257 Folia Zool. – 55(3): 257–263 (2006) Evaluation of sign surveys and trappability of American mink: management consequences Iñigo ZUBEROGOITIA 1 , Jabi ZABALA 2 and Jose A. MARTÍNEZ 3 ¹ Icarus E. M. S. L. Icarus, C/ Pintor Sorolla 6. 1º, E-26007 Logroño, Spain; e-mail: Zuberogoitia@icarus.es ² Sebero Otxoa 45, 5 B, 48480 Arrigorriaga, Biscay, Basque Country, Spain; e-mail: jzabalaalbizua@yahoo.com ³ C/ Juan de la Cierva 43 (S.T. 100), El Campello, 03560 Alicante, Spain; e-mail: qvcocotiers@hotmail.com Received 22 September 2005; Accepted 7 September 2006 A b s t r a c t . We compared different methods of monitoring an American mink, Mustela vison, population in an European mink, Mustela lutreola, area. The study was carried out in the Butron basin, northern Spain. We compared sign surveys and trapping in the same non-continuous 500 m sections of river. We also radio-tagged 10 minks to evaluate the efficiency of the two methods. There was no significant association between the presence of tracks and trapping success. We noticed the presence of minks in sections where none was captured and vice versa. Furthermore, by monitoring the radio-collared minks, we found inter-sexual differences in spatial use that produced differences in the success of the trapping and sign surveys, with males using the main streams and females selecting little streams where tracks were difficult to survey. Moreover, we detected tracks in the sand of 17 minks that actively avoided the traps. Hence, 1) neither sign surveys nor trapping are reliable methods for estimating relative abundances because both are subjected to strong biases. 2) the spatial use of habitat by minks could severely affect management projects that consider only main streams and 3) projects aimed at extracting American minks need to consider those individuals that never trigger traps, females living in very small streams and the floating population living in other habitats. Key words: American mink, field surveys, monitoring programs, relative abundance, trapping effort Introduction Estimates of animal abundance are among the most important requirements of wildlife managers and researchers. Developing methods for collecting distributional data is essential for several purposes: knowledge on geographical distribution, habitat-relation models, effects of land-use changes, effects of human density and disturbance on distribution, relationship between species occurrence and landscape physiognomy and composition, viability models, population-monitoring programs, which ultimately determine the convenience of protecting a species (Z i e l i n s k i & K u c e r a 1996). Estimating the size of wildlife populations can be problematic, especially in mustelids as they are often elusive, nocturnal and may have large home ranges. The ideal method for obtaining reliable results is the capture-recapture method, although trapping carnivores at the intensity required to produce such estimates is not always possible, being difficult, labour-intensive and prohibitively expensive (W i l s o n & D e l a h a y 2001). Therefore, most estimates of relative abundance, expressed as an index value, are often obtained employing alternative methodologies based on detection of field signs (F o r e s m a n & P e a r s o n 1998, S a r g e a n t et al. 1998, W i l s o n & D e l a h a y 2001). When a species is classed as a pest, or, as an alien species is regarded as a threat to endangered indigenous species, the management objective may be to reduce the population