Fig 1.
Study area in Qinghai, China with an example of Argos locations of one antelope-year demarcating the actual migration route.
Winter and summer ranges are based on 50% fixed kernel density estimates of Argos locations. Four wintering sites are FR–Far range; MT–Mountain range; CR–Close range; and RV–River range.
Fig 2.
Net-square displacement (NSD) plot for each migration cycle of each individual Tibetan antelope (Pantholops hodgsonii) (individual ID_year).
The NSD plot represents the distance to the starting point (wintering sites) where Antelopes stay at the beginning of a calendar year. During winter (the beginning of the graphs) the distance remain low until they depart for their calving site in late spring. They arrive at calving site in June when they reach the largest distance from the wintering site. Calving periods are represented by the peak of the graphs (highlighted in red), after which antelopes initiate the returning migration and distance starts to decrease. All antelopes except for 35_2010 and 41_2011 returned to their original wintering sites and the distances drop back to around 0.
Fig 3.
Modelled migration corridor and Argos locations of Tibetan antelopes.
(A)-(D) Distance between Argos locations to modeled migration routes based on circuit theory and least-cost path, respectively, for each wintering site. (E) Distance to the underpass vs. Distance to circuit theory corridor. (F) Distance to the underpass vs. Distance to least-cost path.
Fig 4.
Extra distance that antelopes have to migrate in order to use the Wubei underpass.
X-axis represents ID and is grouped by wintering site. Distances estimated by LCP represents optimal route lengths. Actual migration distances when crossing the railway via the Wubei underpass are estimated by net-squared displacement (NSD) and by directly connecting Argos locations with Wubei as one known point for each individual (Argos). The prolonged distances are calculated by subtracting LCP from NSD or Argos. The full table can be found in S3 Table.