tagged with calein markers - Castillo et al. 2014 https://doi.org/10.1080/02755947.2013.839970 22 = Juvenile Seabass (mean 173 g) tagged with dummy acoustic transmitters in external or intraperitoneal cavity - Begout Anras et al. 2003 https://doi.org/10.1016/S1054-3139(03)00135-8 23 = Juvenile American Eels (113 - 175 mm TL) tagged with micro-acoustic transmitter in body cavity - Mueller et al. 2017 https://doi.org/10.1016/j.fishres.2017.06.017 24 = Juvenile European Eels (7 - 25 g) tagged with 12mm PIT tags - Jepsen et al. 2022 https://doi.org/10.1111/jfb.15183 25 = Adult Atlantic Croaker (147 - 380 mm TL) tagged with VIE tags in caudal fin - Torre et al. 2017 https://doi.org/10.1080/00028487.2017.1360391 26 = Adult Spot (65 - 222 mm FL) tagged with VIE tags in caudal fin - Torre et al. 2017 https://doi.org/10.1080/00028487.2017.1360391
m_mort_c1	A value that represents the slope of the mortality rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, b_mort_c1, m_ret_c1, and b_ret_c1 need to be used. The use of these coefficients will override the err term.
b_mort_c1	A value that represents the intercept of the mortality rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c1, m_ret_c1, and b_ret_c1 need to be used. The use of these coefficients will override the err term.
m_ret_c1	A value that represents the slope of the tag loss (represented as tag loss and missidentification) rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c1, b_mort_c1, and b_ret_c1 need to be used. The use of these coefficients will override the err term.
b_ret_c1	A value that represents the intercept of the tag loss (represented as tag loss and missidentification) rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c1, b_mort_c1, and m_ret_c1 need to be used. The use of these coefficients will override the err term.
m_mort_c2	A value that represents the slope of the mortality rate for the class2 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, b_mort_c2, m_ret_c2, and b_ret_c2 need to be used. The use of these coefficients will override the err term.
b_mort_c2	A value that represents the intercept of the mortality rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c2, m_ret_c2, and b_ret_c2 need to be used. The use of these coefficients will override the err term.
m_ret_c2	A value that represents the slope of the tag loss (represented as tag loss and missidentification) rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c2, b_mort_c2, and b_ret_c2 need to be used. The use of these coefficients will override the err term.
b_ret_c2	A value that represents the intercept of the tag loss (represented as tag loss and missidentification) rate for the class1 individuals. While all the preloaded datasets work in weekly time intervals, these can be customized to any time interval to match the sampling interval. The resulting model will then project mortality and tag loss for 5 times the time interval. If this value is added, then, at minimum, m_mort_c2, b_mort_c2, and m_ret_c2 need to be used. The use of these coefficients will override the err term.

Value

This returns a dataframe datacomp that contains summary information from each mark-recapture effort, several parameters used in the calculation of adjusted recaptures, and basic error values between the expected and observed number of recaptures. The datacomp dataframe can be used in the retentionmort_figure() function to generate some preliminary figures that can be used to assess model performance and factors that influence the error between expected and observed recaptures.

      Values that will be returned include:
        week =      The week in the study
        nT =        The number of tagged individuals per tagging effort
        n_c1 =      The number of tagged individuals in class1 per tagging
                    effort
        TaL =       The cumulative number of tagged individuals at large
        TAsum =     The weekly sum of adjusted number of tags at large
        TDF =       Tag depreciation factor
        YSs =       Resultant survival rate of class1
        YSl =       Resultant survival rate of class2
        YMs =       Resultant tag loss rate of class1
        YMl =       Resultant tag loss rate of class2
        TaLs =      The cumulative number of class1 individuals tagged at
                    large
        TaLl =      The cumulative number of class2 individuals tagged at
                    large
        R =         The number of recaptured individuals per effort
        Rpercent =  The proportion of recaptured individuals
        RA =        The adjusted number of recaptured individuals
        PSE =       The percent standard error between observed and
                    estimated recaptured individuals

Examples

#To formulate a dataset for each example
  ret_env <- new.env()
  data<- test_dataset_retentionmort()
  list2env(data, envir = ret_env)


#Using a preloaded set of model parameters
  datacomp = retentionmort(n_c1=ret_env$n_c1, nT=ret_env$nT, TaL=ret_env$TaL,
                        c=ret_env$c, R=ret_env$R, err=ret_env$err
                        )

#Using custom model parameters for one class type
  datacomp = retentionmort(n_c1=ret_env$n_c1, nT=ret_env$nT, TaL=ret_env$TaL,
                        c=ret_env$c, R=ret_env$R, m_mort_c1=-0.0625,
                        b_mort_c1=1.06, m_ret_c1=-0.113, b_ret_c1=1.05
                        )
            #or
  datacomp = retentionmort(n_c1=ret_env$n_c1, nT=ret_env$nT, TaL=ret_env$TaL,
                        c=ret_env$c, R=ret_env$R, m_mort_c2=-0.0203,
                        b_mort_c2=1.03, m_ret_c2=-0.0541, b_ret_c2=0.993
                        )

#Using custom model parameters for two class types
  datacomp = retentionmort(n_c1=ret_env$n_c1, nT=ret_env$nT, TaL=ret_env$TaL,
                        c=ret_env$c, R=ret_env$R, m_mort_c1=-0.0625,
                        b_mort_c1=1.06, m_ret_c1=-0.113, b_ret_c1=1.05,
                        m_mort_c2=-0.0203, b_mort_c2=1.03, m_ret_c2=-0.0541,
                        b_ret_c2=0.993
                        )

retmort documentation built on April 12, 2025, 1:47 a.m.