Clove Story
A garlic cultivar reference for home gardeners. Eleven groups, 82 cultivars, scored against your winter and your kitchen. Every claim carries the evidence behind it, including where there is none.
Garlic groups are not interchangeable.
Supermarket garlic is almost always a softneck, bred to store long and pack tightly. Those traits matter to a distributor and are optional in a home bed.
The eleven groups differ in what a gardener notices: how many cloves a bulb sets, how easily the skins come away, how many months it keeps, whether the stem braids, and whether it will throw a scape at all in your winter. A bulb sits in the ground for eight or nine months, so the choice is annual and it is made before planting.
Frequently asked questions.
How much cold does garlic actually need?
Measured For cloving, possibly none. The best controlled trial on this grew every plant warm all season and chilled only the cloves beforehand. The unchilled control made normal multi-clove bulbs, the largest in the trial, and had a one-clove bulb rate of zero. Chilling raised the rate of rounds, to 34 percent at its worst, and cut yield by up to three quarters.
Measured For scapes, cold matters a great deal. Bolting went from 3.3 percent unchilled to 66.7 percent at 10°C for 40 days. If you are growing a hardneck for its scape, that is the trait the cold is buying you.
Measured And it is an optimum rather than a gradient. At 40 days, ten degrees beat both 5 and 15. Too little cold and it will not bolt; too much and the plant grows badly and bolts worse anyway.
Cultivar matters more than any of this. Aoba and Takagi's Japanese material would not bulb at all without cooling, the opposite result. Two of Azmi's Indonesian clones needed none. Nobody has published a figure for any variety sold here.
How do I compare two garlics fairly?
Measured Not by eating them one after another. Pardo's panel designed around a specific problem: pungency is not a basic taste but a stimulus that accumulates, so the same garlic tastes hotter the later in a session you meet it. Sequential tasting will tell you the last sample was the strongest whether or not it was.
Their method was a rubbing test on a neutral carrier: small toasts with a few drops of olive oil, each halved clove rubbed five times, all samples prepared identically by one person, then ranked rather than scored. A Mexican panel working on the same problem in 2026 landed on mashed potato as the carrier and green apple or lemon water as the palate cleanser.
Does it matter how big a clove I plant?
Measured Yes, and the return flattens. D'Anna's Sicilian trial: 1.0 g cloves gave 4.6 t/ha, 5.0 g gave 8.8, 6.0 g gave 9.5. Desta's Ethiopian trial: 3 to 3.5 g cloves outyielded 1 to 1.49 g by 25.9 percent and matured 28 days sooner.
Since seed garlic sells by weight, the arithmetic matters. Five times the seed weight buys 91 percent more crop; the last 20 percent of seed weight buys 8 percent. Plant your big cloves and eat the small ones, but do not pay a large premium for the top grade.
When do I plant?
Consensus Northern extension services say within a week or two of the first killing frost, which gives roots four to six weeks before the ground freezes while stopping shoots from emerging above the soil line. South Dakota State gives a soil target of 50°F at four inches. UC Master Gardeners in Santa Clara County invert the rule for mild winters and say two weeks before first frost, mid October through November.
Can I plant supermarket garlic?
One source Purdue advises against it. Temperatures above 77°F can inhibit bulb formation and supermarket stock is usually held warm. Note that the Purdue page carries no measurement, and that the seed-storage figure this site follows is UMass at 50°F, not 40. See Storage in the grow guide.
Should I cut the scapes?
Measured Yes, and where you cut decides whether it works at all. Guo and colleagues compared proper removal at the base against cutting at the flag leaf and at the spathe. Proper removal gained 12.1 percent yield. Both partial cuts gained nothing, because the stub kept growing. If you snip the top off a scape you have done the work and bought none of the benefit.
Measured The size of the gain is unsettled. Guo got 12.1 percent, and 4.6 to 27.0 percent across five cultivars. Rosen and Tong got 15 percent on their low organic matter site and a nonsignificant result on finer soil. Colorado State and Sekhon and Garg both found no significant effect. Wisconsin frames it as a loss avoided, up to 30 percent in poor soil and under 5 percent in well fertilized soil, which reconciles most of it.
Measured Do not cut early. Guo measured an 80.7 percent rise in leaf membrane damage after removal and read it as accelerated senescence. Ontario separately reports that nicking one leaf during removal costs about 17.5 percent of yield, so a careless cut takes more than the scape would have.
How should I store the harvest?
Unsourced Cold wins outright. UC Davis gives 30 to 32°F at 60 to 70 percent humidity and more than nine months, against 3 to 5 months at 60°F and 1 to 2 months warm. The familiar rule that garlic keeps either cold or warm but not in between does not survive those numbers, and the sprouting band itself is disputed: UMass says 40 to 50°F, UC Davis says 41 to 65°F, and neither cites a measurement. A home refrigerator is a poor choice mainly because it is far too humid. Seed garlic is held warmer on purpose, at 50 to 60°F. The grow guide has the full trail.
Can I grow garlic where winters are warm?
Consensus Softnecks and Creoles, yes. UF/IFAS states that hardneck garlic needs significant winter and in some cases will not produce in Florida at all, and recommends softnecks plus elephant garlic. UGA, NC State and UC Master Gardeners agree directionally.
Measured The pre-plant refrigeration workaround has real research behind it, though not the version usually quoted. The chilling literature exists precisely because it substitutes for a winter the plant will not get in the ground: Bandara found 4°C for 45 to 60 days improved cloving in spring-planted garlic. Azmi found 20 days at 5°C sufficient where cold helped at all, and two clones that wanted none.
Note that this is the one case where a storage-chilling number is the right tool. Applying the same figures to fall-planted garlic in a cold region is a category error, because that crop gets its cold in the ground.
Why do group counts differ between sources?
Measured Volk and Stern and Arizona Extension both use ten. Cornell CCE Onondaga lists eleven, adding a Middle Eastern group absent from the peer-reviewed work. Minnesota, Wisconsin and Washington State use a simplified five. Alaska Fairbanks classes Creole as softneck where Arizona and Cornell call it hardneck. This site uses ten Allium sativum groups plus elephant garlic, and flags disputed boundaries on the group itself.
How is confidence assigned?
Two different rules run, so here are both. For the eleven groups: high means three or more independent sources agree on the key figures, medium means two sources or a range spanning several, low means one source or sources in direct conflict. For the 82 cultivars: high means two or more catalogs list it and six or more fields are filled, medium means two catalogs or a well-filled single catalog, low means one thin listing or no group. No figure here is estimated. Where nothing was published, the cell is blank.
Two questions.
The first sets what will grow. The second sets what is worth growing.
Your winter
Cold duration decides whether a hardneck bolts, which is what a scape is. Answer for the ground, not the calendar.
Your kitchen
What you want out of the bulb once it is cured. Pick as many as apply. Some of these pull against each other, and the picker will say so.
Eleven groups of cultivated garlic.
Ten are Allium sativum. Turban and Asiatic bolt weakly and sit between the two neck types. Elephant garlic is not a garlic and belongs to the leek species, included because it is sold beside garlic.
Seven of these groups were genetically differentiated in a USDA AFLP study of 211 Allium sativum and Allium longicuspis accessions. Turban, Creole and glazed purple stripe were underrepresented in that dataset and not clearly separated, so those three rest on morphology rather than molecular evidence.
What garlic needs from winter.
Nearly every garlic page, this one included until it was checked, opens by saying garlic needs cold or it will grow one round bulb instead of splitting into cloves. The best controlled experiment on the question found the opposite. Every claim below carries the kind of evidence behind it.
Measured Cloving does not require cold. Wu and colleagues chilled cloves of a bolting cultivar at 5, 10 and 15°C for 20, 40 or 60 days, then grew every plant warm at 22°C day and 16°C night for the whole season so that only the pre-plant chilling counted. The unchilled control produced normal multi-clove bulbs and the largest bulbs in the trial. In their words, this "contradicts the traditional view of chilling as being essential for cloving and bulbing of garlic."
Measured Chilling produced the round bulbs, not the lack of it. The unchilled control returned a one-clove bulb rate of zero. Chilling raised it, worst at 5°C for 40 days, where 34 percent of bulbs came out as single rounds. Colder meant more rounds at every duration. Longer did not: at 5°C the rate peaked at 40 days and fell again by 60.
Measured A second study, different continent, different method, same answer. Rahim and Fordham stored cloves of a tropical cultivar at 5°C or 20°C for 30 days and then followed clove initiation under a scanning electron microscope. Neither treatment had clove initials at the end of storage. Both cloved. The cold-treated plants matured 30 days sooner, at 87 days against 117, but produced fewer cloves and lighter bulbs, 2.08 g mean dry weight against 2.77 g. Their opening line is precise: low storage temperature advances bulbing. Not enables.
Measured Chilling cost yield, heavily. The unchilled control yielded 7.94 t/ha at a mean bulb weight of 49.7 g, the best of the ten treatments. Every chilled treatment was worse. Sixty days at 5°C dropped it to 2.15 t/ha, about a quarter of the control, at 27 g per bulb.
| Clove treatment | Bolting | One-clove bulbs | Yield t/ha | Mean bulb |
|---|---|---|---|---|
| None, held at 22°C | 3.3% | 0% | 7.94 | 49.7 g |
| 5°C, 20 days | 59.3% | 10.0% | 7.32 | 44.1 g |
| 5°C, 40 days | 22.0% | 34.2% | 5.70 | 30.1 g |
| 5°C, 60 days | 30.0% | 22.9% | 2.15 | 27.0 g |
| 10°C, 20 days | 59.3% | 5.3% | 6.21 | 40.0 g |
| 10°C, 40 days | 66.7% | 6.0% | 5.55 | 38.2 g |
| 10°C, 60 days | 54.0% | 8.4% | 2.98 | 22.7 g |
| 15°C, 20 days | 54.0% | 4.7% | 6.21 | 38.7 g |
| 15°C, 40 days | 56.7% | 1.4% | 5.97 | 35.9 g |
| 15°C, 60 days | 56.0% | 3.3% | 5.95 | 35.0 g |
Wu et al. 2015, cultivar G064, three replicates. Every plant grown warm all season so the clove treatment is the only cold in the experiment.
Measured What cold does reliably is drive the scape. Bolting went from 3.3 percent unchilled to 66.7 percent at 10°C for 40 days. For a hardneck, which is defined by producing a scape, that is the trait cold actually controls. Cloving is far less dependent on it than the advice suggests.
Measured The response is an optimum, not a gradient. At 40 days ten degrees beat both 5 and 15°C for bolting. At 20 days 5 and 10 tied, and at 60 days 15°C came out slightly ahead. The authors state the relationship "did not follow a simple linear relation," and explain it: colder should induce more strongly, but too cold impedes vegetative growth, and a plant that has grown badly bolts badly. Colder and longer is not better.
Measured The requirement is cultivar-specific to the point where no general number exists. Aoba and Takagi's Japanese material failed to bulb at all without cooling, which is the opposite of Wu's G064. Azmi tested Indonesian clones at 0, 5 and 10°C for 20, 40 and 60 days and found two highland clones needed no cold whatsoever, one of them yielding best untreated. Ipek and Utku gave four genotypes an identical 4°C treatment for twelve weeks and sprouting advanced by anywhere between 4 and 26 days.
Measured Long days can substitute for cold. Aoba and Takagi found plants under 20 hours of daylight or more formed bulbs with no cooling at all. Cold does not cause bulbing. Whether the two effects add, and by how much, is not something any of these studies measured. Wu and Atif both found bulbing driven by long days plus warmth of 20 to 30°C.
Unsourced "Six to eight weeks below 40°F" has no research establishing it. Four extension services publish it in near-identical wording, two others publish four to six weeks, and none cites anything. One study lands on those numbers, Bandara recommending 4°C for 45 to 60 days, but it is a Canadian spring-planting trial about avoiding winterkill and its abstract states chilling was not a prerequisite for clove formation. The 40°F figure most plausibly descends from Mann's 1952 University of California report, which used 40°F as a storage treatment level and concluded the field effect was always slight.
Unsourced The hardneck and softneck zone chart. Sellers publish hardneck for zones 1 to 6 and softneck for 7 to 10. Fifteen extension services were checked and none assigns garlic types to zone ranges. NC State's Plant Toolbox, the only one publishing zone numbers at all, gives 4a through 9b for both.
Consensus Warm-winter regions do still struggle with hardnecks. This part holds, and the scape finding explains why: a hardneck that gets no cold does not bolt, and a hardneck that does not bolt is not doing the thing you bought it for. UF/IFAS states hardneck garlic needs significant winter and in some cases will not produce in Florida at all. UGA, NC State and UC Master Gardeners agree directionally.
What nobody has published. No cold requirement for any cultivar a US gardener can buy. Every quantitative study runs on Chinese numbered lines, Israeli accessions, Japanese material, Indonesian clones, Brazilian cv. Ito, or Canadian trials. Wu's own discussion notes the literature contradicts itself on whether pre-plant chilling raises yield, lowers it, or does nothing.
That is why the picker asks what your ground does rather than for a zone number. There is no number to give.
Which traits travel.
Measured Volk and Stern grew ten cultivars, one from each group, on twelve farms across the United States and Canada for two consecutive seasons. Stable across locations: clove arrangement, clove number, clove weight, clove skin color and clove skin tightness. Skin tightness is what peelability is, so the peeling notes here hold wherever you plant. Not stable: bulb wrapper color and its intensity, described in the paper as highly dependent on location and cultivar.
Measured Growers rated Ajo Rojo, the Creole, Excellent. Spanish Roja, the Rocambole, was smallest by bulb weight and rated Poor. One cultivar stood for each group, and the rating measures field performance rather than flavor.
Measured The paper's own recommendation is to buy seed stock grown near you, or to allow a new variety several seasons to adapt, because stock from another region often does not show the size, shape and color the catalog listed.
Two things the research settles.
Most garlic advice is convention. These two are not.
Measured Big seed cloves make big bulbs, and the return flattens fast. D'Anna and colleagues planted graded cloves in Sicily: 1.0 g cloves returned 4.6 t/ha, 5.0 g returned 8.8, and 6.0 g returned 9.5. Desta and colleagues in Ethiopia found 3 to 3.5 g cloves outyielded 1 to 1.49 g cloves by 25.9 percent, 13.13 t/ha against 10.43, with bulbs of 36.1 g against 27.8.
Read the shape, not just the direction. Going from 1 g to 5 g is five times the seed and buys 91 percent more crop. Going from 5 g to 6 g is 20 percent more seed and buys 8 percent more crop. Since seed garlic is sold by weight, the premium on the largest grade is where the return runs out. The middle grades are the value.
Desta also recorded that the largest cloves matured 28 days earlier than the smallest, which no catalog mentions and which matters if you are trying to clear a bed.
Measured Cut the scape at the base or do not bother. Guo and colleagues compared removing the scape properly at the base against cutting it at the flag leaf and at the spathe. Proper removal raised bulb yield 12.1 percent and diameter 6.9 percent over plants whose scapes were left on. Both partial cuts did nothing at all, no significant change in yield, diameter or quality. The residual stub kept elongating, 0.47 cm and 2.76 cm respectively, spending the plant's reserves on a scape that was going nowhere.
Their preliminary trial across five cultivars put the gain from proper removal at 4.6 to 27.0 percent, significant in four of the five. The mechanism is measurable: auxin in the scape fell 57.8 percent, releasing apical dominance, root activity rose 38.2 percent, and nitrogen, phosphorus and potassium accumulation in the bulb rose 52.9, 47.8 and 34.3 percent.
One caution from the same paper. Leaf malondialdehyde, a marker of membrane damage, rose 80.7 percent after removal. The authors read that as accelerated leaf senescence and warn against cutting scapes too early, which shortens the leaf's working life.
Measured But the size of the scape gain is unsettled, and two trials found none. Rosen and Tong got 15 percent on their low organic matter site, significant at p = 0.05, and a non-significant 5 percent at p = 0.10 on the high organic matter site. Colorado State found no significant yield difference on German Extra Hardy. Sekhon and Garg, working in the north Indian plains, reported no effect on days to maturity, bulb diameter, bulb weight, cloves per bulb or cured yield. Wisconsin Extension frames it as a loss avoided, up to 30 percent in poorly fertilized soil and under 5 percent in well fertilized soil, which is the pattern that reconciles most of it: the worse your ground, the more the scape costs you. Pelter and colleagues add timing, from three years and two hardneck cultivars in central Washington: cutting early, between scape emergence and a half turn, raised total yield significantly, waiting until the scape had curled a full turn did not, and the early cut raised average bulb weight 13 and 39 percent by cultivar.
One source The two papers that measured bulb chemistry after scape removal disagree. Guo found soluble sugar up 17.1 percent, reducing sugar up 45.6 and allicin up 6.2. Bažon, working on two Croatian landraces, found dry matter, soluble solids and diallyl disulfide all fell. Nobody has reconciled them.
How to actually grow it.
Forty-nine recommendations across three guides, each with the evidence behind it and the tier that evidence earns. Where the usual advice and the primary literature disagree, both appear and the disagreement is named.
Eighty-two named cultivars.
Every value is what that cultivar's own product page prints. No group-level figure has been copied onto a cultivar, so a blank cell means the supplier does not publish it for that plant. Thirty-three cultivars are listed by more than one supplier; where their numbers differ, both appear.
Volk, Henk and Richards put the trade at more than 200 named garlic clones commercially available in the United States, many of them the same plant renamed as it passed between growers. Eighty-two is what three specialist catalogs document well enough to publish.
Sources are Filaree Garlic Farm, Keene Organics and Hood River Garlic. Supplier copy establishes what is sold and what it is like in the hand. It is not used for climate science, which comes from extension.
Measured About the flavor column. Pungency really does vary between garlic clones, and by a lot. Lamari and colleagues grew 34 Greek genotypes in one field and measured a 15-fold spread in pyruvate, the standard proxy for pungency. Hayat found a six-fold spread in allicin across 28 cultivars.
Unsourced But no study has ever tested the group framework against pungency. Searching the group names against pyruvate, allicin or thiosulfinate returns seller catalogs and nothing else. "Porcelain is hot, Creole is sweet" is untested.
Measured And the one dataset covering North American market names undercuts catalog language directly. Dickerson and Wall measured nine cultivars at three New Mexico sites. 'Mild French' was the most pungent variety at all three. Inchelium Red, a softneck Artichoke, was among the mildest. Their site-to-site spread within a single cultivar was as wide as the spread between cultivars, so where you grow it moves pungency about as much as which one you grow.
Measured And you cannot predict pungency by looking at a bulb. Pardo and colleagues ran 14 Spanish cultivars through instrumental analysis and then put five of them past 25 trained judges. Pungency differed significantly between cultivars, the purple Morasol clearly hottest and the white Ramses mildest. But pungency correlated significantly with nothing they measured: not color, not lightness, not firmness, not pH, not soluble solids, not moisture, not glucose, fructose, sucrose or total sugars. Every correlation came back nonsignificant.
Their cultivars are Spanish and sorted purple, white and Chinese, so this does not test the groups above. What it does show is that the visible and measurable properties of a bulb carry no reliable signal about how hot it will be.
Read the flavor column as what the grower who sells it says it tastes like. That is what it is.
Unsourced About the storage column, and the claim that softnecks keep longer. No research was found testing it. Zhu and colleagues scored 501 accessions from 38 countries for storability over two years and found the full range, decay index from 0 to 100 percent, strongly heritable. They did not analyse it by neck type or by group, and no other study does either. The catalogs agree with each other, which is not the same as evidence.
Tap any row for the per-supplier values, the climate line, and links to every catalog that lists it.
Eleven groups.
| Group | Neck | Storage | Cloves | Flavor | Peel | Braid | Harvest | Winter | Conf. |
|---|
Where the data comes from.
Group traits are anchored to specialist grower catalogs, which are the only published record of most of them. Climate, storage and agronomy are anchored to university extension and peer-reviewed work. Where a catalog and a bulletin disagree, both figures appear and the confidence rating drops.
Peer reviewed, read in full text
- Volk, G.M. & Stern, D. (2009). Phenotypic characteristics of ten garlic cultivars grown at different North American locations. HortScience 44(5):1238–1247. Free access. Table 1 gives the cultivar-to-group mapping and the twelve-farm figures.
- Volk, G.M., Henk, A.D. & Richards, C.M. (2004). Genetic diversity among U.S. garlic clones as detected using AFLP methods. JASHS 129(4):559–569. Free access. Source of the 211-accession figure.
- Wu, C., Wang, M., Dong, Y., Cheng, Z. & Meng, H. (2015). Growth, bolting and yield of garlic (Allium sativum L.) in response to clove chilling treatment. Scientia Horticulturae 194:43–52. The controlled dose-response this page leans on hardest. Ten treatments, three replicates, every plant grown warm all season so the clove chilling is the only cold in the experiment. Source of the cloving, rounds, yield and bolting figures above.
- Guo, Z., Liu, R., Wu, Q., Shi, W., Meng, X., Huang, Y., Ye, Y. & Wang, Y. (2026). Effects of scape removal position on nutrient allocation, yield, and quality of garlic. HortScience 61(6):1348–1357. Found through Harvard's library rather than open search. Source of the cutting-position finding, the 12.1 percent yield gain, the hormone and root-activity mechanism, and the leaf-senescence caution.
- Rahim, M.A. & Fordham, R. (1988). Effect of storage temperature on the initiation and development of garlic cloves. Scientia Horticulturae 37(1–2):25–38. Supplied by the reader after nine open access routes failed. Both treatments cloved; cold advanced maturity by 30 days and cost clove number and bulb weight.
- Lamari, F.N. et al. (2020). Variability in bulb organosulfur compounds, sugars, phenolics, and pyruvate among Greek garlic genotypes. Antioxidants 9(10):967. Open access. 34 genotypes in one field, 15-fold pyruvate range.
- Hayat, S. et al. (2016). Diversity in phytoalexin allicin content among garlic cultivars. Frontiers in Plant Science 7:1235. Open access. 28 cultivars, six-fold allicin range.
- Zhu, Y. et al. (2024). Genetic architecture of bulb storability of plentiful garlic germplasm resources. Horticulture Research 11(12):uhae260. Open access. 501 accessions scored for storability over two years. Does not analyse by neck type, which is why this page will not endorse the softneck storage claim.
- Pardo, J.E., Escribano, J., Gómez, R. & Alvarruiz, A. (2007). Physical-chemical and sensory quality evaluation of garlic cultivars. Journal of Food Quality 30(5):609–622. Free to read on Wiley. Supplied by the reader. 14 Spanish cultivars, 25 trained judges on five of them. Source of the finding that pungency correlates with nothing measurable, and of the tasting method.
- Dickerson, G.W. & Wall, M. (1997). Varietal evaluation of garlic in New Mexico. NMSU Research Report 717. An experiment station report, not a peer-reviewed article, flagged as such. The only measured pungency data on North American market names.
- Rosen, C.J. & Tong, C.B.S. (2001). Yield, dry matter partitioning, and storage quality of hardneck garlic as affected by soil amendments and scape removal. HortScience 36(7):1235–1239. Free access. Table 1 carries both scape results, including the nonsignificant one.
- Pelter, G.Q., Sorensen, E.J., Van Denburgh, R.W. & Hannan, R.M. (2005). Effect of scape removal on garlic bulb yield in Central Washington. Acta Horticulturae 688:323–326. A conference proceedings paper, flagged as such. Harvard does not hold the volume; read in full from an interlibrary loan scan. Three years, two hardneck cultivars under irrigation: early removal, at emergence to a half turn, raised total yield significantly and waiting for the full curl did not. The storage-life penalty its introduction floats is not in its own data, three months of storage showing no significant difference in weight loss, and the claim traces to a 1991 grower handbook from Filaree, not to a measurement.
- Azmi, C. et al. (2022). Temperature and duration of vernalization effect on the vegetative growth of garlic clones in Indonesia. Open Agriculture 7(1):520–528. Open access. The clone-by-clone cold requirements, including the clones that need none.
- Wu, C., Wang, M., Cheng, Z. & Meng, H. (2016). Response of garlic bolting and bulbing to temperature and photoperiod treatments. Biology Open 5(4):507–518. Open access. Note its vernalization was five months of ambient field winter, unrecorded, so it cannot support any cold-hours figure.
- Atif, M.J. et al. (2019). Influence of different photoperiod and temperature regimes on growth and bulb quality of garlic cultivars. Agronomy 9(12):879. Open access. Long days plus warmth drive bulbing.
- Dufoo-Hurtado, M.D. et al. (2015). Low-temperature conditioning of garlic seed cloves. Frontiers in Plant Science 6:332. Open access.
- Morais, É.G. et al. (2023). Treating garlic seed cloves with negative temperatures. Horticulturae 9(9):1022. Open access.
- Mann, L.K. (1952). Garlic bulb studies: effect of day length, temperature during growth, and storage temperature. California Agriculture 6(6):13–14. A University of California research report rather than a peer-reviewed article, flagged as such. The most plausible origin of the 40°F figure, and its own conclusion was that the field effect was slight.
Read as abstract only
- Aoba, T. & Takagi, H. (1971). Studies on the bulb formation in garlic plants III. J. Japan. Soc. Hort. Sci. 40(3):240–245. The most load-bearing citation on this page, and it is an abstract. The English abstract on J-STAGE was read verbatim; the body is in Japanese and was not read. The 5 to 15°C range, the 20 to 30 day duration, the gradient finding and the daylength substitution all come from that abstract.
- Rohkin Shalom, S. et al. (2015). Storage temperature controls the timing of garlic bulb formation via shoot apical meristem termination. Planta 242(4):951–962. Publisher abstract, verbatim.
- Bandara, M.S., Krieger, K., Slinkard, A.E. & Tanino, K.K. (2000). Pre-plant chilling requirements for cloving of spring-planted garlic. Can. J. Plant Sci. 80(2):379–384. Paywalled. The verbatim conference abstract came from AGRIS.
- Kamenetsky, R. et al. (2004). Environmental control of garlic growth and florogenesis. JASHS 129(2):144–151. Listed as open access but returned 403. One publisher-supplied phrase quoted; the rest was a machine reconstruction and is not used here.
- Pooler, M.R. & Simon, P.W. (1993). Garlic flowering in response to clone, photoperiod, growth temperature, and cold storage. HortScience 28(11):1085–1086. Reconstructed abstract only. Not quoted on this page.
- D'Anna, F., Iapichino, G. & Miceli, A. (2000). Effect of clove weight on yield and bulb quality of garlic grown for storage. Acta Horticulturae 533:589–592. Free ISHS abstract, verbatim. It carries the full yield-by-clove-weight figures used here.
- Bažon, I. et al. (2021). Bolting garlic quality and morphological traits are influenced by scape removal. Acta Horticulturae 1320:117–119. Free ISHS abstract. Direction of the quality effects only; magnitudes are behind the paywall.
- Sekhon, N. & Garg, K.S. (2016). Earliness, yield and bulb parameters of hardneck garlic as influenced by leaf knotting and scape removal in north Indian plains. Indexed abstract only, via Harvard's library. Reports no effect of scape removal on maturity, bulb diameter, bulb weight, cloves per bulb or cured yield.
- Ipek, M. & Utku, O. (2013). Effect of pre-planting storage temperature on plant development of garlic clones. Bahçe 42(1–2):1–11. Abstract via AGRIS. Source of the 4 to 26 day genotype spread.
Wanted and could not be read
Gray literature and secondary sources
- Zandstra, B. Scape removal and bulb yield, via the Government of Ontario. The cut-leaf penalty comes from this summary. The 30 percent framing used on this page comes from Wisconsin Extension, not from here. No peer-reviewed primary publication exists. Searches of Acta Horticulturae, HortScience and the Canadian Journal of Plant Science return nothing, and the Ontario page cites "Zandstra 2000" and "Zandstra 2006" with no journal, volume, statistics or tables.
- Crowe, F.J. & Debons, J. DADS trials 1987–1992, Oregon State University COAREC. Extension proceedings, not peer reviewed. The only trial that produced white rot free plots, and the only place the two-application result appears.
- Colorado State University Specialty Crops Program, The Garlic Project 2004. An extension project report, not peer reviewed. Replicated field trial on German Extra Hardy: "no significant statistical differences in garlic yield between the plants with scapes removed and the plants with scapes left on."
Grow guide: read in full text
- Dugan, F.M., Hellier, B.C. & Lupien, S.L. (2007). Pathogenic fungi in garlic seed cloves from the United States and China. Journal of Phytopathology 155(7–8):437–445. The most load-bearing paper in the grow guide. Source of the finding that every commercial seed lot tested carried garlic pathogens, of the wounding requirement for Fusarium, and of the failed fungicide dips.
- Smychkovich, A. & Hashemi, M. (2025). Fall-planted cover crops improve nitrogen use efficiency but reduce garlic yield. J. Sci. Food Agric. 106(2):919–930. UMass Amherst, 'German Extra Hardy', two seasons. The only Northeast nitrogen response curve on this page.
- Desta, B., Tena, N. & Amare, G. (2021). Growth and bulb yield of garlic as influenced by clove size. The Scientific World Journal 2021:7351873. Full text via PubMed Central. All five size classes and their statistics, which the earlier abstract-only reading could not reach.
- Desta, B., Woldetsadik, K. & Ali, M. (2021). Effect of harvesting time, curing and storage methods on storability of garlic bulbs. The Open Biotechnology Journal 15(1):36–45. The harvest-timing trade-off. Note that significance is reported at the ANOVA level without contrast p-values.
- Mouttet, R. et al. (2014). Pest Management Science 70(7):1017–1022. Source of the statement that Ditylenchus dipsaci cannot be detected by eye, and of the biological-race point. Written about alfalfa seed, not garlic cloves, and used here as an analogy.
- Marek, M. et al. (2010). Plant Pathology 59(5):931–943. The more than ten biological races with limited host ranges.
- EPPO (2017). PM 7/87 (2) Ditylenchus destructor and Ditylenchus dipsaci. EPPO Bulletin 47(3):401–419. Diagnostic standard. Molecular methods are specified for low infestation, which concedes that morphological detection is inadequate there.
- Clarkson, J.P. et al. (2006). Plant Pathology 55(3):375–386. Introduction read. States that no commercial Allium cultivar has exploitable white rot resistance.
- Taşkin, H. et al. (2013). The Scientific World Journal 2013:781282. Meristem culture against shoot-tip culture, and why the two are not the same claim.
- Conci, V.C., Canavelli, A.E. & Balzarini, M.G. (2010). Journal of Phytopathology 158(3):186–193. Source of the finding that cloves within a bulb share virus status, so selecting cloves does not select against virus.
- Sopha, G.A. et al. (2024). Aerial bulbils as garlic alternative planting materials, a systematic review. Chilean J. Agric. Res. 84(1):123–131. The two-season figure, and the point that bulbils reduce fungal carryover without cleaning virus.
- Barboza, K. et al. (2022). Crop Science 62(5):1807–1820. Twelve cultivars, four sites. The solids-to-storage correlation, and the hardneck that topped it. The postharvest quality column is a classification descriptor rather than a storage measurement made in that study, which is why this is presented as suggestive rather than as a refutation.
- Madhu, B., Mudgal, V.D. & Champawat, P.S. (2019). Storage of garlic bulbs: a review. J. Food Process Eng. 42(6):e13177. Read to trace the sprouting-temperature figure, which it attributes to the UC Davis produce facts sheet.
Grow guide: read as abstract only
- Crowe, F.J., Hall, D.H., Greathead, A.S. & Baghott, K.G. (1980). Inoculum density of Sclerotium cepivorum and the incidence of white rot of onion and garlic. Phytopathology 70:64–69. The sclerotial density brackets, which are the load-bearing numbers in the white rot section. No DOI is registered for that issue and the publisher full text returned 403.
- Davis, R.M., Hao, J.J., Romberg, M.K., Nunez, J.J. & Smith, R.F. (2007). Plant Disease 91(2):204–208. Over 90 percent sclerotial kill and continued crop loss a year later.
- Elshahawy, I.E. et al. (2019). Heliyon 5(1):e01168. The low-inoculum and high-inoculum split that reconciles the stimulant literature.
- Merriman, P.R., Isaacs, S., MacGregor, R.R. & Towers, G.B. (1980). Annals of Applied Biology 96(2):163–168. The two-site onion oil result, one success and one failure at higher inoculum.
- Coley-Smith, J.R. & Entwistle, A.R. (1988). Susceptibility of cultivars of garlic to Sclerotium cepivorum. Plant Pathology 37(2):261–264. Five cultivars including one sold as resistant. No resistance found in any.
- Coley-Smith, J.R., Mitchell, C.M. & Sansford, C.E. (1990). Long-term survival of sclerotia. Plant Pathology 39(1):58–69. The twenty-year figure, on laboratory-grown sclerotia. The forty-year figure repeated by some extension services traces to nothing we could find.
- Roberts, P.A. & Matthews, W.C. (1995). Disinfection alternatives for control of Ditylenchus dipsaci in garlic seed cloves. Journal of Nematology 27(4):448–456. PMID 19277311. The hot water treatment paper, and the biggest single gap on this page. The PubMed Central full text is gated behind a challenge and returned empty. Every protocol figure here comes from a detailed abstract. Per-treatment efficacy, emergence percentages and cultivar identity are unknown to this page.
- Roberts, P.A. & Greathead, A.S. (1986). Journal of Nematology 18(1):66–73. PMID 19294142. Trace infection survived every treatment including the hot water and formalin dip.
- Poirier, S. et al. (2019). Host range and genetic characterization of Ditylenchus dipsaci populations from eastern Canada. Plant Disease 103(3):456–460. The rotation crops that carry no reproduction. The closest tested match to Northeast garlic stock.
- Testen, A.L. et al. (2014). Plant Disease 98(6):859. Ohio first report, naming introduced planting material.
- Mollov, D.S., Subbotin, S.A. & Rosen, C. (2012). Plant Disease 96(11):1707. Minnesota first report. Nematodes in all eighteen submitted bulbs.
- Lot, H., Chovelon, V., Souche, S. & Delecolle, B. (1998). Plant Disease 82(12):1381–1385. The OYDV and LYSV yield losses. The publisher returned 403 to both the direct link and the resolved DOI, so this is reported as could not retrieve rather than paywalled.
- Conci, V.C. et al. (2003). Plant Disease 87(12):1411–1415. The five-year decay curve.
- Melo Filho, P.A. et al. (2006). European Journal of Plant Pathology 116(2):95–101. The seven-year curve and the 83 percent reinfection figure.
- Bloem, E., Haneklaus, S. & Schnug, E. (2011). J. Agric. Food Chem. 59(9):4442–4447. The sulfur result, and the non-significant nitrogen and storage trend.
- Liu, P. et al. (2024). J. Agric. Food Chem. 72(17):10117–10126. Stable at 0°C for 290 days; a shifted metabolite profile at 22°C.
- Walters, S.A. (2008). HortTechnology 18(2):286–289. The mulch result that disagrees with Colorado on winter survival. The ASHS platform returned a script shell to direct requests and 403 to the fetcher, so this is reported as could not retrieve. Abstract obtained through AGRIS.
- Dugan, F.M., Lupien, S.L. & Hellier, B.C. (2019). Crop Protection 116:43–48. Bibliographic record only, which is less than an abstract. Cited for the direction of the bulbil result and for nothing else.
Grow guide: extension and grey literature
- Cranmer, T. (2024). When is the best time to plant garlic? ONvegetables, Ontario Ministry of Agriculture, Food and Agribusiness. The only fall planting-date trial in a cold climate we could find. Significance is asserted in prose; no tables or p-values are published.
- New England Vegetable Management Guide, garlic. The regional authority for Massachusetts. Its garlic section carries no citations, which is why several of its figures appear here as Consensus or Unsourced rather than Measured.
- Colorado State University Specialty Crops Program. Winter Mulch Study 2002–03; The Garlic Project 2004 (spacing, flame weeding, scape removal, drip irrigation); Garlic Fertilization and Foliar Feeding 2005. Replicated trials with published significance statements and no published yield numbers. The live pages now return 404 to a direct request.
- UMass Extension. Garlic harvest, curing and storage. Source of the ten to fourteen day curing window and the 40 to 50°F sprouting range. No references are given on the page.
- UC Davis Postharvest Technology. Garlic. Gives the sprouting range as 41 to 65°F, the humidity band and the respiration figures. Its reference section reads that references from scientifically validated sources will be added in the future.
- Harris, L.J. (2016). Garlic: safe methods to store, preserve, and enjoy. UC ANR Publication 8568. The origin of the nine months against six months softneck storage claim. Its three references are two garlic-in-oil botulism papers and the UC Davis page above, which contains neither the comparison nor a citation.
- Cornell Cooperative Extension. White rot: a returning problem for garlic in New York. Confirmed presence across eastern, central and western New York in 2016.
- Stewart-Courtens, C. (2018). Early season garlic fertility. Cornell Vegetable Program. Asserts that late nitrogen has little effect on final bulb size, with no trial design, replication or statistics reported.
- SARE Northeast project ONE11-149 final report, New York 2011–2012. The 30 percent to 9.8 percent positivity change, and the finding that bloat nematode did not overwinter in moist New York soil across two years and three sites, which contradicts the standard survives-for-years line.
Grower catalogs
- Filaree Garlic Farm, Washington
- Hood River Garlic, Oregon
- Boundary Garlic Farm, British Columbia
- Keene Organics, Wisconsin
- Territorial Seed, Oregon
Extension