Partial migration is a crucial mobility pattern in animal ecology. Unlike complete migrations which take place when all the individuals in a population migrate with a clear separation of ranges, partial migrations are migrations which often follow modalities and times that vary from animal to animal. As a result the distinction between migratory and non-migratory behavior becomes less defined. %Because of this, discriminating whether animals are resident or migratory is a challenging issue. In this paper, we present an interdisciplinary effort geared to evaluate whether a recent time-aware, density-based clustering technique, called SeqScan, relying on the novel concept of object's \emph{presence} can be effectively applied to the study of partial migrations. To that end, we propose an extended framework centered on SeqScan, comprising a noise model for the detection of fine-grained movement patterns, i.e. excursions and inter-cluster transitions, and an internal time-aware validity index, for clustering evaluation. Further, we contrast SeqScan with a recent technique developed in the context of animal ecology and grounded on statistical methods. For the study, we use real trajectories from two large herbivorous species located in Bavaria. We argue that the classification capabilities of SeqScan are comparable to those of the reference method. Moreover, the SeqScan framework overcomes important limitations of more conventional techniques, offering, in particular, the opportunity of quantifying the mobility behavior of individuals.

Introducing ‘presence’ and ‘stationarity index’ to study partial migration patterns: an application of a spatio-temporal clustering technique / M.L. Damiani, H. Issa, G. Fotino, M. Heurich, F. Cagnacci. - In: INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE. - ISSN 1365-8816. - 30:5(2016), pp. 907-928.

Introducing ‘presence’ and ‘stationarity index’ to study partial migration patterns: an application of a spatio-temporal clustering technique

M.L. Damiani
Writing – Original Draft Preparation
;
H. Issa;
2016

Abstract

Partial migration is a crucial mobility pattern in animal ecology. Unlike complete migrations which take place when all the individuals in a population migrate with a clear separation of ranges, partial migrations are migrations which often follow modalities and times that vary from animal to animal. As a result the distinction between migratory and non-migratory behavior becomes less defined. %Because of this, discriminating whether animals are resident or migratory is a challenging issue. In this paper, we present an interdisciplinary effort geared to evaluate whether a recent time-aware, density-based clustering technique, called SeqScan, relying on the novel concept of object's \emph{presence} can be effectively applied to the study of partial migrations. To that end, we propose an extended framework centered on SeqScan, comprising a noise model for the detection of fine-grained movement patterns, i.e. excursions and inter-cluster transitions, and an internal time-aware validity index, for clustering evaluation. Further, we contrast SeqScan with a recent technique developed in the context of animal ecology and grounded on statistical methods. For the study, we use real trajectories from two large herbivorous species located in Bavaria. We argue that the classification capabilities of SeqScan are comparable to those of the reference method. Moreover, the SeqScan framework overcomes important limitations of more conventional techniques, offering, in particular, the opportunity of quantifying the mobility behavior of individuals.
data analysis; clustering; migrations; animal ecology
Settore INF/01 - Informatica
2016
29-lug-2015
Article (author)
File in questo prodotto:
File Dimensione Formato  
mainijgis-with-authors-pubblico.pdf

accesso riservato

Tipologia: Publisher's version/PDF
Dimensione 2.57 MB
Formato Adobe PDF
2.57 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/292983
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 24
  • ???jsp.display-item.citation.isi??? 23
social impact