Individual Movement Strategies Revealed through Novel Clustering of Emergent Movement Patterns - PubMed Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Mar 8:7:44052.
doi: 10.1038/srep44052.

Individual Movement Strategies Revealed through Novel Clustering of Emergent Movement Patterns

Denis Valle  1 , Sreten Cvetojevic  1 , Ellen P Robertson  2 , Brian E Reichert  3 , Hartwig H Hochmair  1 , Robert J Fletcher  2
Affiliations

Individual Movement Strategies Revealed through Novel Clustering of Emergent Movement Patterns

Denis Valle et al. Sci Rep. .

Abstract

Understanding movement is critical in several disciplines but analysis methods often neglect key information by adopting each location as sampling unit, rather than each individual. We introduce a novel statistical method that, by focusing on individuals, enables better identification of temporal dynamics of connectivity, traits of individuals that explain emergent movement patterns, and sites that play a critical role in connecting subpopulations. We apply this method to two examples that span movement networks that vary considerably in size and questions: movements of an endangered raptor, the snail kite (Rostrhamus sociabilis plumbeus), and human movement in Florida inferred from Twitter. For snail kites, our method reveals substantial differences in movement strategies for different bird cohorts and temporal changes in connectivity driven by the invasion of an exotic food resource, illustrating the challenge of identifying critical connectivity sites for conservation in the presence of global change. For human movement, our method is able to reliably determine the origin of Florida visitors and identify distinct movement patterns within Florida for visitors from different places, providing near real-time information on the spatial and temporal patterns of tourists. These results emphasize the need to integrate individual variation to generate new insights when modeling movement data.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Figure 1
Figure 1. Simulated networks (left panels) and inference provided by different methods (middle and right panels).
Left and middle panels (panels A, B, D, E, G, H, J and K) depict the simulated links and the links estimated by IBC, respectively, between sites assuming a 0.02 cuttoff. Right panels depict groups of locations estimated by the map equation method (ME; panels C and F) and modularity maximization (MM, panels I and L). Solid circles denote locations visited by a single group while hollow circles are locations visited by multiple groups. Panels A-F: simulated networks and inference for scenarios with different amount of mixed membership sites. Increased mixed membership is emphasized in panels D-F with grey polygons. The MM and latent cluster (LC) algorithms identified only 3 groups for scenario 1. For scenario 2, MM and LC algorithms identified 3 and 1 group, respectively (results not shown). Panels G-L: simulated networks and inference for scenarios where movement patterns change through time.
Figure 2
Figure 2. Individual variation in movement strategies by snail kites illustrate that these strategies vary by birth year and that some individuals show high site fidelity while others move throughout the geographic range.
Maps: Numbers between parentheses refer to the proportion of all individuals in each group. Inserted text refers to the name of the main critical connectivity sites, where TOHO, KISS, OKEE, and WCA3A stand for Lake Tohopekaliga, Lake Kissimmee, Lake Okeechobee, and Water Conservation Area 3A, respectively. The size of blue circles is proportional to visitation rate ψkl and dashed lines connect all sites for which ψkl > 0.05. Boxplots: this plot shows the distribution of year of birth for individuals in each group. Bottom right graph: Visitation rate for different sites for the 7 groups (different colored lines). Sites are ordered from south to north (left to right). Maps were created in R (version 3.3.1) and the base map comes from the freely available GADM database.
Figure 3
Figure 3. Changes in landscape use through time with increased prey availability from an exotic snail invasion.
Panels depict changes in visitation rate for the five groups that jointly represent 97% of all individuals. The y-axis displays the difference of visitation rate for each site between periods t1 and t2, given by formula image. Circle sizes scale with the number of individuals in each group. Significant differences between time periods are shown in black (non-invaded sites) and red (invaded sites) while non-significant differences are shown in grey (non-invaded sites) and pink (invaded sites). Differences for which the 95% credible interval did not include 0 were deemed statistically significant.
Figure 4
Figure 4. High correspondence between our estimates and independent estimates of state of origin of national visitors (left panel) and country of origin of international visitors (right panel).
Independent estimates of visitation rate were extracted from ref. . The estimated visitation rate is based on summer data but is identical when using winter data. A 1:1 line was added for reference (diagonal grey line).
Figure 5
Figure 5. Differences in movement patterns for national and international Florida visitors.
Left panels: National visitors travel more within Florida than international visitors and both types of visitors travel more within Florida during the winter than during the summer. Results from all 25 groups are presented simultaneously. Circle size for location l is proportional to the probability of visiting this location and belonging to group k (i.e., formula image). Darker lines indicate links from groups with more individuals. Right panels: Visitation rate for the 10 largest groups (y-axis) is shown for a subset of Florida counties. Only counties that were visited 2% of the times by more than one group are shown. Red boxes highlight the counties with the four largest international airports in Florida, as determined by passenger boarding in 2013. Maps were created in R (version 3.3.1) and base map comes from the freely available database in the “maps” package in R (version 3.1.1).
Figure 6
Figure 6. Differences in movement patterns according to state or country of origin of Florida visitors.
Left panels: visitors from Georgia travel more, with greater visitation rate of northern Florida counties, than visitors from New York. Right panels: visitors from the United Kingdom (UK) travel more within Florida than Brazilian visitors. Results from all 25 groups are presented simultaneously. Circle size for location l is proportional to the probability of visiting this location and belonging to group k (i.e., formula image). Darker lines indicate links from groups with more individuals. Maps were created in R (version 3.3.1) and base map comes from the freely available database in the “maps” package in R (version 3.1.1).
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Eagle N. & Pentland A. Reality mining: sensing complex social systems. Pers Ubiquit Comput 10, 255–268 (2006).
    1. Lu X., Bengtsson L. & Holme P. Predictability of population displacement after the 2010 Haiti earthquake. Proc Natl Acad Sci USA 109, 11576–11581 (2012). - PMC - PubMed
    1. Wesolowski A. et al.. Quantifying the impact of human mobility on malaria. Science 338, 267–270 (2012). - PMC - PubMed
    1. Tatem A. J. & Smith D. L. International population movements and regional Plasmodium falciparum malaria elimination strategies. Proc Natl Acad Sci USA 107, 12222–12227 (2010). - PMC - PubMed
    1. Guo D. Visual analytics of spatial interaction patterns for pandemic decision support. Int J Geogr Inf Sci 21, 859–877 (2007).

Publication types