Survey data and code to estimate abundance of Brachyramphus murrelets, Icy Bay, Alaska, USA
Data files
Oct 01, 2024 version files 1.09 GB
-
DataCode_BRMU_KISSLING_et_al.zip
1.09 GB
-
README.md
11.63 KB
Abstract
A fundamental aspect of ecology is identifying and characterizing population processes. In population studies, we almost always use sampling to make inference about the biological population, which is the population of interest, and the part of the population at risk of sampling is referred to as the statistical population. Ideally, the statistical population is the same as, or accurately represents its corresponding biological population. However, in practice, they rarely align in space and time, which can lead to biased inference. We often view a population misalignment as a temporary emigration process and resolve it with replicate and/or repeat sampling, though this approach is not feasible for all species and habitats. We developed a hierarchical framework to estimate abundance of a biological population of the Kittlitz’s murrelet (Brachyramphus brevirostris), a highly mobile, non-territorial, ice-associated seabird of conservation concern in Alaska and eastern Russia. Our model combines datasets from boat and telemetry surveys to account for all components of detection probability, specifically using telemetry locations to estimate probability of presence (pp) and line transect distance sampling to estimate probability of detection (pd). By estimating pp directly, we were able to account for temporary emigration from the sampled area, which changed with shifting icefloes between sampling occasions. Between 2007 and 2012, annual pp was highly variable, ranging from 0.33 to 0.75 (median=0.50, standard deviation=0.02), but was not predictable using five environmental covariates. In years when two boat surveys were conducted, our model reduced the coefficient of variation (CV) of abundance estimates for the biological population compared to the statistical population by 13–35%, yet in the year with only one boat survey (2009), the CV skyrocketed about 10-fold, emphasizing the importance of a second survey if pp varies. Although we increased the precision of annual abundance estimates by accounting for pp, we did not see the same improvement in the temporal trend estimate. This result indicates that while we reduced within-year variance, we failed to account for a source(s) of variation across years, which we suspect is related to the propensity for murrelets to skip breeding in some years. Our model to account for a population misalignment is simple, flexible, and scalable for generating unbiased and precise abundance estimates of highly mobile species that occupy dynamic habitats where other open population models are not possible . Importantly, it improves inference of the biological population. We urge ecologists to think critically about the population in which they want to draw inference, especially as tracking technology improves and model complexity increases.
README: Survey data and code to estimate abundance of Brachyramphus murrelets, Icy Bay, Alaska, USA
Data description and file structure:
These data were analyzed and summarized in a manuscript submitted to PCI Ecology as part of a special issue for the EURING Analytical Conference Proceedings, April 2024. The title of the manuscript is “Using multiple datasets to account for misalignment between statistical and biological populations for abundance estimation.”
File structure is as follows:
DataCode_BRMU_KISSLING_et_al
|- Data
| |- AbundanceResults
| |- BoatSurveys
| |- TelemetrySurveys
|- Code
| |- JAGS
| | |- AbundanceEstimation
| | |- ProbPresenceEstimation
| | |- TrendEstimation
| |- R
| | |- AbundanceEstimation
| | |- ProbPresenceEstimation
| | |- TrendEstimation
|- SupplementalInformation
Data folder
All raw data files are included in the folder "Data." Boat and telemetry survey data were collected between 2005 and 2012, using methods described in the manuscript. Briefly, boat survey data were gathered along line transects with distance sampling, usually twice per year during 1–15 July (deviations occurred in 2009 and are described in the manuscript). Telemetry data were acquired from radio-tagged murrelets tracked throughout the breeding season (May–August) from 2007 to 2012. Ice condition data were georeferenced and summarized in a GIS. Weather data were downloaded from https://www.ncdc.noaa.gov/cdo-web/ for the Icy Bay weather station.
AbundanceResults folder
Includes one CSV file titled "AbundanceByWindow." Data fields are as follows:
- Year - survey year
- Type - window length (none = without probability of presence; mean = posterior mean probability of presence from 2007 to 2012; 1day = 1-day window of boat survey, etc.; see manuscript for more details)
- TypeCat - window length by numerical category
- Estimate - mean posterior estimate of abundance
- SD = standard deviation of estimate
- LCI = 95% lower credible interval
- UCI = 95% upper credible interval
- CV = coefficient of variation
NA and blank cells indicate no data are available.
BoatSurveys folder
Includes 14 CSV files titled BRMU_IB_Date_BS#, which indicates Brachyramphus Murrelet (BS), Icy Bay (IB), Survey Date, and Boat Survey (BS) Identifier. Note that data for 2016 and 2017 are not publicly available. See "Data and Code Availability Statement" in manuscript. Minimum data fields are as follows (not all fields are available in all years):
- Date - mm/dd/yyyy
- Transect - identification number
- Strata - Main Bay (main), Taan Fjord (taan)
- Weather - 0 = <50% cloud cover; 1 = >50% cloud cover; 2 = fog; 3 = mist to light rain; 4 = medium to heavy rain
- SeaState - following Beaufort classification; 0 = glossy, 1 = rippled with small chop < 1 ft; 2 = small chop with few whitecaps 1-2 ft; 3 = small chop with whitecaps 2-3 ft; 4 = choppy with lots of whitecaps/spray > 3 ft
- Swell - nearest foot (ft)
- Ice - none; trace (<1% surface area); 1%; 3%; 5%; 5-100% increments within 300 m of boat
- Lat - latitude (WGS 84)
- Long - longitude (WGS 84)
- Species - KIMU = Kittlitz's murrelets; MAMU = marbled murrelet; BRMU = unidentified Brachyramphus murrelet
- GroupSize - number in group within 3 m of each other
- Distance - perpendicular from transect line to center of group; measured in meters
- Behavior - on water or flying
- Comments - as needed
- Plumage - alternate, basic, molting
- ClosestDepth - measured in feet from closest endpoint
- ClosestEndpoint - closest endpoint to bird location
- Secchi - measured in meters
- SST - sea surface temperature in Celsius
- 0.5mTemp - ocean temperature at 0.5 m depth in Celsius
- 1mTemp - ocean temperature at 1.0 m depth in Celsius
- Time - time of location
- Tide - tidal stage (falling, rising)
- BinomialTide - binomial tidal stage (0 = falling, 1 = rising)
- FirstTideTime - time of previous tidal switch
- SecondTideTime - time of next tidal switch
- FirstTideHeight - height in feet of previous tidal switch
- SecondTideHeight - height in feet of next tidal switch
- TideFraction - fraction of tidal cycle complete
- RelativeStrength - relative strength of tidal cycle based on rule of twelfths (see Kissling et al. 2007 for explanation)
- TotalTideHeight - total tidal height swing in feet
- RelativeTideCurrentStrength - RelativeStrength multiplied by TotalTideHeight (see Kissling et al. 2007)
NA and blank cells indicate no data are available.
TelemetrySurveys folder
Includes three CSV files for (a) ice coverage in Icy Bay study area during each telemetry survey (IceAreaIB), (2) Yakutat weather (YAK_weather), and (3) telemetry survey data (TxLocsIceAnalysis_StrataWNests). Data fields are as follows:
1. IceAreaIB
- Date - mm/dd/yyyy
- AllIceArea - Area (km2) of total ice (all categories) in Icy Bay study area
- IceMinusBBArea - Area (km2) of all categories of ice except brash ice (sometimes called bergy bits; BB) in Icy Bay study area
- AllIceCoreArea - Area (km2) of total ice (all categories) in Icy Bay state (where boat surveys were conducted)
- PropIceCoreArea - Proportion of Icy Bay state inaccessible due to ice
2. YAK_weather
- Date - mm/dd/yyyy
- PRCP - total precipitation (inches)
- AWND - average daily wind speed (miles per hour)
3. TxLocsIceAnalysis_StrataWNests
- RecordID - unique record identifier
- Species - KIMU = Kittlitz's murrelets; MAMU = marbled murrelet
- BirdNo - unique bird identifier
- Year - year of data collection
- BirdId - unique tagged bird identifier (some birds were captured and/or tagged in more than one year)
- Frequency - telemetry frequency
- Attachment type - method used to attach transmitter
- FlightID - unique identifier for flight
- Date - mm/dd/yyyy of flight
- JulianDate - of flight
- Latitude - latitude of bird location (WGS 84)
- Longitude - longitude of bird location (WGS 84)
- Time - of location
- Activity - whether bird was diving, resting, or at a nest
- FlightDay - number of days after first tag deployed
- SeasonAttach - during early or late capture session (see Kissling et al. 2015 for more details)
- Age - age of bird
- Sex - sex of bird
- ReproStatus - reproductive status of bird
- Ice_cond - category of ice condition in which bird was located based on flight ice map; based on Bowditch classification
- Tide - tidal stage (falling, rising)
- BinomialTide - binomial tidal stage (0 = falling, 1 = rising)
- FirstTideTime - time of previous tidal switch
- SecondTideTime - time of next tidal switch
- FirstTideHeight - height in feet of previous tidal switch
- SecondTideHeight - height in feet of next tidal switch
- TideFraction - fraction of tidal cycle complete
- RelativeStrength - relative strength of tidal cycle based on rule of twelfths (see Kissling et al. 2007 for explanation)
- TotalTideHeight - total tidal height swing in feet
- RelativeTideCurrentStrength - RelativeStrength multiplied by TotalTideHeight (see Kissling et al. 2007)
- State - spatial state in which bird was located
- Strata - spatial strata in which bird was located
- InCore - binomial value describing whether bird was present in core sampled area or not (0 = not in core; 1 = in core)
NA and blank cells indicate no data are available.
Code folder
Includes two folders; one for JAGS code and another for R code.
JAGS folder
Includes three folders for each analysis: (a) AbundanceEstimation, (b) ProbPresenceEstimation, and (c) TrendEstimation.
1. AbundanceEstimation
Includes models to estimate abundance as follows:
MultiSurveyAbundance_BioPopN_Avg_pp - estimates biological population in years when two boat surveys were conducted and no telemetry data were available (2005, 2016, 2017)
MultiSurveyAbundance_BioPopN_W_pp - estimates biological population in years when two boat surveys were conducted and telemetry data were available (2007, 2008, 2010-2012)
MultiSurveyAbundance_StatPopN_No_pp - estimates statistical population in years when two boat surveys were conducted (all years except 2009)
SingleSurveyAbundance_BioPopN_W_pp - estimates biological population in years when one boat survey was conducted and telemetry data were available (2009)
SingleSurveyAbundance_StatPopN_No_pp - estimates statistical population in years when one boat survey was conducted (2009)
2. ProbPresenceEstimation
Includes models to estimate and cross validate probability of presence (pp):
ProbPresence - estimates probability of presence with all data
ProbPresence_CrossValidation - estimates probability of presence with portion of data and predicts probability of presence with other portion
3. TrendEstimation
Includes state space model to estimate trend of statistical and biological populations:
TrendEstimation_StateSpace - estimates temporal trend in abundance
R folder
Includes three folders for each analysis: (a) AbundanceEstimation, (b) ProbPresenceEstimation, and (c) TrendEstimation.
1. AbundanceEstimation
Includes R code to estimate abundance of statistical and biological populations for 2005 and 2007-2012.
2. ProbPresenceEstimation
Includes R code to estimate and cross validate probability of presence (pp).
3. TrendEstimation
Includes R code to estimate trend of statistical and biological populations.
Supplemental Information folder
Includes results of each model run by year and the temporal trend across years. A folder for each year and trend includes summary results (.txt), model output as an object (.RDS), and the MCMC diagnostic plots. Note that these files have a different naming format, with "no_corep" referencing the statistical population and "corep" referencing the biological population. All of the annual results are summarized in the AbundanceByWindow.csv file (Data/AbundanceResults), which was used to estimate trend.
Sharing/Access information
Boat survey data also available in the North Pacific Pelagic Seabird Database: https://www.usgs.gov/centers/alaska-science-center/science/north-pacific-pelagic-seabird-database
Other publications that describe and use these data:
Kissling, M.L., P.M. Lukacs, S.B. Lewis, S.M. Gende, and J. Waite. 2016. Breeding decisions of a declining seabird, the Kittlitz’s Murrelet. Marine Ornithology 44:171–182.
Kissling, M.L., S.M. Gende, S.B. Lewis, and P.M. Lukacs. 2015. Reproductive performance of Kittlitz’s Murrelet in a glaciated landscape, Icy Bay, Alaska. The Condor: Ornithological Applications 117:237–248.
Kissling, M.L., P.M. Lukacs, S.M. Gende, and S.B. Lewis. 2015. Multi-state mark-recapture model to estimate survival of a dispersed-nesting seabird, the Kittlitz’s murrelet. Journal of Wildlife Management 79:20–30.
Kissling, M.L., Lukacs, P.M., Lewis, S.B., Gende, S.M., Kuletz, K.J., Hatch, N.R., Schoen, S.K., and Oehlers, S. 2011. Distribution and abundance of the Kittlitz’s Murrelet Brachyramphus brevirostris in selected areas of southeastern Alaska. Marine Ornithology 39(1):3–11.
Kissling, M.L., M. Reid, P.M. Lukacs, S.M. Gende, and S. B. Lewis. 2007. Understanding abundance patterns of a declining seabird: implications for monitoring. Ecological Applications 17:2164–2174.
Code/Software
The RProject titled "DataCode_BRMU_Kissling_et_al" is provided for convenience. All code files are included in the folder "Code" and sub-folders "Jags" and "R," and are organized in folders for estimating abundance, probability of presence, and trend. Code files are annotated.
Methods
These data used to estimate abundance and trend of Kittlitz's murrelets were collected in Icy Bay, Alaska, USA between 2005 and 2012. Boat survey data were gathered along line transects using distance sampling, and telemetry data were collected by flying aerial telemetry flights to locate radio-tagged murrelets during the breeding season. Methods are summarized in numerous publications, including the current manuscript in review at Peer Community In Ecology and several others described in the README file.