Diseases and parasites are important drivers of population dynamics in wild mammal populations. Small and endangered populations that overlap with larger, reservoir populations are particularly vulnerable to diseases and parasites, especially in ecosystems highly influenced by climate change. Sarcoptic mange, caused by a parasitic mite (Sarcoptes scabiei), constitutes a severe threat to many wildlife populations and is today considered a panzootic. The Scandinavian arctic fox (Vulpes lagopus) is endangered with a fragmented distribution and is threatened by e.g., red fox (Vulpes vulpes) expansion, prey scarcity and inbreeding depression. Moreover, one of the subpopulations in Scandinavia has suffered from repeated outbreaks of sarcoptic mange during the past decade, most likely spread by red foxes. This was first documented in 2013 and then again 2014, 2017, 2019, 2020 and 2021. We used field inventories and wildlife cameras to follow the development of sarcoptic mange outbreaks in this arctic fox subpopulation with specific focus on disease transmission and consequences for reproductive output. In 2013-14, we documented visual symptoms of sarcoptic mange in about 30% of the total population. Despite medical treatment, we demonstrate demographic consequences where the number of arctic fox litters plateaued and litter size was reduced after the introduction of S. scaibei. Furthermore, we found indications that mange likely was transmitted by a few arctic foxes travelling between several dens, i.e., “super-spreaders”. This study highlights sarcoptic mange as a severe threat to small populations and can put the persistence of the entire Scandinavian arctic fox population at risk.

Study population

The sarcoptic mange epizootic occurred in the Swedish subpopulation Borgafjäll (65.05N 14.33 E) located at the border between Jämtland and Västerbotten counties in Sweden (Fig. 1a). The area consists of 1676 km² sub-Arctic tundra and is surrounded by valleys with boreal forest. The arctic fox population in Borgafjäll has been under conservation actions since 2001, including supplementary feeding and red fox culling (Angerbjörn et al. 2013). Feeding stations containing commercial dog pellets are placed in close connection to active den sites. Since the conservation actions started, the number of litters has increased two-fold from early in the century until today (Wallén et al. 2021).

Disease appearance and monitoring

 The first suspected case of sarcoptic mange was observed the 14^th of April 2013. This individual was utilised and sent to the Swedish Veterinary Institute (SVA) for autopsy which confirmed the diagnosis of sarcoptic mange (U130513-0007). After the discovery of this first individual, we put out wildlife cameras on inhabited den sites (n=22) and combined this with intensive field monitoring in order to estimate the extent of the infection. During 2013 - 2015, we followed the population in a rather detailed way. However, for the period between 2016 - 2021 we only conducted ordinary inventories of most known dens and had some wildlife cameras out. We have therefore divided the data in three parts; i.e., before the mange infestation 2004 - 2012, intensive monitoring 2013 - 2015, and an extensive monitoring during 2016 - 2021.

During all years the population was monitored by field inventories and the individuals at inhabited den sites were visually examined on distance using binoculars in order to detect symptoms of sarcoptic mange. Further the den sites were monitored for the occurrence of litters. The field inventories were combined with wildlife cameras at some dens in order to monitor the abundance of individuals with severe symptoms (Fig. 1b).

In December 2013, a mountain rescue group reported a suspected case of sarcoptic mange that was observed at an old hut. Due to severe weather conditions, field work could not be initiated directly but new cases of sarcoptic mange was confirmed in January 2014. In January, two weeks of trapping was conducted in order to collect blood samples and monitor symptoms of infected individuals. In total, 6 individuals could be captured and sampled. Blood samples were collected on Nobuto blood sampling paper and was analysed by SVA using sarcoptic ELISA test.

Wildlife cameras

For the intensive period we tried to identify individuals with severe symptoms of sarcoptic mange (Fig. 1b), and we had wildlife cameras on all inhabited dens. The cameras were set to be triggered by motion and took 3 pictures every third minute when triggered. The pictures were analysed for red or arctic foxes with or without symptoms of sarcoptic mange. All pictures were then classified into visits in order to be able to compare the occurrence of infected individuals independent of the duration of the visit. A visit was considered as unique if no pictures with animals had been triggered during the last 20 minutes. By this, an individual that e.g., spent 6 h in front of the camera contributed to only one visit, despite several hundreds of pictures. During this intensive monitoring cameras were active during three periods; May-June 2013, September-November 2013 and February-April 2014. Due to logistical restrictions and harsh climate, the cameras have not always been active for the same number of days, mainly dependent on the snow conditions. The minimum number of working days for a camera during one period were 20.

During fur moulting in May, we could estimate the minimum number of individuals by fur characters and thereby determine the number of individuals at specific den sites. In July, no individual identification was possible due lack of individual markers in the fur.

For the extensive monitoring 2016 - 2021, the population was monitored by the county board administration through wildlife cameras. This recorded arctic fox activity, reproduction, litter size and foxes with visual signs of sarcoptic mange infection. However, data on the detailed dynamics of transmission across den sites has not been recorded in this period.

Treatment regime

The treatment was provided non-invasively through meat or sausage injected with Dectomax or Nexgard with the active substance doramectin or afoxolaner. A minimum of five pieces were put out on each den site hidden in the entrances or feeding stations. The pieces were hidden in order to avoid consumption by birds. The treatment was started immediately after the disease was detected and was repeated at the infected inhabited den sites every second to third week until no signs of mange was recorded, repeated one to six times. All treatment was conducted by the county board administration in Västerbotten and Jämtland. Dens or feeding stations that were supplied with medicine were monitored through wildlife cameras and directly by field personnel.

Statistical analysis

We tested for difference in the proportion of foxes with sarcoptic mange observed at visits between years using the General Linear Model (GLM). The relationship between geographic distance from the core area (i.e., the area where mange was first detected) was correlated to the proportion of visits from foxes with clear symptoms of sarcoptic mange.

Based on the recorded number of litters over the study period, we calculated population growth rate, lambda = ln (Nt/Nt-1). Then, in order to decrease the sensitivity to rodent abundance, we used a three-year moving average. The difference in 3 years average growth rate during years without mange compared with years with manage was tested for with a t test. To investigate if the litter size changed in response to sarcoptic mange outbreaks, we used linear mixed models (LMM) fitted with maximum likelihoods, using litter size as response variable and period (before vs. after) as a binary fixed effect explanatory variable. We accounted for prey abundance by including small rodent phase (classified as ‘low’, ‘increase’ and ‘peak’) as a fixed effect explanatory variable. Den site was included as a random variable. Further, we also tested for a difference in litter size between years with documented sarcoptic mange (2013-14, 2017, 2019-21) and years with no records of mange (2015-16, 2018) within the later time period (2013-2021). For this, we used the same approach as above, but used years with and without documented sarcoptic mange as a binary fixed effect explanatory variable.