As mentioned above, wild or cultivated Cannabis plants have not been reported in Montcortès local and regional vegetation surveys (Carreras et al., 2005-2006; Mercadé et al., 2013), and hemp retting has not been practiced in the lake since the late 19^th century, as indicated by historical records and supported by the absence of bRNA (Rull et al., 2022). Therefore, a reasonable explanation for the 20C event seems to be long-distance pollen dispersal mediated by wind (anemophily). However, the possibility of local or regional illegal crops cannot be disregarded (Munuera et al., 2002). In both cases, a conspicuous increase in these activities would be needed to explain the 20C. Therefore, pollen dispersal patterns and historical trends of Cannabis cultivation are critical points to be addressed.

The high productivity and dispersal power of anemophillous Cannabis pollen are well known (e.g., Small & Antle, 2003; Li et al., 2017). On the IP, several studies have documented relatively high amounts of airborne Cannabis pollen in areas far from known plant sources. For example, a northern African origin for the airborne Cannabis pollen found in some localities from the southeastern IP has been proposed by several authors after detailed aerobiological studies including the use of long-distance dispersal modeling tools. These localities include the cities of Murcia (Munuera et al., 2002; Aznar et al., 2022), Málaga (Cabezudo et al., 1997; Aboulaich et al., 2013) and Córdoba (Cariñanos et al., 2004) (Fig. 1), which are situated 150-300 km north of the African coasts. Pollen from northern Africa comes primarily from illegal crops (Aboulaich et al., 2013). The possibility of nearest sources, such as illegal local/regional crops, was suggested by Munuera et al. (2002). Recently, Aznar et al. (2022) found a significant increase in airborne Cannabis pollen between 2017 and 2020 in the region of Murcia (Fig. 1), which coincided with the predominance of local/regional wind patterns, and attributed this increase to the spread of illegal crops in the region rather than to long-distance dispersal from Africa. The occurrence of high Cannabis pollen percentages in modern sediments near crop fields of this plant has recently been documented (Boutahar et al., 2023). A typical feature of these crops is their high mobility and spatial turnover, which is a strategy to avoid detection and makes their identification difficult (Aboulaich et al., 2013). These studies come from the southernmost part of the IP, and no comparable information exists for other regions. However, it is expected that the dispersal of Cannabis pollen follows similar spatial patterns. Similar surveys should be conducted in the Montcortès region to evaluate the potential contribution of long-distance vs. local/regional pollen sources. These studies could also contribute to the detection of illegal cannabis crops in the area.

Regarding production, hemp cultivation almost disappeared from Spain in the early 1970s, but a renewed interest in this crop began in 1972. Until the 1960s, hemp was cultivated mainly on irrigated semiarid areas from the eastern IP (Alicante) to provide thread, ropes and fabric (Fig. 1). In 1972, however, the paper pulp industry started using hemp as a raw material instead of worn textiles, and in the early 1980s, the main production center shifted from the east to the northeast (Catalonia) in lowland areas (400-900 m elevation) of higher precipitation (600-700 mm per year), thus avoiding the need for continuous irrigation (Gorchs & Lloveras, 2003; Gorchs et al., 2017). This geographical shift coincided with a considerable production increase (<1000 tons in 1980 to >8000 tons in 1998) favored by the onset of EU subsidies to hemp cultivation, which started in 1985 (40 € ha^-1) and significantly increased (∼780 € ha^-1) in 1996-97 (Fig. 4). This production increase was centered on the southern Pyrenees, where Lake Montcortès (slightly above 1000 m elevation) lies, under wet climates, as a spring crop in rotation with winter wheat. This rotation significantly increased wheat yields, which has been attributed to the improvement of soil structure and the contribution of hemp to control pests, diseases and particularly weeds, as its rapid early growth outcompetes most weed species, thus reducing the weed burden for the ensuing winter crops (Gorchs et al., 2017).

The 20C event, as recorded in Lake Montcortès was coeval with this 1980s hemp production acceleration, which strongly suggests a causal relationship (Fig. 4). Due to its high dispersion power, hemp pollen could have easily reached Lake Montcortès from adjacent regional pre-Pyrenean hemp crops, which could be confirmed by modeling the Cannabis pollen dispersal around the lake area. Hemp cultivation declined again shortly after the 1990s and was abandoned by 2006, due to the relocation of the hemp straw processing industry (Gorchs et al., 2017). This decline was also paralleled by pollen trends (Fig. 2), thus reinforcing the potential causal relationship. However, Cannabis pollen remained close to the retting threshold until at least 2015 (Figs. 2 and 3), which deserves an additional explanation. Additional hemp pollen sources from illegal crops cannot be disregarded, although the available evidence is less robust. In recent decades, a relevant rise in large-scale Cannabis plantations has been recorded in the IP, especially along the Mediterranean coasts, in a trend to increasingly replace hashish imports from northern Africa (Alvarez et al., 2016; Belackova et al., 2016). The potential manifestation of this peninsula-wide rising trend in the Montcortès region remains to be documented, and is worth to be addressed in future studies.

The geographical extent of the 20C event remains unknown, as it has not been recorded in other IP localities. This lack of evidence could suggest that this event could be a local/regional phenomenon, but methodological constraints should not be dismissed. On the one hand, only 23 of the 56 records included in the peninsular-wide CHIP database (Rull et al., 2023) contain the 20^th century, and of them, Cannabis pollen occurs in only four (Fig. 1), excluding Lake Montcortès. Of these four sites, Cannabis pollen is abundant (maximum of 25%) in only one (ET), whereas in the others this pollen has maxima of only 5% (CE) or shows single scattered occurrences (AD, EN). On the other hand, the resolution of these records, usually centennial to millennial, is not enough to resolve eventual changes that occurred within the 20^th century. For example, in ET, which is the most similar to Montcortès regarding pollen abundance, the record ends in 1991, and only a couple of samples are available for the whole 20^th century (Riera et al., 2004). Therefore, the lack of high-resolution records including the 20^th century is a major handicap to assessing the spatial patterns of the 20C across the IP.

Another interesting observation is that the major vegetation shift documented above and the 20C occurred at or shortly after the onset of the “Anthropocene”, as currently defined by the Anthropocene Working Group (AWG) (Waters et al., 2023). Indeed, the AWG proposes that this boundary should be situated in the mid-20^th century, coinciding with the Great Acceleration, when the Earth System underwent a global transformation overcoming the range of variability of the Holocene (Steffen et al., 2015). Our results, together with the very high resolution (seasonal) of the Montcortès record, suggest that the sediments of this lake could be a suitable candidate to identify and precisely date an eventual “Anthropocene” GSSP (Global Stratotype Section and Point) (Waters et al., 2018). The proxies identified by the AWG as the best suited stratigraphic markers (radioisotopes, fly ash, stable carbon and nitrogen isotopes) have not yet been analyzed in the Montcortès record, but efforts in this direction are worth doing, especially after the detection of significant Hg and Pb increases during the Industrial Revolution and the Great Acceleration in these sediments (Corella et al., 2017).