What is the effect of the exclusion of less accurate 2D locations?

The exclusion of less accurate 2D locations (e.g. Lewis et al. 2007), reduces sampling effort and may therefore weaken statistical inferences of spatial utilization patterns (Frair et al. 2004).

What did Frair et al. (2004) say about the RSF model?

Frair et al. (2004) as well as Johnson & Gillingham (2008) however noted that RSF model coefficients are sensitive to location error and Frair et al. (2004) suggest that this was due to type II errors caused by data loss.

What is the reason why the accuracy of the methods is so low?

despite the apparent increase in location accuracy through the application of these methods, data lost due to missing locations cannot be remedied because loss is usually non-random and related to habitat variables (Rettie & McLoughlin 1999).

What is the effect of the filtered data on the number of 2D locations removed?

The number of 2D locations removed by filtering increased significantly with increasing canopy cover (2 × 3 contingency table: χ2 = 17.1, P = 0.0002, see Fig. 1c) but only marginally with slope (2 × 4 contingency table: χ2 = 9.65, P = 0.022, see Fig. 1d).

What is the argument for the inclusion of 2D locations in such analyses?

This may argue in favour of the inclusion of 2D locations in such analyses, especially because estimates of distances between locations are accurate when they are large relative to measurement error (Jerde & Visscher 2005), as was the case in their study.

What is the average fix rate for the canopy trials?

Fix rate was high (with an average of 70 ± 38.8% (mean ± S.D.) for canopy trials and 90 ± 18.8% for slope trials), decreasing significantly with increased canopy cover (Kruskal-Wallis non-parametric test: H = 9.78, P = 0.008 and Dunn’s post-hoc test, P < 0.05), but was not a function of slope (KruskalWallis non-parametric test: H = 6.57, P = 0.087).

What is the significance of the exclusion of 2D locations?

the exclusion of 2D locations significantly improved location accuracy at both the 50 and 95% CEP under various canopy (Wilcoxon signed rank test: W = 36.0, P = 0.008 and W = 45.0, P = 0.004, respectively), and slope categories (Wilcoxonsigned rank test: W = 153.0, P = 0.0003 and W = 171.0, P = 0.0002, respectively).

What did Hebblewhite et al. (2007) show?

Hebblewhite et al. (2007) show that collars on steep slopes have better satellite coverage than those on shallow slopes and this may explain why the authors recorded higher accuracies on steep slopes.

Open AccessJournal Article10.3957/056.040.0204

Inferred spatial use by elephants is robust to landscape effects on GPS telemetry

Q: What have the authors contributed in "Inferred spatial use by elephants is robust to landscape effects on gps telemetry" ?

Global Positioning System ( GPS ) telemetry provides data to study spatial utilization patterns of animals. The authors concluded that although the exclusion of 2D locations improved the accuracy of locations per se, it resulted in significant data loss. Most ( 75 % ) of the locations recorded in closed woodland were 2D locations, suggesting that the exclusion of 2D locations may reduce estimates of the utilization of such habitats.

Q: What test was used to determine if CEP was significantly lower for filtered data?

To determine if CEP was significantly lower for filtered than for unfiltered data for canopy and slope trial categories, the authors used the Wilcoxon signed rank non-parametric t-test.

Q: How many sites were used for each of the cover and slope trials?

The collars were moved to new pre-selected sites at least 500 m away from one another at ~24-hour intervals over a four-day period, yielding 15 and 20 sites for each of the cover and slope trials, respectively.

Q: How many collars have been used for elephant studies?

For instance, since 1999 a single supplier has assembled at least 800 collars for elephant (Loxodonta africana) studies in Africa (Sophie Haupt, pers.

Theresia Ott, +1 more

- 01 Oct 2010

- South African Journal of Wildlife Resear...

- Vol. 40, Iss: 2, pp 130-138

TL;DR: In this article, the exclusion of 2D locations from records obtained for GPS units deployed on free-ranging elephants (Loxodonta africana) increased daily displacement distances, and changed the number of locations per habitat.

Abstract: Global Positioning System (GPS) telemetry provides data to study spatial utilization patterns of animals. Spatial uncertainty due to poor accuracy and fix rates, however, may detract from inferences based on such data. The exclusion of two-dimensional (2D) locations may improve such inferences, but the prevalence of 2D locations may be a factor of landscape properties. In trials conducted using GPS units stationed at known positions, fix rate decreased significantly with increased canopy cover, but was unaffected by slope. Most (75%) of the locations recorded in closed woodland were 2D locations, suggesting that the exclusion of 2D locations may reduce estimates of the utilization of such habitats. Excluding 2D locations from records obtained for GPS units deployed on free-ranging elephants (Loxodonta africana) increased daily displacement distances, and changed the number of locations per habitat. However, selection ratios and estimates of home range area were not influenced by filtering location data. We concluded that although the exclusion of 2D locations improved the accuracy of locations per se, it resulted in significant data loss. This loss could alter inferences on patterns of spatial utilization.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What have the authors contributed in "Inferred spatial use by elephants is robust to landscape effects on gps telemetry" ?

Global Positioning System ( GPS ) telemetry provides data to study spatial utilization patterns of animals.. The authors concluded that although the exclusion of 2D locations improved the accuracy of locations per se, it resulted in significant data loss.. Most ( 75 % ) of the locations recorded in closed woodland were 2D locations, suggesting that the exclusion of 2D locations may reduce estimates of the utilization of such habitats.

2. What test was used to determine if CEP was significantly lower for filtered data?

To determine if CEP was significantly lower for filtered than for unfiltered data for canopy and slope trial categories, the authors used the Wilcoxon signed rank non-parametric t-test.

3. How many sites were used for each of the cover and slope trials?

The collars were moved to new pre-selected sites at least 500 m away from one another at ~24-hour intervals over a four-day period, yielding 15 and 20 sites for each of the cover and slope trials, respectively.

4. How many collars have been used for elephant studies?

For instance, since 1999 a single supplier has assembled at least 800 collars for elephant (Loxodonta africana) studies in Africa (Sophie Haupt, pers.

Table 1. Description of the canopy and slope categories used for stationary trials and field data. The canopy categories were based on structural vegetation units classified from satellite imagery in study sites across southern Africa. The shrubland category was only applied to field data.

Fig. 2. The 50% (a) and 95% (b) Circular Error Probable (CEP) as functions of location type during stationary trials. Significant differences are denoted by ***.

Fig. 1. Fix rate (a & b) and the number of locations (c & d) as a function of canopy cover and slope during stationary trials. Significant and non-significant differences are denoted by *** and NS, respectively.

Fig. 3. The effect of the exclusion of 2D locations on inferences on elephant spatial utilization: a, the number of locations recorded in different canopy categories; b, daily displacement distances; c & d, estimates of Kernel home range sizes;e, Resource Selection Function (RSF) forage ratios.The diagonal line (c–e) represents a one-to-one line where the differences between effects on variables calculated with 2D + 3D and 3D would be zero.The horizontal and vertical lines indicate the standard error of the mean for the 3D + 2D and 3D selection ratios for elephants within each park, respectively. Significant and non-significant differences are denoted by *** and NS, respectively.

Citations

Journal Article•10.1002/WSB.524

Retained satellite information influences performance of GPS devices in a forested ecosystem

Katie M. Moriarty, +1 more

- 01 Jun 2015

- Wildlife Society Bulletin

TL;DR: This study demonstrates that GPS tests should use the same fix schedules as field deployments, and GPS data quality in dense cover improves with short fix intervals, andMiniature GPS units are a promising tool, but the study design should be carefully considered.

...read moreread less

Title: Habitat Use and Movement Behavior of Pacific Marten (Martes caurina) in Response to Forest Management Practices in Lassen National Forest, California.

Clinton W. Epps

- 01 Jan 2014

TL;DR: In this article, the authors collected detailed movement data on Pacific martens (Martes caurina) in Lassen National Forest, California, during 2010-2013, to understand martens' behavior in forest patches that were altered by thinning to remove ladder fuels.

...read moreread less

•Dissertation

Determinants of protected area boundary crossings by savannah elephants, Loxodonta africana

Tamara Ann Lee

- 12 Apr 2013

TL;DR: In this article, the authors used telemetry locations of collared elephants in primary protected areas (PPAs) throughout southern Africa and investigated whether a range of intrinsic and extrinsic variables could explain why elephants crossed the boundaries of the PPAs.

...read moreread less

Journal Article•10.1007/S10329-012-0307-5

Evaluating home range techniques: use of Global Positioning System (GPS) collar data from chacma baboons

Paula A. Pebsworth, +2 more

- 05 Apr 2012

- Primates

TL;DR: The results demonstrate the need for primatologists to develop a standardized reporting method which documents the tool, screening protocol, and smoothing parameter used in the creation of home range estimations in order to make comparisons that are meaningful.

...read moreread less

•Posted Content•10.1186/S40317-021-00265-9

How often should dead-reckoned animal movement paths be corrected for drift?

Richard Gunner, +33 more

- 09 Jun 2021

- Animal Biotelemetry

TL;DR: Simulations were performed to examine the extent of dead-reckoning error, relative to VPs, as a function of Verified Position correction (VP correction) rate and the effect of this on estimates of distance moved and it is demonstrated how sources of measurement error can arise within VP-corrected dead- Reckoning tracks.

...read moreread less

References

Journal Article•10.2307/1938423

Kernel methods for estimating the utilization distribution in home-range studies

B. J. Worton

- 01 Feb 1989

- Ecology

TL;DR: Kernel methods are of flexible form and can be used where simple parametric models are found to be inappropriate or difficult to specify and give alternative approaches to the Anderson (1982) Fourier transform methods.

...read moreread less

4.3K

•Book

Resource selection by animals

D. Borchers, +3 more

- 01 Jan 1993

1.1K

Journal Article•10.1111/J.1365-2664.2005.01010.X

Effects of radio‐collar position and orientation on GPS radio‐collar performance, and the implications of PDOP in data screening

Robert G. D'eon, +1 more

- 01 Apr 2005

- Journal of Applied Ecology

TL;DR: It is suggested that collar orientation can safely be disregarded, whereas radio-collar position remains a large potential source of error and bias, which has major implications regarding animal activity and GPS radio-telemetry research.

...read moreread less

370

•Journal Article•10.1111/J.1365-2664.2007.01286.X

Effects of habitat on GPS collar performance: using data screening to reduce location error

Jesse S. Lewis, +3 more

- 01 Jun 2007

- Journal of Applied Ecology

TL;DR: In this paper, the authors quantify how canopy cover and terrain affected GPS collar performance within a mountainous region of northern Idaho, USA, and evaluate different data-screening options for GPS location data from stationary test collars and free-ranging black bears Ursus americanus.

...read moreread less

360

Journal Article•10.1006/ANBE.2001.1685

The implications of location accuracy for the interpretation of satellite-tracking data

Graeme C. Hays, +4 more

- 01 May 2001

- Animal Behaviour

TL;DR: Once more detailed aspects of Argos tracking data are considered, such as the speed of travel or small-scale movements, then location accuracy is likely to become a much more important issue.

...read moreread less

307

...

Expand

Inferred spatial use by elephants is robust to landscape effects on GPS telemetry

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What have the authors contributed in "Inferred spatial use by elephants is robust to landscape effects on gps telemetry" ?

2. What test was used to determine if CEP was significantly lower for filtered data?

3. How many sites were used for each of the cover and slope trials?

4. How many collars have been used for elephant studies?

5. What is the effect of the exclusion of less accurate 2D locations?

6. What did Frair et al. (2004) say about the RSF model?

7. What is the reason why the accuracy of the methods is so low?

8. What is the effect of the filtered data on the number of 2D locations removed?

9. What is the argument for the inclusion of 2D locations in such analyses?

10. What is the average fix rate for the canopy trials?

11. What is the significance of the exclusion of 2D locations?

12. What did Hebblewhite et al. (2007) show?

Figures

Citations

Retained satellite information influences performance of GPS devices in a forested ecosystem

Title: Habitat Use and Movement Behavior of Pacific Marten (Martes caurina) in Response to Forest Management Practices in Lassen National Forest, California.

Determinants of protected area boundary crossings by savannah elephants, Loxodonta africana

Evaluating home range techniques: use of Global Positioning System (GPS) collar data from chacma baboons

How often should dead-reckoned animal movement paths be corrected for drift?

References

Kernel methods for estimating the utilization distribution in home-range studies

Resource selection by animals

Effects of radio‐collar position and orientation on GPS radio‐collar performance, and the implications of PDOP in data screening

Effects of habitat on GPS collar performance: using data screening to reduce location error

The implications of location accuracy for the interpretation of satellite-tracking data

Related Papers (5)

Estimating habitat selection when GPS fix success is less than 100

Effects of sampling effort based on GPS telemetry on home-range size estimations

Detecting autocorrelation problems from GPS collar data in livestock studies

The trade-off between fix rate and tracking duration on estimates of home range size and habitat selection for small vertebrates.

Applying mixture models to derive activity states of large herbivores from movement rates obtained using GPS telemetry