Dates
Choose a date range from within your data file for analysis. If left blank, all data for the selected animals will be included in the analysis.
-
Animals
Choose one or more animals for analysis and visualisation. ZoaTrack can generate layers for multiple animals at a time; however, the processing time is positively related to the number of animals and location fixes included in the analysis.
Select all
Vero Tutuala - Olive ridley sea turtle
Lencawonu Jaco - Hawksbill sea turtle
Dailelesi - Hawksbill sea turtle
Kitawelai - Hawksbill sea turtle
Liboma - Hawksbill sea turtle
Lulu - Green sea turtle
Cynthia - Green sea turtle
Niquole - Green sea turtle
Asirara Jaco - Hawksbill sea turtle
Salauni - Green sea turtle
Project layers

The trajectory is the animal movement path created from the location fixes in chronological order. Unless the date range has been specified, ZoaTrack plots the trajectory from the first to the last location fix in the uploaded file.

Detections are a collection of data points containing temporal (i.e. the time and date of collection) and spatial data (i.e. the geographical coordinates). Unless the date range has been specified, ZoaTrack plots all the detections in the uploaded file.

This function adds colouration to the points that indicate the start (green) and end (red) of the animal track data-set.

Analysis layers

Otherwise known as a convex hull, this approach uses the smallest area convex set that contains the location data (Worton 1998). This calculation is undertaken within R using the adehabitatHR package (Calenge 2008). ZoaTrack will return the MCP calculation in the Analysis Results window. An image of the MCP will be produced for visualisation over the map image, as well as a KML file for viewing in Google Earth.

References

Calenge, C. (2006) The package adehabitat for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling, 197, 516-519

Worton, B.J. (1995) A convex hull-based estimator of home-range size. Biometrics, 51, 1206-1215.

Read more
Percent

ZoaTrack offers the feature to strip the Minimum Convex Polygon (MCP) to different levels based on percentage. At 100% the MCP will be the equivalent to the area covered by all locations within the dataset. Inserting a lower % value into the box will result in only this percentage of locations being contained in the final MCP.

%

The fixed kernel density estimator is a non-parametric method of home-range analysis, which uses the utilization distribution to estimate the probability that an animal will be found at a specific geographical location. This fixed method of kernel smoothing ignores the temporal sequence whereby locations were obtained, and assumes that all locations from that individual are spatially autocorrelated. This means that the location of an individual at a particular point implies an increased probability that it frequents neighbouring locations as well. The kernel UD accurately estimates areas of high use by the tagged animal, providing that the level of smoothing is appropriate.

These calculations are undertaken within R using the adehabitatHR library of functions (Calenge 2008).

References

Calenge, C. (2006) The package adehabitat for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling, 197: 516-519

Seaman, D.E., Powell, R.A. (1996) An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology, 77: 2075-2085.

Silverman, B.W. (1986) Density estimation for statistics and data analysis. Chapman and Hall, London, UK

Worton, B.J. (1989) Kernel methods for estimating the utilization distribution in home-range studies. Ecology 70: 164-168

Read more
Percent

The user can specify what volume contour they wish to extract from the utilisation distribution. This contour represents the boundary of that area which contains x% of the volume of the utilisation distribution. Inserting a lower value into the box will extract areas of a higher probability of usage. The 95% and 50% volume contour are those most commonly adopted as the home-range and core-area UD, respectively.

%
h estimator

There are a number of different smoothing parameters that have been adopted in kernel estimates, and no single parameter will perform well in all conditions. ZoaTrack offers three options for selecting the kernel smoothing parameter. Two of these are automatically generated using either the ad hoc method for a bivariate normal kernel (reference bandwidth = ‘href’), or the least-squares cross-validation (‘LSCV’) algorithm. Note, the LSCV approach can take some time to compute the parameter h and, in some cases, will fail to compute this h value. See Seaman and Powell (1998) for more information.

http://cran.r-project.org/web/packages/adehabitatHR/adehabitatHR.pdf

The Kernel Brownian Bridge approach calculates the utilization distribution to estimate the probability that an animal will be found at a specific geographical location. Unlike the classical Fixed Kernel approach, the Kernel Brownian Bridge incorporates serial autocorrelation between fixes (i.e. the time the animal took to move between locations) into the calculation (Bullard 1992). Brownian Bridges, therefore, only contribute to utilisation distributions when sequential locations (i.e. XYn at timen and XYn+1 at timen+1) occur close to one another in time (Kie et al. 2010).

This function uses information on the animal’s trajectory, how long it took to move between locations, how fast the animal moves on average (Sig1) and the uncertainty around each location fix (Sig2) to control the degree of kernel smoothing (Bullard 1992).

These calculations are undertaken within R using the adehabitatHR library of functions (Calenge 2008).

References

Bullard, F. (1991) Estimating the home range of an animal: a Brownian bridge approach. Master of Science, University of North Carolina, Chapel Hill

Calenge, C. (2006) The package adehabitat for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling, 197, 516-519

Horne, J.S., Garton, E.O., Krone, S.M. and Lewis, J.S. (2007) Analyzing animal movements using brownian bridge. Ecology, in press

Kie J.G., Matthiopoulos J., Fieberg J., Powell R.A., Cagnacci F., Mitchell M.S., Gaillard J.M., Moorcroft P.R. 2010. The home-range concept: are traditional estimators still relevant with modern telemetry technology? Phil. Trans. R. Soc. B 365, 2221–2231

Read more
Percent

The user can specify what volume contour they wish to extract from the utilisation distribution. This contour represents the boundary of that area which contains x% of the volume of the utilisation distribution. Inserting a lower value into the box will extract areas of a higher probability of usage. The 95% and 50% volume contour are those most commonly adopted as the home-range and core-area UD, respectively.

%
sig1

ZoaTrack offers two smoothing parameters for the Brownian Bridge method. Sig1 is related to the speed of the animal, describing how far from the line joining two successive locations the animal can go in one time step. This can be entered manually or can be estimated within R using the liker function in adehabitatHR library of functions (Calenge 2008). See the adehabitatHR package notes for more information.

http://cran.r-project.org/web/packages/adehabitatHR/adehabitatHR.pdf

sig2

The second smoothing parameter for the Brownian Bridge method is Sig2. This is related to the imprecision of the animal locations, which is assumed to be known from static trials, and is the equivalent of the h parameter in the classical fixed kernel approach (Calenge 2008). See the adehabitatHR package notes for more information.

http://cran.r-project.org/web/packages/adehabitatHR/adehabitatHR.pdf

The alpha hull home range estimation is a generalisation of the convex hull but objectively crops low use areas from the polygon surface. Alpha hulls are generated by connecting all locations as a Delauney triangulation, then systematically removing vertices until only those vertices that are shorter in length than the chosen parameter value alpha are retained. The smaller the value of alpha, the finer the resolution of the hull and the greater the exposure of non-use areas. As alpha increases, the polygon surface will increase until it is equivalent to a 100% minimum convex polygon.

This calculation is undertaken within R using the alphahull package (Pateiro-Lopez & Rodriguez-Casal 2011). The analysis is heavy on computing resources and can take up to 20 minutes to calculate depending on the number of location fixes. Be patient.

References

Burgman, M.A. & Fox, J.C. (2003) Bias in species range estimates from minimum convex polygons: implications for conservation and options for improved planning. Animal Conservation, 6, 19-28.

Read more
Alpha

The smaller the value of alpha, the finer the resolution of the hull and the greater the exposure of non-use areas. As alpha increases, the polygon surface will increase until it is equivalent to a 100% minimum convex polygon.

The Local Convex Hull (LoCoH) method estimates individual utilisation distributions based on the local nearest-neighbour convex hulls. These are formed by constructing convex hulls around each location in the animal’s trajectory then jointing these hulls together, iteratively, to form isopleths (Getz 2007). This is a useful home-range estimator when the movements of the animal have been constrained along hard edges such as roads, fences and rivers.

This calculation is undertaken within R using the LoCoH series of functions within the adehabitatHR library of functions (Calenge 2008). Users may either fix the number of nearest neighbours (k-1) to the root point (i.e. the fixed k-LoCoH), or fix the maximum radius from root points when generating local hulls (i.e. the fixed r-LoCoH). This analysis is heavy on computing resources and can take up to 20 minutes to calculate depending on the number of location fixes. Be patient.

References

Calenge, C. (2006) The package adehabitat for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling, 197, 516-519

Getz, W.M. & Wilmers, C.C. (2004). A local nearest-neighbor convex-hull construction of home ranges and utilization distributions. Ecography, 27, 489–505.

Getz, W.M., Fortmann-Roe, S.B, Lyons, A., Ryan, S., Cross, P. (2007). LoCoH methods for the construction of home ranges and utilization distributions. PLoS ONE, 2: 1–11.

Read more
Percent

For the Local Convex Hull, the lower the percent value for the home range contour/isopleth, the greater the probability of finding the individual in that region. The 100% isopleth contains all the locations, while the 50% isopleth contains 50% of the locations.

%
Neighbours

By entering a value in this field, ZoaTrack runs the fixed k-LoCoH (function LoCoH.k) contained within the adehabitatHR package in R. Here neighbours = the number of neighbours to include (k) – 1 with which to construct the convex hulls. See the adehabitatHR package notes for more information.

http://cran.r-project.org/web/packages/adehabitatHR/adehabitatHR.pdf

Radius

By entering a value in this field, ZoaTrack runs the fixed r-LoCoH (function LoCoH.r) contained within the adehabitatHR package in R. Here radius = the distance (in meters) from the root point with which to include locations in convex hulls. See the adehabitatHR package notes for more information.

http://cran.r-project.org/web/packages/adehabitatHR/adehabitatHR.pdf

m

This generates a grid over the study area and uses a coloured gradient to visually identify areas of high usage by the tagged animal. These can be applied to either points or connectivity lines between points. The size of the grid cells (in meters) can be specified. This ZoaTrack tool utilises the spatstat package in R (Baddeley & Turner, 2005)

References

Baddeley, A. & Turner, R. (2005) spatstat: An R package for analyzing spatial point patterns. Journal of Statistical Software, 12,6

Read more
Show absence
Grid size m
Colours

This generates a grid over the study area and uses a coloured gradient to visually identify areas of high usage by the tagged animal. These can be applied to either points or connectivity lines between points. The size of the grid cells (in meters) can be specified. This ZoaTrack tool utilises the spatstat package in R (Baddeley & Turner, 2005)

References

Baddeley, A. & Turner, R. (2005) spatstat: An R package for analyzing spatial point patterns. Journal of Statistical Software, 12,6

Read more
Show absence
Grid size m
Colours

This filter applies a state-space model combined with a kalman filter to your location data (logitude and latitude) to help predict the ‘most probable’ track through time. This combines the location and temperature data. More information on this particular method (i.e. the extended kalman filter) can be found in Sibert et al. (2003), Sibert et al. (2006) and the kftrack package documentation on which this code is based at https://code.google.com/p/geolocation/wiki/ArticleKftrack.

References

Sibert, J.R., Musyl, M.K. & Brill, R.W. (2003) Horizontal movements of bigeye tuna (Thunnus obesus) near Hawaii determined by Kalman filter analysis of archival tagging data. Fisheries Oceanography, 12(3):141-151.

Sibert, J. R., Lutcavage, M.E., Nielsen, A., Brill, R.W. & Wilson, S.G. (2006) Interannual variation in large-scale movement of Atlantic bluefin tuna (Thunnus thynnus) determined from pop-up satellite archival tags. Canadian Journal of Fisheries and Aquatic Sciences 63: 2154-2166.

Read more
Start date
Start longitude
Decimal degrees east
°
Start latitude
Decimal degrees north
°
End date
End longitude
Decimal degrees east
°
End latitude
Decimal degrees north
°

If tag-recorded sea-surface temperature (SST) data are available for each location, this filter applies a state-space model combined with a kalman filter to your location data (logitude, latitude) and SST to help predict the ‘most probable’ track through time. In an attemt to improve the model’s predictions, the tag-recorded SST is matched with external SST data collected by the National Oceanic and Atmospheric Administration (NOAA). This model is adapted from the kfsst function, in the ukfsst R package https://code.google.com/p/geolocation/wiki/ArticleUkfsst. More information on this particular method can be found in Lam, Nielsen & Sibert (2008).

References

Lam, C.H., Nielsen, A. & Sibert, J.R. (2008) Improving light and temperature based geolocation by unscented Kalman filtering. Fisheries Research, 91: 15-25.

Read more
Start date
Start longitude °
Start latitude °
End date
End longitude °
End latitude °
The graphs are interactive. The smaller graph is a navigator window for the larger graph. You can:
  • Zoom in by double-clicking on the graph canvas, or by using a mouse scroll
  • Zoom out by shift-double-clicking
  • Drag on either canvas to pan the graph
  • Hover over a data point to see detail
  • Click on an animal in the legend to add/remove it from the graph

Project Summary

Species

Chelonia mydas, Eretmochelys imbricata
green sea turtle, hawksbill sea turtle

Location

Papua New Guinea, Solomon Islands, Timor L'Este

Date Range

2017-12-05 to 2018-10-07

Animals

10 animals

Spatial Reference System

EPSG:3577

Open Access

The data in this project is publicly available under a Creative Commons Attribution-NonCommercial-NoDerivs License. If you use these data in any type of publication then you must cite the project DOI (if available) or any published peer-reviewed papers associated with the study. We strongly encourage you to contact the data custodians to discuss data usage and appropriate accreditation.

Contact