Given that the vast majority of participants began drinking during teenage years and must therefore recall multiple decades of alcohol use, estimates of alcohol consumption will naturally deviate from the true amount of alcohol exposure. Consequently, it is recommended that our estimates related to alcohol use and neurocognition be interpreted conservatively with a greater emphasis on directionality than exact magnitude. The cross-sectional nature of our data prevents us from disentangling the effects of alcohol and MA use from longstanding individual differences in neurocognitive capacities . However, the inclusion of the WRAT Reading subtest as a covariate in all regression models increases our confidence that the observed effects of substance use on neurocognitive performance are not attributable to premorbid functioning. Furthermore, the application of demographic corrections to neuropsychological test scores improves the comparison of results between the MA groups despite differences in education. The positive association between WRAT scores and global neurocognition highlights the incremental predictive value of the WRAT above and beyond demographic effects, most notably education. These findings align with prior substance use studies that suggest that intellectual enrichment, as indicated by high IQ, can increase cognitive reserve and mitigate the deleterious effects of stimulant-induced neural injury on neurocognition . Unsurprisingly, MAþ individuals more frequently met criteria for lifetime dependence for other substance use than MA– individuals. However, study exclusion criteria necessitated that such dependence be episodic in nature and remote . Additional individual differences that we were not able to capture in the present study include potential genetic differences in vulnerability to alcohol effects ; however, these would presumably be equally distributed among MAþ and MA– individuals.
The unexpected finding that alcohol reduces the likelihood of neurocognitive impairment in MAþ individuals raises intriguing biologically driven theories of neuroprotection that we unfortunately cannot answer with our data. Simultaneous administration of MA and alcohol versus non-overlapping periods of single substance use is an issue central to conceptualizing the interaction between MA and alcohol use. Many primary MA users report alternating use of MA and alcohol throughout a given binge in order to titrate their subjective experience of intoxication . This coordinated pattern of MA and alcohol use may attenuate MA-related sleep disturbances,indoor hydroponic system but may also increase risky behaviors due to decreased perceptions of intoxication . Although the lifetime average daily alcohol use metric captures lifetime alcohol patterns, it does not capture chronicity and persistence of alcohol use nor does it distinguish periods of concurrent MA and alcohol use from intervals of monosubstance use among the MAþ individuals. Such a distinction between lifetime periods of simultaneous intoxication versus non-overlapping intoxication would permit for a more nuanced understanding of the aforementioned neurophysiological hypotheses. Additionally, although our neurocognitive variables reflect the behavioral outputs of neural functioning, they do not directly measure the integrity of neural circuitry and neurobiological activity. Therefore, the inclusion of genetic, neuroimaging, and fluid-based biomarker data that more directly reflect neurobiological pathways is recommended for future studies of polysubstance use.As of January 2020, recreational use of cannabis is legal in Uruguay, Canada and 12 US states, and medical use is partially or fully legal in 36 countries . As legal markets for cannabis develop, policy makers are tasked to regulate its production, distribution and consumption in new ways. With rising liberalization, researchers have taken a growing interest in the potential environmental impacts of cannabis – a dynamic partly fueled by growing public concerns and news coverage of the topic, which increased by over 500% from 1992 to 2019 . If implemented successfully, legalization could give regulators a chance to anticipate and regulate the environmental outcomes of the cannabis industry as it expands.
Some current regulatory schemes already reflect this priority through the inclusion of specific language meant to reduce environmental impacts which can arise from land, water and energy use, application of chemicals, or other pathways . There are four primary classes of cannabis production which may impact the environment through different pathways and at different magnitudes . These production systems are not always clearly distinct in practice: for instance, in a single farm, mother plants may be kept indoors while cloning occurs in mixed-light and full crops are produced outdoors. Aside from trespass systems , which we describe separately due to the specific practices associated with them, the cannabis production systems we describe can exist legally or illegally. There are distinct trade-offs between production systems. Indoor systems are associated with few concerns about wildlife habitat destruction, water diversion or pollution, but require high external inputs such as energy and fertilizers. Conversely, outdoor farms may require fewer resource inputs, but poor management or siting could disrupt surrounding ecosystems.We note that trespass grows are generally only associated with negative environmental impacts. Researchers investigating interactions between cannabis and the environment have faced historic hurdles – often due to cannabis’ legal status – which include societal stigma, funding restrictions, safety concerns and difficult access related to remote cultivation sites, as well as regulatory obstacles such as complex licensing requirements and restrictions on cultivar testing . Despite such limitations, a new science around cannabis and the environment is starting to emerge. Our objective here is to review existing literature documenting environmental impacts of cannabis, to identify significant research findings and knowledge gaps and to suggest policy recommendations. As shown in Fig. 3, before 2012 only a handful of studies suggested links between cannabis and environmental degradation . Recent empirical studies, however, have started to quantify specific environmental impacts of cannabis cultivation and consumption. While limited in size and scope, this first generation of studies provides an opportunity to identify and summarize both what is known about cannabis and the environment, and what knowledge gaps persist. This review highlights the emerging science around cannabis and the environment. We hope it can serve as a catalyst to encourage more research in this area and as a resource to provide science-based guidance for policy-makers.
We evaluated peer-reviewed and non-peer-reviewed sources that quantified the effects of cannabis cultivation or consumption on the environment. We excluded studies and reports that: addressed other impacts of cannabis such as on human health; focused on other plants or other illicit drugs; or commented on environmental impacts without providing data. Based on published commentaries on cannabis and the environment , we identified a list of terms to search the Web of Science for relevant studies in June-July 2019 . We screened titles and abstracts of resulting studies according to the three eligibility criteria noted above, yielding a total of 14 peer-reviewed articles for which we reviewed the full text. We incorporated nine additional studies referenced in these studies in our final review . We also searched for non-peer-reviewed literature on Google in July-August 2019 and included documents found in the first five pages of results. Our final review includes two non-peer-reviewed reports and a book series . We found six peer-reviewed studies that investigated the water footprint of cannabis cultivation , all of which focus on northern California. Bauer, et al. used satellite imagery to estimate the number of cannabis plants in northern California and used this to predict that watershed-scale water consumption may exceed local stream flow during the growing season. These results were based on assumptions that: on average, a cannabis plant consumes 22.7 liters of water per day throughout the growing season; this water is predominantly accessed through surface-water diversions; and water application equals water extraction. The authors suggested that during dry years, cannabis farming could completely dewater some streams. Butsic and Brenner applied a similar methodology to estimate annual water use for cannabis irrigation at 11,000 m3 – equivalent to 0.001% of annual agricultural water use – in Humboldt County, California.These findings highlight the potential impacts of cannabis on water resources, but their accuracy is limited by a lack of actual water use data. Three additional studies in California examined cultivator-reported water use for cannabis at the farm scale. High variability in water use and extraction practices was documented – likely driven by variation in seasonal growing patterns, farm size or cultivation methods. Wilson, et al. and Dillis, et al. both confirmed that water use rates among California cannabis farmers approximated the 6 gallon per-plant figure reported by Bauer, et al. . However, this was only the case during peak growing season and respondents reported lower water use rates throughout the rest of the year. Wilson, et al. also documented monthly water use on average-sized farms in California and found that while water application to cannabis plants exceeded this rate during cannabis’ growing season,microgreen flood table water extraction from rainwater, surface and sub-surface sources remained far below it for most of the year. In separate assessments of farm scale water extraction practices, Wilson, et al. and Dillis, et al. showed that sub-surface wells, rather than surface-water diversions, may be the primary source of water for many northern Californian growers. Sub-surface water extraction may threaten connected watersheds if annual extraction exceeds recharge rates, as sub-surface water reserves tend to recover more slowly from overuse than surface sources. We found one peer-reviewed study and one gray literature report focused on cannabis and energy use. Mills estimated that indoor US cannabis production uses 20 TWh of electricity annually, leading to the annual emission of 15,000,000 tons of CO2. This value is equivalent to the energy consumption of the entire US agricultural sector , or to 1% of US total national electricity use.
Mills’ calculations were based on national cannabis cultivation estimates and assumed “typical” energy use for indoor production and relevant transportation processes. A more recent report combined estimated US cannabis demand and cultivation area with self-reported data from cultivators to provide a detailed assessment of current cannabis energy use. Combined illicit and legal cultivation were estimated to consume 4.1 MWh annually, equivalent to 472,000 tons of associated CO2 emissions. These estimates did not account for off-grid energy use, transportation, fertilization or irrigation, but were significantly lower than the numbers reported by Mills . We note that Mills’ findings may not accurately represented energy use by the US cannabis sector today, as cultivation practices have likely become more efficient in recent years. Studies quantifying land-use impacts of cannabis remain scarce despite reports of significant cannabis cultivation activity in North and Sub-Saharan Africa, the Americas and Asia . We found five empirical studies from the US which assessed cannabis and land-use dynamics. Satellite data for California showed a high concentration of cultivation sites in remote, ecologically sensitive areas . In Humboldt County, cannabis’ impact on land cover change from 2000 to 2013 was relatively limited, contributing 1.1% of forest canopy area loss compared to 53.3% from timber harvest . However, remote cultivation sites were linked to landscape perforation as they created gaps in forest patches, reducing forest core areas and increasing open edges. This could contribute to landscape-wide forest fragmentation and resulting wildlife habitat degradation if current expansion rates persist . The spatial distribution of cannabis farms, in addition to total land-use footprint, may thus be significant determinant of potential environmental impacts. These reported spatial dynamics suggest that the factors driving the location of both legal and illegal cannabis cultivation are distinct from those of other crops. Cannabis prices and law enforcement related risks emerged as important factors determining siting decisions in California, Oregon and Washington’s illicit markets . Butsic, et al. documented strong network effects amongst growers in Humboldt County, which led to clustering of cultivation sites and appeared to be more important than biophysical factors such as soil quality or terrain. Klassen and Anthony identified state enforcement capacities and poverty and unemployment rates as potential factors leading to a decline in illegal farms discovered in Oregon, but not Washington, following legalization in both states. Although pesticides used in cannabis production are likely to impact the environment, to our knowledge no quantitative studies have documented these impacts on private land or legal cannabis production systems. We found five peer-reviewed studies which focused on impacts of anticoagulant rodenticides on local wildlife species in trespass grows. ARs are presumably used to control rodent populations; they are frequently encountered on trespass production sites in California and can bio-accumulate in the food chain . In northern and central California, field-studies documented contamination by highly toxic ARs in an endangered predator, the Pacific fisher , using a combination of field-data collection, lab data analysis and spatial correlation .