11.07.2015 Views

BIRD POPULATIONS - Birdpop.org

BIRD POPULATIONS - Birdpop.org

BIRD POPULATIONS - Birdpop.org

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

ROBERT L. WILKERSON AND RODNEY B. SIEGELFIGURE 14. Probability of detecting owls during the 2006-2007 survey in blocks where owls were detected in1991-1993 as a function of the number of owl pairs detected on the block in 1991-1993. The predicted probabilityof detection in 2006-2007 is shown by the curve. Data points below the curve are blocks on which owls weredetected during both surveys; data points above the curve represent blocks where owls were detected in 1991-1993 but not detected in 2006-2007. Data points are plotted at their 1991-1993 owl pair (x-axis) values andrandomly jittered in the probability (y-axis) space (below or above the curve, depending on whether owls werepresent in 2006-2007) to show the distribution of the data.reliably detect all owls in sampled areas (Conwayand Simon 2003, Conway et al. 2008), particularlyin areas with limited or no road access may tendto bias our estimates low. Additionally, observersgenerally assumed that whenever they detected asingle adult Burrowing Owl, it represented abreeding pair. To the extent that unmated adultbirds may have been detected, this could result inan upward bias in our estimate of breeding pairs.Another potentially confounding factor was thatsurveyors were unable to gain access to somemilitary installations and private landholdings; ifsuch areas were more or less likely to be occupiedby owls than other areas, bias in one direction orthe other could have been introduced into ourestimates. Finally, our survey methodologyincorporated no means for assessing detectionprobability, which in some environments (such asdesert areas with very low road density) mayhave been quite low. Perhaps of even greaterconcern than detection probability being low isthat it could have varied substantially acrosssurvey blocks or survey regions with differentphysiographic characteristics.Nevertheless, we believe the sheer volume ofdata collected counterbalances some of themethodological limitations described above, andensures that the broader patterns in distributionand abundance are meaningful. Additionally,because our methods adhered to thoseestablished by DeSante et al. (2007), any biasesaffecting our results likely affected the 1991-1993study, too, so that comparisons between the twosurveys are appropriate. Finally, our surveydocumented the exact locations of 1,758Burrowing Owl pairs (18.9% of the estimatedtotal) across California, information that shouldbe of great use for ongoing and future conservationefforts.The generally large variances associated withour regional and statewide population estimatesextrapolated from randomly-selected blocksindicate that our statistical power to detectchanges in abundance was rather weak. Indeed,the Northern Central Valley was the only regionfor which our 2006-2007 population estimatediffered significantly from the 1991-1993 estimateof DeSante et al. (2007). Moreover, many of[30]

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!