Short answer. It isn't just nostalgia. In nearly 90,000 lab results from six US states, THC-dominant flower fell into three terpene groups, and two of them (caryophyllene–limonene and myrcene–pinene) held about 87 % of samples with terpene data [1]. Thousands of strain names sit on top of that narrow chemistry [1][3]. Meanwhile the traditional landraces that once supplied distinct profiles are being hybridised away in their home regions [6][7], and retail prices rise with THC, not aroma [9]. The evidence shows convergence. GenetixBay's answer is to preserve, document and share what's distinctive.
House position, carefully framed. The founder's view, after growing thousands of cultivars, is that modern commercial lines are converging, distinctive old profiles are being bred out, and the best is kept by a few. We don't claim a plot. The data below show convergence driven by ordinary forces: everyone breeding from the same few popular parents, the market paying for THC, and landraces disappearing at source. Those are things a community can do something about.
What the data show
| Study | What they measured | Finding |
|---|---|---|
| Smith et al. 2022 [1] | 89,923 flower lab results, six US states; 42,843 with terpenes | 96.5 % THC-dominant. THC-dominant samples clustered into three terpene groups: I caryophyllene–limonene, II myrcene–pinene, III terpinolene–myrcene. Groups I and II held 87.4 %. Commercial labels matched chemistry poorly |
| Reimann-Philipp et al. 2020 [3] | Nevada medical cannabis, chemistry and genetic markers | Hundreds of strain names, but terpene profiles fell into a few clusters: names "misrepresent" the real diversity |
| Watts et al. 2021 [2] | Genomes and chemistry of retail samples | "Indica" and "sativa" labels were genetically indistinct genome-wide; only a few terpenes tracked the labels |
| Schwabe & McGlaughlin 2019 [11] | Same-named strains from dispensaries in three states | Most named strains had genetic outliers; names don't guarantee a distinct genotype |
| Lynch et al. 2016; Sawler et al. 2015 [4][5] | Genotypes of hundreds of drug-type varieties | A few broad genetic groups underlie the drug-type market, and reported ancestry only partly matches genotype |
Why it's happening
1. Everyone breeds from the same parents. Modern drug-type cannabis is a hybrid swarm built from a limited set of founders, and popular lines get crossed with each other repeatedly (see how hybrid strains emerged). The genetic groups found in commercial samples are few and broad [4][5].
2. The market pays for THC. In about 30 million Washington State flower sales, price per gram rose with THC content [9]. Labels, lab results and shelf placement lead with a THC number. When buyers pay for THC, breeders select for it, and aroma diversity isn't what gets selected.
3. Landraces are disappearing at source. McPartland and Small describe the high-THC domesticates as endangered: 50 years of introgressive hybridisation "obliterated differences" between them [6]. In Morocco's Rif, traditional kif has largely been replaced by introduced high-yielding hybrids [7]. Similar replacement is widely reported elsewhere [8], though documented less systematically.being reviewed
4. Names hide the sameness. With thousands of names and few chemotypes, the shelf looks diverse when it isn't [1][3][10]. A new name is easier to produce than a new terpene profile.
What the data don't show
- Not that diversity is gone. Cluster III (terpinolene-led) and rarer profiles exist [1]; heirloom and landrace lines still circulate.
- Not who is hoarding what. "The best is kept by a few" is the founder's experience, not something these studies measure. We present it as his view.
- Not that THC selection is anyone's plan. It's what a price signal does.
The gas and candy trend
House position: the gas and candy trend is killing the classics. The founder's view, and GenetixBay's: market selection for "gas" and "candy" is crowding out the old profiles (incense, hash, pine, haze). Below is what the chemistry shows about those two trends and why they narrow the pool. How fast the classics are losing shelf space hasn't been measured; that part is the founder's experience.
What "gas" is, chemically. The skunky, fuel-like note that buyers call gas comes mostly from a family of prenylated volatile sulfur compounds, not from terpenes. Oswald and colleagues found seven of them across 13 cultivars; the key one, 3-methyl-2-butene-1-thiol (labelled VSC3), was the most abundant sulfur compound in the samples a panel rated most pungent, and adding it to a mix of major cannabis aroma compounds recreated the characteristic smell [12]. These compounds rose sharply towards the end of flower, peaked during curing and then fell off in storage [12]being reviewed. Sulfur compounds are smelled at tiny concentrations, so a trace of them dominates a jar.
What "candy" and "tropical" are. A follow-up study of 31 rosin samples found that the sweet, fruity, "exotic" aromas track minor, non-terpene compounds: esters, and a class of tropical sulfur compounds built on a 3-mercaptohexyl group (3-mercaptohexanol, 3-mercaptohexyl acetate and 3-mercaptohexyl butyrate), the same family that gives some hops and tropical fruit their character [13]. Savoury "chemical" notes tracked skatole [13]. The key line: terpene profiles were "remarkably similar" across varieties whose aromas differed sharply, and terpenes contributed mainly to the generic cannabis smell, not to what made each variety distinctive [13].
Why that narrows the pool.
- The backbone stays the same. If the "new" aromas ride on trace compounds, a menu of candy and gas cuts can sit on the same couple of terpene backbones the lab data already show: caryophyllene–limonene and myrcene–pinene, with 87 % of THC-dominant samples between them [1]. Different top notes; same underlying chemistry, often the same few parents (see same cut, a thousand names).
- The market rewards loud. When consumers rated strains by smell, citrus-sweet-pungent samples were judged more potent and worth more per gram than earthy-woody ones, yet perceived potency didn't correlate with measured THC at all [14]. Loud sells, so breeders cross loud to loud, and every popular gas or dessert cut becomes the parent of the next wave.
- The classics lose by default. Incense, hash, pine and old haze profiles are mostly earthy, woody, spicy or terpinolene-led. Terpinolene-dominant flower is already the smallest of the three commercial groups, around an eighth of THC-dominant samples [1]. Nobody bans a classic; it just stops being crossed, stops being stocked, and survives only where someone keeps it.
What the data don't show. No study we've found counts menu share by aroma type over time, so "dessert and candy crosses dominate menus" is a widely shared observation, including the founder's, not a measured trendbeing reviewed. Nor does the chemistry say gas or candy is bad. They're real, interesting compounds. The problem is a market that selects for them at the expense of everything else.
What to do about it. Keep and share the classic profiles (cuts, seed, males in the Stud Book), describe aroma on GB-ID records in words and chemistry, and when you hunt, keep the odd incense or pine keeper even if it isn't what's trending.
What GenetixBay does about it
- Preserve. Wanted-line and revival programmes, clone backups, males in the Stud Book, seed vaults (see landraces of the world and clone-only preservation).
- Document chemistry, not names. GB-IDs carry lab-tested terpene profiles, so a distinctive line can be found by what it is, not what it's called (see chemotype inheritance).
- Reward aroma, not just THC. On the Bay, potency numbers never earn rewards by themselves; documentation, males and preservation do.
- Share openly. Distinctive lines registered, described and shared in legal lanes stay in the commons (see patents and the gene pool).
For growers: finding something different
- Ask for the certificate of analysis and read the terpene section, not just THC.
- Look for the less common dominants: terpinolene, ocimene, or high levels of guaiol and eudesmols (associated with some Central Asian lines [2][6]).
- Grow heirloom and landrace seed with a documented history, and pheno-hunt for aroma (see selection and pheno-hunting).
Evidence
- Commercial THC-dominant flower clustering into a few terpene groups: Supported by large lab datasets [1][3].
- Labels and names misrepresenting diversity: Supported [1][2][3][11].
- Landrace erosion by hybridisation: Supported [6][7]; extent outside Morocco less documented.
- Price tracking THC: Supported in one large market [9].
- "The best is hoarded by a few": founder's view, not measured.
- "Gas" aroma driven by prenylated volatile sulfur compounds (VSC3): Supported by one detailed study [12].
- Candy/tropical aromas driven by minor non-terpene compounds, with terpene profiles similar across divergent aromas: Supported by one study of 31 extracts [13].
- Consumers rating loud, sweet-pungent flower as more potent and valuable without any link to measured THC: Supported [14].
- The gas/candy trend crowding out classic profiles: house position (founder's view); menu share by aroma type not yet measured.
