The necropsy found that the female was reproductively active with developing follicles in the ovary, was emaciated, and heavily parasitized with large numbers of Leucocytozoon spp. protozoa in red blood cells and Elmeria spp., coccidia and Capillariid spp. in the intestine. The female weighed 490 g when found, which was 73.9% of the mean weight for female northern spotted owls [24]; no baseline weight had been taken for this individual when first captured in 2008. No other abnormalities, including trauma, were detected. The owl tested negative for avian influenza viruses, West Nile virus, and exposure to lead. Brain cholinesterase levels were not depressed, suggesting no acute exposure to cholinesterase-inhibiting toxicants, such as organophosphate or carbamate pesticides. Proximate cause of death was diagnosed as emaciation and parasitism.

Brodifacoum was the only rodenticide detected in the liver (33.3–36.3 ng/g) or the blood (0.48–0.54 ng/ml) (Table 1). In this case, exposure to brodifacoum was not the primary cause of death of the northern spotted owl examined here, as there was no sign of internal hemorrhage indicative of AR poisoning. However, the levels of brodifacoum residues found in the owl’s liver have been associated with lethal AR poisoning in other owl species, such as great horned owls (Bubo virginianus) with liver residues as low as 10 ng/g [25]. Two northern spotted owls submitted to the NWHC in the 1990s had brodifacoum liver residues of 50.0 and 100 ng/g and signs of hemorrhaging (NWHC Case Numbers 10128 and 13799). Although not directly linked as the cause of mortality, exposure to brodifacoum in the owl we found may have resulted in a sub-lethal exposure. While sub-lethal effects of brodifacoum on non-target wildlife are poorly understood, they are hypothesized to include anemic lethargy that impairs hunting ability leading to loss of body mass, and increased susceptibility to disease [26, 27]. For example, sub-lethal exposures to brodifacoum slowed growth in Japanese quail [28] and was associated with an outbreak of notoedric mange in bobcats (Lynx rufus), which led to a 64% reduction in survival [29]. Combined with the northern spotted owl’s reproductive status and heavy parasitism, brodifacoum may have been an additional contributor to the owl’s death.

Brodifacoum is commonly used in household, industrial and agricultural settings and is often found at illegal clandestine MGOs [5–8] to prevent rodents from damaging the stalks of marijuana plants where it is applied at plant bases and around MGO perimeters [6]. The dead northern spotted owl was found within 669–1347 m of at least 7 active MGOs on private inholdings within a National Forest (Fig. 1). In 2015, an illegal clandestine MGO with ~ 23 kg of brodifacoum-laced bait was discovered 450 m from the recovery location of this dead owl (M. Gabriel, unpublished data, Fig. 1), indicating that other undetected MGOs may have been nearby. Given the documented use of brodifacoum on MGOs, it is highly likely that the source of the brodifacoum residues found in the dead owl were from one or more of the MGOs within its territory (Fig. 1); we were unable to identify any other potential sources.

One probable mechanism of AR transmission from MGOs to northern spotted owls in this region of California is through ingestion of dusky-footed woodrats, which are a dominant prey of spotted owls in this area [30]. Dusky-footed woodrats are abundant in early-seral stages, such as openings created by fire, timber harvests, or (presumably) MGOs, and have specialized gut microbiomes that allow them to digest toxic secondary plant compounds and fibrous plant material [31], typical of marijuana plants. In addition, dusky-footed woodrats incorporate plants with high monoterpene content [32, 33], such as California bay (Umbellularia californica), into their nests, with evidence that these plants act as larvicides and repellants against fleas [34, 35]. Cannabis contains similar monoterpenes and can also act as a larvicide against mosquito and other insect larvae [36–38]; anecdotal observations indicate that woodrats incorporate marijuana stalks into their nests [13]. Both California bay and marijuana plants are aromatic [34, 39] and woodrats may be able to detect these on the landscape through olfactory cues. Northern spotted owls in California also tend to forage near edges of openings when woodrats predominate in their diet [40]. In forested landscapes in Humboldt County, California, MGOs have generated increased edge with forest areas and increased patch shape complexity [41], landscape elements that also contribute to high-quality habitat for northern spotted owls [30].

Based on this, we propose a mechanistic hypothesis on the linkage between MGOs and secondary AR poisoning of northern spotted owls (Fig. 2), where woodrats are attracted to MGOs by the presence of marijuana plants for food and larvicidal nest material and encounter ARs in rodent baits while foraging on MGOs. Northern spotted owls are also attracted to MGOs because of the edge habitat created by MGOs and increased prey movement across those edges. This edge becomes the area where owls then prey on AR-contaminated woodrats.