Range-wide sources of variation in reproductive rates of northern spotted owls

Abstract

We conducted a range-wide investigation of the dynamics of site-level reproductive rate of northern spotted owls using survey data from 11 study areas across the subspecies geographic range collected during 1993-2018. Our analytical approach accounted for imperfect detection of owl pairs and misclassification of successful reproduction (i.e., at least one young fledged) and contributed further insights into northern spotted owl population ecology and dynamics. Both nondetection and state misclassification were important, especially because factors affecting these sources of error also affected focal ecological parameters. Annual probabilities of site occupancy were greatest at sites with successful reproduction in the previous year and lowest for sites not occupied by a pair in the previous year. Site-specific occupancy transition probabilities declined over time and were negatively affected by barred owl presence. Overall, the site-specific probability of successful reproduction showed substantial year-to-year fluctuations and was similar for occupied sites that did or did not experience successful reproduction the previous year. Site-specific probabilities for successful reproduction were very small for sites that were unoccupied the previous year. Barred owl presence negatively affected the probability of successful reproduction by northern spotted owls in Washington and California, as predicted, but the effect in Oregon was mixed. The proportions of sites occupied by northern spotted owl pairs showed steep, near-monotonic declines over the study period, with all study areas showing the lowest observed levels of occupancy to date. If trends continue it is likely that northern spotted owls will become extirpated throughout large portions of their range in the coming decades.

Keywords: barred owl; detection; misclassification; multistate; nondetection; occupancy; reproductive rate; spotted owl.

FIGURE 1

Location of 11 study areas used in the northern spotted owl reproductive rate analysis. Light green shaded area denotes the Northwest Forest Plan footprint.

FIGURE 2

The probability of a site being unoccupied by a pair of northern spotted owls $\left({\varphi }_{i,t}^{\left[0\right]}\right)$ on 11 study areas in Washington (a), Oregon (b), and California (c), USA, 1993–2018. Error bars are one standard error. CAS, South Cascades; CLE, Cle Elum; COA, Coast Range; GDR, Green Diamond; HJA, HJ Andrews; HUP, Hoopa; KLA, Klamath; NWC, NW California; OLY, Olympic; RAI, Rainier; TYE, Tyee.

FIGURE 3

The probability of a site being occupied by either an unproductive pair $\left(\text{orange},,,{\varphi }_{i,t}^{\left[1\right]}\right)$ or productive pair $\left(\text{blue},,,{\varphi }_{i,t}^{\left[2\right]}\right)$ of northern spotted owls over time in 11 study areas in Washington, Oregon, and California, USA, 1993–2018. Error bars are one standard error. CAS, South Cascades; CLE, Cle Elum; COA, coast range; GDR, Green Diamond; HJA, HJ Andrews; HUP, Hoopa; KLA, Klamath; NWC, NW California; OLY, Olympic; RAI, Rainier; TYE, Tyee.

FIGURE 4

Probability that a pair of northern spotted owls successfully reproduced $\left(R_{i,t},=,{\varphi }_{i,t}^{\left[2\right]},/,\left({\varphi }_{i,t}^{\left[1\right]}||+{\varphi }_{i,t}^{\left[2\right]}\right)\right)$ in 11 study areas in Washington (a), Oregon (b), and California (c), USA, 1993–2018. CAS, South Cascades; CLE, Cle Elum; COA, Coast Range; GDR, Green Diamond; HJA, HJ Andrews; HUP, Hoopa; KLA, Klamath; NWC, NW California; OLY, Olympic; RAI, Rainier; TYE, Tyee.