Incorporating animal spatial memory in step selection functions

Luiz Gustavo R. Oliveira-Santos, James D Forester, Ubiratan Piovezan, Walfrido M. Tomas, Fernando A.S. Fernandez

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Memory is among the most important and neglected forces that shapes animal movement patterns. Research on the movement-memory interface is crucial to understand how animals use spatial learning to navigate across space because memory-based navigation is directly linked to animals' space use and home range behaviour; however, because memory cannot be measured directly, it is difficult to account for. Here, we incorporated spatial memory into step selection functions (SSF) to understand how resource selection and spatial memory affect space use of feral hogs (Sus scrofa). We used Biased Random Bridge kernel estimates linked to residence time as a surrogate for memory and tested four conceptually different dynamic maps of spatial memory. We applied this memory-based SSF to a data set of hog relocations to evaluate the importance of land cover type, time of day and spatial memory on the animals' space use. Our approach has shown how the incorporation of spatial memory into animal movement models can improve estimates of habitat selection. Memory-based SSF provided a feasible way to gain insight into how animals use spatial learning to guide their movement decisions. We found that while hogs selected forested areas and water bodies and avoided grasslands during the day (primarily at noon), they had a strong tendency to select previously visited areas, mainly those held in recent memory. Beyond actively updating their memory with recent experiences, hogs were able to discriminate among spatial memories encoded at different circadian phases of their activity. Even though hogs are thought to have long memory retention, they likely relied on recent experiences because the local food resources are quickly depleted and slowly renewed, yielding an uncertain spatial distribution of resources.

Original languageEnglish (US)
Pages (from-to)516-524
Number of pages9
JournalJournal of Animal Ecology
Volume85
Issue number2
DOIs
StatePublished - Mar 1 2016

Fingerprint

animal
animals
space use
swine
learning
resource selection
Sus scrofa
resource
relocation
habitat preferences
home range
land cover
habitat selection
body water
navigation
residence time
grasslands
grassland
spatial distribution
food

Keywords

  • Animal movement
  • Biased Random Bridge kernel estimation
  • Cognitive maps
  • GPS tracking
  • Habitat selection
  • Spatial memory
  • Sus scrofa

Cite this

Oliveira-Santos, L. G. R., Forester, J. D., Piovezan, U., Tomas, W. M., & Fernandez, F. A. S. (2016). Incorporating animal spatial memory in step selection functions. Journal of Animal Ecology, 85(2), 516-524. https://doi.org/10.1111/1365-2656.12485

Incorporating animal spatial memory in step selection functions. / Oliveira-Santos, Luiz Gustavo R.; Forester, James D; Piovezan, Ubiratan; Tomas, Walfrido M.; Fernandez, Fernando A.S.

In: Journal of Animal Ecology, Vol. 85, No. 2, 01.03.2016, p. 516-524.

Research output: Contribution to journalArticle

Oliveira-Santos, LGR, Forester, JD, Piovezan, U, Tomas, WM & Fernandez, FAS 2016, 'Incorporating animal spatial memory in step selection functions', Journal of Animal Ecology, vol. 85, no. 2, pp. 516-524. https://doi.org/10.1111/1365-2656.12485
Oliveira-Santos, Luiz Gustavo R. ; Forester, James D ; Piovezan, Ubiratan ; Tomas, Walfrido M. ; Fernandez, Fernando A.S. / Incorporating animal spatial memory in step selection functions. In: Journal of Animal Ecology. 2016 ; Vol. 85, No. 2. pp. 516-524.
@article{69e1b27bf3b9414cbd1519553bfb79f7,
title = "Incorporating animal spatial memory in step selection functions",
abstract = "Memory is among the most important and neglected forces that shapes animal movement patterns. Research on the movement-memory interface is crucial to understand how animals use spatial learning to navigate across space because memory-based navigation is directly linked to animals' space use and home range behaviour; however, because memory cannot be measured directly, it is difficult to account for. Here, we incorporated spatial memory into step selection functions (SSF) to understand how resource selection and spatial memory affect space use of feral hogs (Sus scrofa). We used Biased Random Bridge kernel estimates linked to residence time as a surrogate for memory and tested four conceptually different dynamic maps of spatial memory. We applied this memory-based SSF to a data set of hog relocations to evaluate the importance of land cover type, time of day and spatial memory on the animals' space use. Our approach has shown how the incorporation of spatial memory into animal movement models can improve estimates of habitat selection. Memory-based SSF provided a feasible way to gain insight into how animals use spatial learning to guide their movement decisions. We found that while hogs selected forested areas and water bodies and avoided grasslands during the day (primarily at noon), they had a strong tendency to select previously visited areas, mainly those held in recent memory. Beyond actively updating their memory with recent experiences, hogs were able to discriminate among spatial memories encoded at different circadian phases of their activity. Even though hogs are thought to have long memory retention, they likely relied on recent experiences because the local food resources are quickly depleted and slowly renewed, yielding an uncertain spatial distribution of resources.",
keywords = "Animal movement, Biased Random Bridge kernel estimation, Cognitive maps, GPS tracking, Habitat selection, Spatial memory, Sus scrofa",
author = "Oliveira-Santos, {Luiz Gustavo R.} and Forester, {James D} and Ubiratan Piovezan and Tomas, {Walfrido M.} and Fernandez, {Fernando A.S.}",
year = "2016",
month = "3",
day = "1",
doi = "10.1111/1365-2656.12485",
language = "English (US)",
volume = "85",
pages = "516--524",
journal = "Journal of Animal Ecology",
issn = "0021-8790",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - Incorporating animal spatial memory in step selection functions

AU - Oliveira-Santos, Luiz Gustavo R.

AU - Forester, James D

AU - Piovezan, Ubiratan

AU - Tomas, Walfrido M.

AU - Fernandez, Fernando A.S.

PY - 2016/3/1

Y1 - 2016/3/1

N2 - Memory is among the most important and neglected forces that shapes animal movement patterns. Research on the movement-memory interface is crucial to understand how animals use spatial learning to navigate across space because memory-based navigation is directly linked to animals' space use and home range behaviour; however, because memory cannot be measured directly, it is difficult to account for. Here, we incorporated spatial memory into step selection functions (SSF) to understand how resource selection and spatial memory affect space use of feral hogs (Sus scrofa). We used Biased Random Bridge kernel estimates linked to residence time as a surrogate for memory and tested four conceptually different dynamic maps of spatial memory. We applied this memory-based SSF to a data set of hog relocations to evaluate the importance of land cover type, time of day and spatial memory on the animals' space use. Our approach has shown how the incorporation of spatial memory into animal movement models can improve estimates of habitat selection. Memory-based SSF provided a feasible way to gain insight into how animals use spatial learning to guide their movement decisions. We found that while hogs selected forested areas and water bodies and avoided grasslands during the day (primarily at noon), they had a strong tendency to select previously visited areas, mainly those held in recent memory. Beyond actively updating their memory with recent experiences, hogs were able to discriminate among spatial memories encoded at different circadian phases of their activity. Even though hogs are thought to have long memory retention, they likely relied on recent experiences because the local food resources are quickly depleted and slowly renewed, yielding an uncertain spatial distribution of resources.

AB - Memory is among the most important and neglected forces that shapes animal movement patterns. Research on the movement-memory interface is crucial to understand how animals use spatial learning to navigate across space because memory-based navigation is directly linked to animals' space use and home range behaviour; however, because memory cannot be measured directly, it is difficult to account for. Here, we incorporated spatial memory into step selection functions (SSF) to understand how resource selection and spatial memory affect space use of feral hogs (Sus scrofa). We used Biased Random Bridge kernel estimates linked to residence time as a surrogate for memory and tested four conceptually different dynamic maps of spatial memory. We applied this memory-based SSF to a data set of hog relocations to evaluate the importance of land cover type, time of day and spatial memory on the animals' space use. Our approach has shown how the incorporation of spatial memory into animal movement models can improve estimates of habitat selection. Memory-based SSF provided a feasible way to gain insight into how animals use spatial learning to guide their movement decisions. We found that while hogs selected forested areas and water bodies and avoided grasslands during the day (primarily at noon), they had a strong tendency to select previously visited areas, mainly those held in recent memory. Beyond actively updating their memory with recent experiences, hogs were able to discriminate among spatial memories encoded at different circadian phases of their activity. Even though hogs are thought to have long memory retention, they likely relied on recent experiences because the local food resources are quickly depleted and slowly renewed, yielding an uncertain spatial distribution of resources.

KW - Animal movement

KW - Biased Random Bridge kernel estimation

KW - Cognitive maps

KW - GPS tracking

KW - Habitat selection

KW - Spatial memory

KW - Sus scrofa

UR - http://www.scopus.com/inward/record.url?scp=84958999229&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958999229&partnerID=8YFLogxK

U2 - 10.1111/1365-2656.12485

DO - 10.1111/1365-2656.12485

M3 - Article

C2 - 26714244

AN - SCOPUS:84958999229

VL - 85

SP - 516

EP - 524

JO - Journal of Animal Ecology

JF - Journal of Animal Ecology

SN - 0021-8790

IS - 2

ER -