The quantitative analysis of 12 elements (Pb, Cu, Ca, Mg, Mn, Zn, Cd, As, Hg, Fe, Na, and K) in a total of 46 samples, of which 34 samples were Cannabis samples and 12 soil samples associated with the cannabis regions, was carried out using ICP-MS. The soil and cannabis farm samples all came from Nsawkaw, Badu, and Boti areas in Ghana. While the high concentration of eight of the analyzed elements (Cu, Ca, Mg, Mn, Zn, Fe, Na, K) is not concerning as they are essential for both the plant and the soil, four of the analyzed elements (Hg, Pb, As, Cd) are highly toxic metals and metalloids, which are all known to be carcinogenic, as well as to produce a number of serious adverse health effects. The distribution of the detected elements and the skewness are shown in Figure 1, Figure 2, Figure 3 and Figure 4, while the individual concentrations for each element for both cannabis and soil samples is illustrated in Tables S5–S7 (Supplementary). The levels of concentrations for the eight essential elements in cannabis ranged from non-detectable (N.D.) to 643 ppm for Na, 2830 to 12,939 ppm for Mg, 20,772 to 60,537 ppm for K, 10.699 to 45,652 Ca, 73 to 2363 ppm for Mn, 131 to 3335 ppm for Fe, 6 to 73 ppm for Cu and 20 to 197 ppm for Zn.

These values are comparable to values for cannabis leaves published in other studies. For example, a study by Zafeiraki et al. reported Cannabis samples from various regions in Greece with analogous concentrations: 7.1 to 19.8 thousand ppm for Mg, 24.0 to 123.2 thousand ppm for Ca, 76.8 to 518 ppm for Mn, 135 to 1338 ppm for Fe, 8.2 to 64.2 ppm for Cu and 23 to 157 ppm for Zn [23]. As in the study of Greece, the predominant elements from those analyzed in Ghana cannabis samples was Ca and K, followed by Mg, Mn and Fe. Of interest are the high levels recorded for Fe, which is because most of the cannabis farms are in high rocky areas with rocks containing high levels of the element. During the breakdown of the rocks, Fe is known to enter the soil. The farmers often choose these areas because they want to hide from security agencies. Ca was recorded as the highest metal/metalloid concentration in the cannabis samples for Nsawkaw, Badu, and Boti, with Boti recording the highest, which is consistent with data reported by Kuras and Wachowicz [24]. In addition, all the three farms showed the presence of Cu, Fe, Mg, Mn and Zn, as reported by Shibuya et al. [25], proving that Cannabis sativa can absorb metals and metalloids from the soil irrespective of where it was cultivated, but the levels may be dependent on natural and anthropogenic sources [26].

Moreover, the soils from the Cannabis cultivation regions were analyzed to associate their levels of trace metals/metalloids to cannabis. As a result, the concentrations of the essential elements under study in the 12 soil analyzed samples ranged as follows: up to 86 ppm for Na, 351 to 729 ppm for Mg, 921 to 2042 ppm for K, 400 to 1277 ppm for Ca, 204 to 467 ppm for Mn, 4536 to 14,007 ppm for Fe, 5 to 13 ppm for Cu and 8 to 25 ppm for Zn (Table S4).

With the exemption of Hg of which no detectable concentrations were found in both cannabis samples and soils, the concentration levels of the highly toxic Cd, As and Pb were determined at high levels, but again, they were comparable to previous studies in Cannabis samples from different geographical regions. In particular, Pb ranged from 11 to 854 ppb, As up to 242 ppb, and Cd was determined up to 181 ppb (Figure 3). For comparison, Zafeiraki et al. reported concentrations of 0.01 to 0.85 for Pb, 0.01 to 0.1 ppm for As and 0.003 to 0.18 ppm for Cd. The concentration of Pb in the soil samples ranged from 2.0 to 5.0 ppm, As ranged from 1.2 to 13 ppb, while Cd and Hg were not detectable in any soil samples (Figure 4). Other literature reports on Pb, Cd and As concentrations on cannabis report values averaging 0.5 ppm indicating that the values determined in this study are high but comparable to the literature.

Of the amounts recorded for the highly toxic metals, Pb concentration was the highest and can harm anyone who takes Cannabis contaminated to this level. Pb contamination in the environment arises from both naturally occurring reasons as it is a component in the earth’s crust, and human activities such as mining, burning of fossil fuels and manufacturing. Particularly for soil, it is known that Pb particles from anthropogenic activities settle on soil and can last for years [27]. In Ghana, Pb is ubiquitous in soil, especially near mining areas, industrial sites, incinerators, farms, landfills, and waste sites. The farms from which samples were collected are in close proximity to landfills, which explains the high levels of the toxic metal [28]. The toxic metal is carcinogenic, and Pb poisoning very effectively occurs by inhalation. In concentrations higher than 0.1 ppm, Pb is known to affect the gastrointestinal tract and the central nervous system, causing a variety of adverse effects on the human organs, particularly to kidneys [29].

The presence of Cd in cannabis originates from atmospheric Cd emissions, metal production, sewage mud, fertilizers, and disposal of batteries and other Cd-containing sources [30]. CdO also exists as small particles in the air (fumes) resulting from smelting, soldering, or other high-temperature industrial processes. Since the soil from Boti had no Cd content, but the plants had, it can be suggested that the metal source is airborne. Table S6 shows the level of Cd recorded for 22 seized cannabis samples from 2017 to 2020. Out of the 22 samples, only 2 showed no amount of Cadmium. The three towns used for this research are known for illegal cultivation and big bust of Cannabis by the security agencies. With cadmium level as a varying factor between the visited farms in Bono and Eastern regions, the seized Cannabis can be grouped according to the presence or absence of Cadmium in them. This can be a preliminary way of determining the source of Cannabis when seized. The toxic metal is a known carcinogen which affects the skeletal, urinary, reproductive, cardiovascular, central and peripheral nervous, and respiratory systems. Although, the presence of Hg in Ghana soils has been recorded in the literature, neither the soil areas under study, nor the cannabis samples contained any detectable concentrations.

The amounts of As in both cannabis samples and soils also raises concern as the toxicity of arsenic depending on exposure dose, frequency, duration, biological species, age, and gender, as well as on genetic and nutritional factors, can be very dangerous with cancer known as one of the long-term environmental exposure to it [31,32].

As Discriminant analysis (DA) has been used by others as a possible technique to link a suspect to a crime scene [11,21,33], it was applied to the data from this study in order to observe if DA is appropriate for a real-world scenarios and in particular for case of seized by the police cannabis samples. For the discriminant analysis plots all data were processed the same, using JMP 14. Data was normalized by using z-scores. As a result, DA is depicted for the cannabis samples using 11 elements in Figure 5. The data set includes data from Farms 1, 2 and 3 as well as data from all the seized samples. Due to the nature of the samples (police seized) the locations are kept confidential.

As it is suggested by the DA analysis, the seized samples from the year 2020 do not overlap with any of the three farms, this indicates that there is a high probability that these samples did not come from any of the three farm areas. Since Farms 1 and 2 did not have any overlap with any of the seized samples they can be excluded from the possibility that they came from these sites. Farm 3 had the most overlap with the seized samples, and overlapped with the years 2017, 2018 and 2020. Six samples from 2019 overlapped with the ellipsoid of farm 3. A comparison of percent differences is shown in Table 1.

As it is evident, the elements that had the smallest percent difference in Table 1 between samples 34 and 20 were Mg, K, Cu, Zn, As and Cd. Samples 34 and 20 had nearly identical amounts of Mg, As and Cd. Even though samples 8 and 14 were taken in two different years, there is a high probability that these samples could have come from the same area based on their location in the DA plot in Figure 4 and the percent differences in Table 1. Even samples that come from the same farm (samples 31 and 34) have varying percent differences in elemental composition as demonstrated in Table 1, as it can be seen in Figure 6 the DA of the soil samples that shows the soil samples were better separated and this could be due to the fact that the soil samples have a more consistent elemental concentration compared to the cannabis samples which included flowers and leaves.