Long time series are important because they extend back to an era when animal populations were less influenced by habitat loss and climate change. Annual fluctuations in harvest yields are good proxies for large changes in population size and may reveal underlying ecological processes. From a variety of sources, we built a 142-year long time series representing the mean daily catch (CPUE) of willow ptarmigan Lagopus lagopus in southeastern Norway. CPUE decreased over the 142 years, from 35 birds shot per day in early years to around two in the last years. There were three periods in the time series: a first period with 3-5 year cycles of high peaks and low troughs (1872-1900), a short second period with similarly high peaks, variable depth of troughs and variable cyclicity (1901-1916), and a third long period with much lower peaks and faded cycles (1917-2013). Yearly variation in CPUE was best explained by an interacting effect of small rodent peak years and period, with a reduced positive effect of rodents in the last period, and a positive effect of the North Atlantic Oscillation index in spring and early summer. None of the weather variables with significant time trends explained any variation in CPUE and we could therefore not attribute the decline in CPUE directly to climate change. We postulate that a long-term dampening of the amplitude in small rodent cycles combined with an increase in red fox numbers, have increased predation on alternative prey like ptarmigan, and prevented the populations from reaching their earlier peaks. Even though the present population of willow ptarmigan is only a fraction of what it was in former days, we recommend light hunting to motivate for monitoring and to keep public attention on the bird.  

Willow ptarmigan harvest data (CPUE)

Introduced by British sportsmen in the first half the 1800`s, hunting of willow ptarmigan by pointing dogs became a popular pastime of the Norwegian upper-class (Søilen 1995). This has resulted in continuous written documentation of harvest, but the sources and type of information vary throughout the study. We have used three types of data and provide the annual estimates and sources in Table S1 in the electronic supplementary material.

Years 1872-1939. From 1872, hunters were organized in the Norwegian Hunting and Fishing Association (NJFF) and soon after a monthly journal appeared. Already in the first issues of the journal, the editorial staff asked hunters, each fall, to report the number of ptarmigan shot and the number of days hunted, or just the average number of birds bagged per day. These data were processed by employees of the Norwegian Forest Service from 1872 to 1881 (Barth 1891; Skinnemoen 1979), and by the editorial staff of the journal from 1882 to 1939, and presented in annual issues, and compose our measure of CPUE for the first part of the time series (Table S1).

The number of hunters reporting their catch varied between years, from 200 to 400 during the years 1913-1917 and 1926-1936, but with unknown number in other years. Furthermore, the information given in the journal varied between very detailed such as the exact number of birds shot per hunter, to less precise interval-type assessments. A typical example of the latter is the report for 1891, a very good year (Anon. 1891): “Most hunters have shot around 30 birds on short field days, sometimes the double and up to more than 100 birds on a single day, by those that wanted to make a record”. The next year, 1892, saw a complete breakdown of the population: “The bag was generally between 1 and 5 birds per hunter/day” (Anon. 1892). The very high peak years occurred only during the years 1872 to 1912 (Fig 1). The CPUE for these years was conservatively set to 35 willow ptarmigans shot per hunter day.

Years 1940 -1971. During 1932 - 1971, the Norwegian Wildlife Research Institute conducted a yearly inquiry on the sizes of willow ptarmigan populations. The game board of each municipality was asked to give an evaluation of population sizes compared to those of the last 5 – 6 years (Myrberget 1974, 1982b). Sizes of populations were rated using a quantitative index from 0.1 to 1.4 with 1 as the average population size (Fig. S1). These indexes have been converted to CPUE by a simple calculation: 1949 is the only year from which both the game boards estimates and a CPUE from the study area is available (Holt 1950). This year, the CPUE was 1,9 birds (average of 49 reports). The CPUE for any one year is estimated as: 1,9 x A/B, where A and B are from the index of the inquiry. A is the index of the year for which the CPUE is to be estimated and B the CPUE for 1949. These estimated CPUEs corresponds well with the CPUEs collected by NJFF for the overlapping years 1932 to 1939, and also for the final years, 1963 to 1971 (when yearly proportion of juveniles in southern Norwegian ptarmigan populations was determined from hunter bags as indicated by wing samples) (Myrberget 1974). Although the reports from the municipalities cover the whole country, we assume that they represent a good proxy for the CPUE from our study area for the period 1932-1971. Interestingly the game boards took a rather pessimistic view of population sizes. During one year only (1966) of the entire 40-year period, ptarmigan populations were regarded as above average (Fig. S1).

Years 1972-2013. From 1972 to present, Statistics Norway (www.ssb.no) has collected data on the number of ptarmigan shot each year and for later years also the number of hunters and days hunted. The CPUEs for the study area may be calculated directly from the numbers provided by Statistics Norway (Fig. S2), but they would not be directly comparable to the CPUE for the first part of the study. Ptarmigan hunters of later decades are a highly mixed group. Main differences consist of use of dog or not, level of experience, quality of terrain (including if it is private or public ground) and whether they hunt only the first week or the entire season. Those that reported their catch in earlier days were a much more homogenous group. They were generally wealthy people, very devoted to hunting, they used dogs which were possibly better trained than the average dog of today, and they could choose terrains with high population density and hunt only during the first part of the season when hunting conditions were the best. The present-day hunters that most resemble those of earlier centuries, are those that owns or rent private terrains with generally high ptarmigan populations. They use well-trained dogs and have a thorough knowledge of their terrain and where to find the birds. Considering this, we choose the number of bagged ptarmigan at private terrains in Dalsbygda, situated in the northeastern mountain regions of southern Norway, to compare present figures of CPUE with those of older days. This area generally has high density of willow ptarmigan. The best catches of willow ptarmigan in these terrains were in 1988, with an average of 6.8 birds reported shot/hunter/day (Pedersen 1995). This also corresponds with the best year recorded by Statistics Norway covering all of Norway (Fig. S2). To estimate the daily catch for each year during the period 1971 – 2013 we used the same approach as above: 6.8 x A/B, where A is the number of birds shot during the year for which the CPUE is to be estimated and B is the CPUE shot in 1988 (A and B from Statistics Norway).

The data file is stored as excel file and tab delimed text file. It contains all data needed to reproduce the analyses in the manuscript, including rodent data, North Atlantic Oscillation, and weather data, with the same names as used in the Wildlife Biology paper "The roles of climate and alternative prey in explaining 142 years of declining willow ptarmigan hunting yield" by Hjeljord and Loe.