2. Refining the Shortest Paths in-water

Index

1. Preparing the data
2. Refining the Shortest Paths in-water
3. Calculating utilization distribution
4. Calculating overlaps
5. Calculating areas in steps

The runRSP() function is used to recreate the shortest paths between pairs of acoustic detections. By default, the analysis will run for all transmitters detected, but you can determine also which transmitters you would like to include using tags. The detection data, station coordinates and the group of each tracked animal is passed on to RSP automatically by actel through the argument input. You must also include the name of the transition layer you created using the argument t.layer, e.g.:

library(actel)
library(RSP)

filtered_data <- explore(tz = "Europe/Copenhagen")

water.shape <- loadShape(shape = "my_study_area.shp", size = 0.0001)  
water.transition <- transitionLayer(water.shape, directions = 16)

rsp.results <- runRSP(input = filtered_data, t.layer = water.transition, coord.x = "Longitude", coord.y = "Latitude")

Note:

coord.x and coord.y must be the column names containing the coordinates in the spatial data frame.

The detection ranges of each listening station are also taken into account in the runRSP(). These will be used as the location errors for the dBBMM when calculating UD areas. A Range column can be included in the spatial.csv file for specifying the detection ranges (in metres) for each acoustic station if these are known. If the ‘Range’ column is not found, a default detection range of 500 m is automatically considered for each receiver with the warning:

Warning: Could not find a 'Range' column in the spatial data; assuming a range of 500 metres for each receiver.

Note:

The ‘Range’ column must already be present in the spatial.csv file when you run the explore() function for it to be incorporated in the analysis.

There is some uncertainty regarding the trajectory taken while animals move between a pair of consecutive acoustic detections. This uncertainty increases proportionally to the time taken to go from one place to another. By default, consecutive detections separated by more than 24 hours will be broken by the runRSP() into separate ‘tracks’ (defined by the max.time = 24 argument, in hours). This avoids the estimation of unrealistic behaviour when the animals do not get detected in any array for exceedingly long periods of time. Detections that occur isolated (e.g. more than 24-h before or after any other detection) are automatically excluded from analysis. The runRSP() will return the percentage of raw detections that can be used for refining the shortest paths when the analysis is finished:

M: Percentage of detections valid for RSP: 99.8%

Pairs of detections can occur either at the same station or at different stations. For consecutive detections on different stations, estimated positions are added at intervals of approximately a given distance argument in metres (250 m by default). Note that the added positions will be centered relative to the total distance, e.g., if the distance between two stations is 600 m, then two RSP positions will be added; one at 200 m and one at 400 m. On the other hand, if an animal is detected consecutively at the same station (with a time interval greater than the stipulated at the time.step argument), then estimated positions are added at that station location, over intervals of approximately time.step minutes. E.g. if a fish is detected at a station twice with a 22 minute interval, and time.step is set to 10, two estimated positions will be included.

While moving away from the first detection, the position errors gradually increase for each estimated position. This increase defaults to a 5% rate of the distance argument, but it can be specified in metres using er.ad. When the animal reaches half of the elapsed time/distance between the first and the second detection, the errors of estimated positions now gradually decrease as it approaches the second station where it got detected. This principle is used for both pairs of detections on different stations, and for consecutive detections at the same station:

A: consecutive detections on the same station; B: consecutive detections on different stations.

Note how the distances between consecutive RSP positions (points) vary around the distance argument (250 m by default) as they depend on the estuary shape and the shortest distances between stations. It is also possible to observe how the errors of the estimated positions (lines) gradually increase/decrease as the animals move between stations.

The dynamic Brownian Bridge Movement Model accounts for the speed at which animals move between consecutive detections to expand/contract the UD areas. Consequently, depending on your array configuration, estuary shape and species being tracked, you may find useful to adjust the distance and time.lapse arguments for recreating the most plausible movement patterns of the monitored animals.

Note: If your tracked animals were recaptured during the monitoring period you may want to provide these aditional locations to runRSP(). To do so, please provide the following dataset named recaptures.csv:

Please make sure you provide the Recapture.date timezone at local time. To include the respective recapture locations in the estimations of refined shortest paths, make sure to set recaptures = TRUE in runRSP(). The recapture locations will be used to start a new RSP track begining at the recapture location:

2.1. Exploring the runRSP results

Here are some examples of the runRSP() output:

1. In the $tracks object you can find metadata, stored individually for each tracked transmitter, on the identified tracks (Track) and their corresponding number of total acoustic detections (original.n), duration in hours (Timespan), and their corresponding validity (Valid):

Only the valid tracks are used by RSP to recreate the shortest in-water paths of tracked animals. The tracking data can be retrieved from the list $detections in which data is saved individually for each transmitter.

The RSP data can be found in the $detections object, stored individually for each tracked animal. There are two main types of RSP interpolation:

2. For consecutive detections on the same station:

Note:

Various columns were omitted in this display for simplicity.

The Position column in this dataset identifies the two consecutive acoustic detections (Receiver) from this animal. We can notice that they occurred on the same Station (Standard.name column); the first on 2018-03-07 00:43:49 and the second on 2018-03-07 01:11:45 (slightly less than 30 minutes from each other). Because this time difference is longer than the default time.lapse (10 minutes), the runRSP() estimated the intermediate positions (RSP) by repeating the station’s Longitude and Latitude and changing the Error parameter at a rate of 5% from the default distance argument (250 metres = 12.5 metres). Notice how the elapsed time was distributed evenly between the position intervals.

3. For consecutive detections on different stations:

Note:

Various columns were omitted in this display for simplicity.

Here the animal was detected first at the station St.1 on 2018-04-27 05:27:10, and then at the station St.2 on 2018-04-27 07:04:39. The runRSP() now calculated the shortest in-water path between stations, and we can see how the Error of added locations increased up to half-way, (575 metres on 2018-04-27 06:15:54), and then decreased back to 500 as the track approached the second station.

RSP also has a plotDensities() function, which allows you to investigate the distribution of elapsed time between consecutive acoustic detections:

You can set a group argument to generate a density plot for a particular group of interest. By default all monitored groups are used.

2.2. Visualizing the runRSP outputs

We can use plotTracks() to plot any of the tracks from the runRSP() output:

plotTracks(rsp.data1, base.raster = water.shape1, type = "both", tag = "R64K-4125", track = "Track_3")
plotTracks(rsp.data1, base.raster = water.shape1, type = "both", tag = "R64K-4125", track = "Track_3", land.col = "darkgreen")
plotTracks(rsp.data2, base.raster = water.shape2, type = "both", tag = "R64K-4545", track = "Track_9")

You can use the function suggestSize() in your base raster (shapefile) file to get suggested dimensions for a projected plot, and easily change the colour of the land mass using land.col. The function addStations() can be used to add the station locations to your plot.:

You can easily change the colour of the land mass using land.col.

If the animal you want to plot was recaptured and you want to include the recapture locations, you can use addRecaptures():

plotTracks(input = rsp.data, base.raster = water, tag = "A69-9004-485", track = 3) + 
  addStations(rsp.data) + addRecaptures(Signal = "485")

2.3. Distances travelled exclusively in-water

The getDistances() function can be used to obtain the distances travelled (in metres) during each RSP track. The column Loc.type shows you whether the distances were calculated only using the station locations or if they were calculated also accounting for the interpolated positions (added by RSP).

dist.table <- getDistances(rsp.results)
dist.table

We can use plotDistances() to compare the total distances travelled by each animal calculated using only the station locations and also including the RSP estimations:

plotDistances(dist.table)

Note:

You can view the data for a single group by using the argument group.

plotDistances(dist.table, compare = FALSE)

Note:

If you set compare = FALSE, only the RSP total distances travelled will be returned.

Proceed to Calculating utilization distribution

Back to top.