Monitoring anthropogenic activity in the Hardangervidda wild reindeer range. Possible applications of crowdsourced Strava-data in remote settings

Abstract

Seen in light of the increasing interest of nature-based tourism and recreational outdoor activities in Norway the last decades (Haukeland, Grue & Veisten, 2010), spatiotemporal information on human activity in remote areas and knowledge about how this activity may affect wildlife and nature is a crucial part of a knowledge-based management (Gundersen et.al., 2011, p.14; Gundersen, Strand & Punsvik, 2016, p.166). Hardangervidda is the largest national park in mainland Norway and is also home to the largest population of wild mountain reindeer (Rangifer tarandus tarandus), a specie of international responsibility in management and conservation and recently added to the Norwegian red list (Kjørstad et.al., 2017, p.26; Artsdatabanken, 2021). In this thesis, I calibrated crowdsourced data from the Strava-app with visitor data from automatic counters from 2016-2019, to identify spatiotemporal activity patterns in Hardangervidda area (inside and outside the national park) and to evaluate the use of Strava as a methodological approach for monitoring visitor traffic in remote areas. The spatiotemporal patterns on human activity in summer in the period 2016-2019 derived from Strava-data was presented in combination with a habitat suitability map for wild mountain reindeer and GPS-positions from GPS-collared wild reindeer during the same period, to discuss potential conflict-areas regarding human – wild reindeer interferences on Hardangervidda. Because there is a lack of knowledge about how automatic counters work in remote settings in wintertime, counter accuracy was tested by comparing measures between two automatic counters in the exact same location at specific trail-segments during winter 2018. My results showed that the relationship between Strava-data and data from automatic counters was consistent during summer season (July-Sept.), while data from winter season (March and April) showed no consistent pattern between the two measures. The relationship between the two measures of human activity was assumed to be related to the number of observations from Strava, as there were considerably more observations during summer than winter. The results showed that the consistency between the counter-pairs across all locations was ρc = 0.82 and r > 0.7 for 15 of the counter-pairs, indicating high counting accuracy for automatic counters also in remote settings in wintertime. Observations of human activity from Strava showed more observations outside the national park than inside the national park, in all years and all months. The seasonal variation in Strava-observations showed a distinct peak during summer (July-August), and lower numbers for the remaining months. There was an increase in Strava-observations over the study years both inside and outside the national park, however, this increase is assumed to be related to an increased use of the Strava-app and cannot be concluded to be evidence for increased human activity on Hardangervidda. The main emphasis of the Strava-data was in the surrounding edge zone and in relation to the marked trails and public tourist cabins in the Hardangervidda area. Using Strava-data as a single monitoring method is not recommended, as there are challenges concerning representability and reliability in the dataset. This is especially evident for the off-trail recreational use, were my results suggest that this activity is rarely detected by Strava. However, Strava-data does provide large amounts of spatiotemporal data on human activity, and my results complement results from previous studies that Strava-data has potential as a valuable supplement to describe the main use pattern in an area and can increase data from other monitoring systems. The marked trails, especially the coherent routes south-north and east-west may function as a barrier for the wild reindeer population, preventing them to use suitable summer-pastures in the northern and western part of the range, with consequence of very high density of reindeer in a limited area in the south-east. An increased level of visitors on Hardangervidda with higher frequency of human activity on the marked trails may lead to long-term disturbance effects for the wild reindeer population, like habitat alteration, avoidance of large areas of suitable habitats and potential cascade-effects like increased grazing-pressure in certain areas (Andersen & Hustad, 2004, p.40; Gundersen & Singsaas, 2021). Further studies on visitors’ impact on the wild reindeer population on Hardangervidda should be carried out to achieve a sustainable human use of this area and to secure the wild reindeer populations’ livelihoods and sustainable population dynamics.

Keywords: wild reindeer, Hardangervidda, Strava, automatic counters, visitor data, anthropogenic disturbance effects