Monitoring Svalbard rock ptarmigan: Distance sampling and occupancy modeling

• Published in: The Journal of Wildlife Management
• Main Authors: Pedersen, Åshild Ø., Bårdsen, Bård‐Jørgen, Yoccoz, Nigel G., Lecomte, Nicolas, Fuglei, Eva
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2011
• Subjects: Nature and Landscape Conservation; Ecology; Ecology, Evolution, Behavior and Systematics; Arctic; Svalbard; Norway; Survey Point; Lagopus muta; Lagopus muta hyperborea; rock ptarmigan; Svalbard Rock Ptarmigan
• Online Access: http://dx.doi.org/10.1002/jwmg.276; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fjwmg.276; http://onlinelibrary.wiley.com/wol1/doi/10.1002/jwmg.276/fullpdf

Abstract The only resident terrestrial herbivorous bird species in high‐Arctic Svalbard, Norway is the endemic Svalbard rock ptarmigan ( Lagopus muta hyperborea ) of which little is known of its population dynamics. We assessed temporal and spatial variability of the pre‐breeding population of Svalbard rock ptarmigan males using: 1) distance sampling to estimate density (2000–2009) and 2) occupancy modeling to determine the proportion of survey points being occupied in relation to a habitat index for ptarmigan habitat suitability (2005–2009). Data were collected using a point‐transect sampling design. We split the analysis according to type of survey point (non‐random, random, and survey points combined). Our estimated spring densities were low (1.3–3.1 territorial male/km 2 , non‐random survey points, 2000–2009) with limited annual variability. The best models describing occupancy rates of territorial males at 2 different spatial scales (ptarmigan males observed ≤250 m and ≤450 m from the sampling point) were independent of spatial scales and the type of survey points. Occupancy dynamics were related to the habitat index whereas detection probability was year dependent. Extinction probability was negatively related to habitat quality (good habitats had lower extinction probability). We could not estimate the habitat effect on colonization precisely because initial occupancy rates were high at both spatial scales (estimated average initial occupancy at scale ≤250 m = 0.96; scale ≤450 m = 0.97). Colonization appeared to be positively related to the habitat index for the random survey points (including mainly marginal habitats), but the small sample size led to large uncertainty in the parameter estimate. Detection probabilities varied greatly between study years, thus demonstrating the importance of estimating detection probability annually. We recommend that future surveys are stratified with respect to habitat quality and to integrate the 2 methodologies in population monitoring of Svalbard rock ptarmigan. © ...