Introduction

The potato crop of commerce is a heterozygous tetraploid so does not “breed true” as botanical seed propagules. In traditional breeding programs one selects a single elite seedling which then must be propagated clonally as much and as long as there is a market for that cultivar. Thus, every year in the USA, about 6% of the potato crop is devoted to use as tuber “seed” that will be stored and planted the following year. In contrast to dried seeds of grains, for example, the seed and the commercial crop of potato is about 75% water, so is more vulnerable to freezing temperature-induced damage and must be carefully stored to keep a constant supply for year-around marketing of fresh and processed potato products, as well as propagules for the next season’s crop.

Tissue suffers catastrophic damage from ice crystals, but water can stay liquid to temperatures well below 0C (supercool) when there is no initiation of crystals (nucleation). This is a well-recognized and investigated mechanism of survival in overwintering plants. The exact limit of supercooling can be easily measured because a burst of heat (low temperature exotherm, LTE) occurs when ice forms. LTEs are influenced by cooling rates and the duration of exposure to freezing temperatures (Burke et al. 1976; Levitt 1980).

The study of freezing tolerance of potato tubers has a long history. Wright and Peacock (1929) simply spread tubers on the floor of a room cooled to 25F, tossed them in a basket to stimulate freezing, spread them out again, and observed damage, emergence and yield when planted. They reported that previous work and their own shows that cultivar tuber tissue generally experiences lethal freezing at just one to two degrees below zero C, and damage may occur if the tissue is cooled for a longer time and to lower temperature. Boydsten et al. (2006) provide a good review of the physiology, influential factors and metrics of freezing damage of potato tubers. In agreement with previous results, they found potato tubers of various cultivars are able to supercool to about -1C but freeze at temperatures around -2C. For example, they found that tubers exposed to -2C for 12 h were killed and did not produce viable sprouts. We detect no previous published screening on tubers of wild potato relatives, although cold tolerance of the above-ground parts of exotic Solanum species germplasm has been studied for many years with a view to breeding, being more relevant to growing a crop successfully in frost-prone regions (Palta et al. 1997).

Survival of potato tubers at freezing temperatures has several important implications. Freezing injury to tubers makes it more difficult to maintain processing quality and avoid rots in storage (Bethke and Fishler 2018). Cold storage in the absence of tissue damage is a well-known way to maintain dormancy. In the U.S. Potato Genebank, standard clonal maintenance and distribution of germplasm is in the form of in vitro shoots (Bamberg et al. 2016a). However, storing germplasm in the form of tubers also has applications. We previously reported dormancy of over 10 years for Solanum jamesii tubers stored at 6C (Bamberg 2010), so survival at below-freezing temperatures might maintain dormancy and viability of tubers even longer for germplasm purposes. After harvest, some tubers are inevitably left in the soil. If tubers are not killed by winter frost, they will grow to become weeds in the subsequent crop, carrying over disease and pest concerns into the next season (the major theme of Boydsten et al. 2006). Finally, knowing how tuber freezing tolerance gives fitness in the wild could provide insights into species distribution and genetic diversity there. There may also be related archeological implications to tuber survival at freezing temperatures. For example, if ancient inhabitants of the southwest USA who cultivated jamesii (Louderback and Pavlik 2017) actively moved it north as far as southern Utah and Colorado, did tuber survival at freezing temperatures make that possible?

Materials and Methods

General Methods of Tuber Generation and Handling

Most wild potato species do not tuberize in the Wisconsin field season, being adapted to short days, so most tubers for testing were generated in greenhouse pots at the Genebank during the fall/winter using standard cultivation methods. Tubers were held after harvest for at least 3 mo at 6C in paper bags in a walk-in tuber storage. Wild species tubers are naturally numerous and small (~10 mm diameter). Four tubers from each population were placed in each well of 50-well plastic trays of the kind typically used for seedling transplants, (4.4 cm diameter and 5.7 cm deep, Landmark Plastic, Akron Ohio). Replicate trays for exposure to progressively colder temperature were nested and enclosed in a plastic bag to retain moisture. A custom cooling box with 5 cm polystyrene walls was constructed to fit inside a standard household chest freezer (Frigidaire, 15 cf). A digital thermostat (Johnson Controls model A421) regulated the temperature inside the freezer, and a thermograph (Dickson Model KT8P2, Addison IL) confirmed slow, smooth equilibration to target temperatures inside the polystyrene box where tubers were present. Progression of increasing cold was applied by adjusting the thermostat down by hand very slowly to reflect the gradual natural cooling of soil. A control replicate was held in the treatment chamber only to 0C. After being at the target temperature for the planned duration (usually 1 wk per degree C), a single tray was removed from the freezer and returned to 6C storage for an additional week, then placed on shelves at room temperature. Thus, cold exposure was cumulative, such that tubers experiencing a week at -5C, for example, had also experienced a week at -4C and each of the warmer temperature increments. After the freezing challenge, tubers were monitored weekly. Tubers killed (leaking, soft, shriveling, rotted) or viable (sprouting) were recorded and removed until all tubers were accounted for.

General Themes of the Numerous Small Exploratory Studies

  1. 1.

    A preliminary survey of 25 diverse potato species to compare S. jamesii tuber survival to that of other wild and cultivated potatoes.

  2. 2.

    A systematic test of jamesii compared to a related wild species and cultivated S. tuberosum controls.

  3. 3.

    Tests of the effect of soil on tubers by comparison of tubers dusted with dried soil versus clean tubers to determine if artificial growing conditions in soil-less medium in greenhouse pots confers cold survival (perhaps due to the absence of ice-nucleating particles).

  4. 4.

    Tests of effects of ploidy comparing tubers of jamesii artificially raised to the tetraploid level versus tubers from plants of natural diploids.

  5. 5.

    Tests looking for differences among 128 jamesii populations grown as field and greenhouse replicates.

  6. 6.

    Tests to determine the limit of jamesii tuber viability with respect to the degree and duration of cold exposure, including response of F1 hybrids of jamesii and a related wild species and other species for comparison.

  7. 7.

    Tests of tuber tissue death by in vitro culture and tetrazolium when tubers were firm and otherwise appeared viable after long exposure to very low temperatures, but failed to sprout.

  8. 8.

    Tests of the mechanism of tolerance done by precisely-controlled cooling and sensors that detected exotherms when ice formed.

Details of Materials and Methods

1. Preliminary survey of Solanum species. A standard spectrum of 25 representative potato species, the Genebank’s “Mini-core” collection was used for initial surveys in 2015 and 2016. Three of these species are cultivated, the balance wild. Each species was represented by three populations. Results of initial species survey in 2015 at -1C to -4C showed diploid wild species jamesii tubers to be unaffected, so the replicate in 2016 was challenged at -2C to -6C. See Hardigan et al. (2015) and (GRIN 2019) for detailed identities of these Mini-core materials. Tubers were held for 1 wk at each degree C. No statistical analysis was done, but results of the number of tubers that sprouted from each year’s trial for each species at each level of cold were put in a single chart to visualize trends.

2. Systematic test of jamesii compared to a related wild species and cultivated S. tuberosum controls. A more systematic focus on S. jamesii was also undertaken in 2015, comparing a single genotype from each of eight different jamesii populations (PI 673353, 673354, 673355, 673356, 673357, 673360, 673364, 673366) to the related species S. cardiophyllum (PI 558041), and a diploid cultivated S. tuberosum as control (cultivar haploid US-W 0004 = Genebank accession number GS 2), cooling all to -7C. Each replicate flat had tubers in 10 wells for cardiophyllum and tuberosum, and triplicate 10-well representation of each jamesii at each temperature. Tubers were held for 1 wk at each degree C. Two-factor (temperature and species) ANOVA was calculated on over-all species average sprouting per temperature, and also separately for the eight jamesii populations.

3&4. Test of soil-dusted tubers and tetraploidy. As noted in the introduction section, tissue can supercool without ice nucleation. In 2016, to test whether freezing survival of greenhouse tubers grown in artificial potting mix was an artifact of being too clean, we moistened tubers and dusted one treatment with powdered field soil. This clean/dirty tuber freezing test on diploid jamesii PI 673357 included a ploidy treatment using tubers of an artificial chromosome-doubled tetraploid line of jamesii (origin PI unknown). In light of the apparent general survival of jamesii tubers to -7C previously observed, this trial progressed to a much colder -14C. Each temperature treatment of jamesii ploidy with or without soil was represented by five wells in each flat, allowing ANOVA of two factors (temperature and tuber treatment) and their interaction. Tubers were held for 1 wk at each degree C.

5&6. Test comparing field-grown and greenhouse-grown tubers of 128 jamesii populations, including test to detect the limit after which tubers would no longer sprout, and including jamesii interspecific hybrids and other wild species. In 2017 and 2018, we sought to look for differences within jamesii and possible interaction with growing environment. We produced a uniform generation of tubers of all 128 jamesii populations from the Genebank (see Bamberg et al. 2016b) in a greenhouse. The germplasm in the Genebank had been sourced over many years from its natural habitat in the southwest USA states of Arizona, New Mexico, Texas, Colorado and Utah, where it is adapted to emerge in response to monsoon rains in July, and tuberize into the late fall. In 2016, we produced tubers of the same set of 128 populations in an irrigated field at the New Mexico State Agricultural Research Station near Farmington New Mexico, in the soil and climate in which jamesii evolved. In this 2017 trial, tubers were evaluated after cooling to -7C. The experiment was repeated in 2018 held at 0C, -3C, -5C, -10C, -12C, -15C for 1 wk, with cooling at the rate of one degree per day per degree C between these levels, and an additional treatment held at -15C for 3 mo (-15Clong).

The trial in 2018 also used greenhouse-grown tubers of other potato species observed to be most cold tolerant in the initial 2015 survey of the Mini-core as controls. It also incorporated tests of hybrids of jamesii and cardiophyllum tested in 2015. Two factor ANOVA was calculated to assess effects of temperature and tuber pedigree. A single coefficient for each population was calculated that weighed percent sprouting with degree of cold challenge by multiplying the degrees below C by the percent tuber survival, and summing over all temperatures. These coefficients were evaluated with Tukey’s test (Steel and Torrie 1980).

7. Tests of tuber tissue death by In Vitro culture and tetrazolium when tubers were firm and otherwise appeared viable after long exposure to very low temperatures, but failed to sprout. The eyes of firm but non-sprouting tubers of -15Clong treatment were examined microscopically, their flesh was observed for natural color and after tetrazolium assay (Sacher and Iritani 1982). Tuber tissue was also placed in vitro on callus induction media to nurse any growth that would be evidence of life.

8. Test of the co-occurrence of ice formation and tuber death. Differential thermal analysis (DTA) was used to detect super-cooled water in tubers. DTA was performed with S. tuberosum cultivars Russet Burbank and Snowden, and jamesii tubers using a modification of the combined methods of Mills et al. (2006) and Einhorn et al. (2011). Thermoelectric modules (TEMs) (model HP-127-1.4-1.5-74, TE Technology, Traverse City, MI) were used to detect the exotherms. TEMs were placed in individual hinged tin-plated steel containers lined with 5 mm thick pieces of open-cell foam to reduce air turbulence. Ten TEM units were evenly spaced and attached to each of four 30 × 30 cm perforated aluminum sheet pieces (hereafter called “trays”), and wired to a single 24-pin D-sub connector. A copper-constantan (Type T) thermocouple (22 AWG) was positioned on each of two of the trays to monitor temperature in proximity to the TEM units. Trays were positioned vertically in a Tenney Model T2C programmable freezing chamber (Thermal Product Solutions, New Columbia, PA) and connected to a Keithley 2700-DAQ − 40 multimeter data acquisition system (Keithley Instruments, Cleveland, OH). TEM voltage and thermocouple temperature readings were collected at 15-s intervals via a Keithley add-in in Excel (Microsoft Corp., Redmond, WA).

Tubers selected for DTA were approximately 8–15 mm on their longest axis for both the tuberosum cultivars and jamesii. Each tuber was wrapped in a small piece of moist Kim Wipe tissue (to provide a source of ice nucleation) and then aluminum foil. Single wrapped tubers were placed inside the units in direct contact with the TEMs. Each of two freezing runs consisted of a temperature ramp from room temperature to 4C, followed by a hold for 0.5 h to ensure equilibration of the system. The freezing program continued with a ramp to -25C at a rate of 1°C/h.

RESULTS and DISCUSSION

1. Preliminary survey of Solanum species. Table 1. summarizes the results of screening the Mini-core spectrum of species to -4C (2015) and -6C (2016). Control tubers cooled to 0C for 1 wk are not shown since all remained firm and sprouted. S. jamesii is shaded to emphasize that only this species had most tubers of all populations survive at the coldest temperatures in both years.

Table 1 Survey of 25 potato species. Data given is the number of three maximum populations with any observed tuber sprouting after the indicated temperature challenge
Full size table

2. Systematic test of jamesii compared to a related wild species and cultivated S. tuberosum controls. Fig. 1. presents the results of comparing cold response of tubers of species jamesii and cardiophyllum, one of the few other potato species that hybridize easily with jamesii, with diploid tuberosum (cultivated) included for comparison. The effects of temperature, species, and interaction were highly significantly different. Differences in cold response within the jamesii individuals were also highly significant (data not shown), but since replicates were from a single tuber lot, the differences could have been due to how tuber lots were handled rather than genetics. Replicated tuber lots were examined in tests 5&6 below in an effort to detect differences between jamesii populations.

Fig. 1
figure 1

Superior survival of S. jamesii tubers after low temperature challenge

Full size image

3&4. Test of soil-dusted tubers and tetraploidy. When tubers of jamesii at the natural diploid or artificial tetraploid levels were compared with or without dusting with soil to promote ice nucleation, the effects of tuber treatment and temperature were highly significant as was the interaction. Fig. 2. shows that the significant cold tolerance effect associated with tuber treatment is wholly due to being diploid, not being clean. This suggests that the survival of jamesii tubers after exposure to freezing cold temperatures does not depend on being artificially grown in potting medium in a greenhouse.

Fig. 2
figure 2

Survival of 4x and 2x jamesii tubers, clean and dirty, after cold challenge

Full size image

5&6. Test comparing field-grown and greenhouse-grown tubers of 128 jamesii populations and limits of survival, including hybrids and other species for comparison. Tubers of 128 populations of jamesii were generated both in the Genebank greenhouse and irrigated field at Farmington New Mexico in 2016 and tested in 2017 to -7C. In light of jamesii being a very heterogeneous species (Bamberg et al. 2015) and tetraploid jamesii tubers being cold sensitive, we hoped cooling tubers to -7C would reveal effects of population or growing environment—that is, reveal the existence of some jamesii populations with tuber cold sensitivity. But essentially 100% of tubers of all populations from both locations sprouted after exposure to -7C (data not shown), providing no basis to discriminate populations.

The comprehensive test on jamesii populations of 2017 was repeated in 2018 with the same tubers lots to lower temperature (to -15C for 1 wk as in 2017, but also held for 3 mo at -15C, denominated as -15Clong). This trial also incorporated several greenhouse-grown control tuber types, re-testing a selection of Mini-core populations that showed the most cold tolerance in 2015, and hybrids of jamesii and cardiophyllum.

As suggested by Fig. 3., when the mean performance of each of the four types of treatments were compared by a single coefficient weighting sprouting percent with all levels of cold challenge, and evaluated with Tukey’s test, the highly significance (p = 0.01) relationships were as follows: (Greenhouse jamesii = Field jamesii) > (jamesii-cardiophyllum hybrids = Mini-core species).

Fig. 3
figure 3

Average percent sprouting after cold treatment: S. jamesii tubers generated in the greenhouse versus New Mexico field. Hybrids of jamesii with the related species S. cardiophyllum and the most hardy species from the Mini-core set (see Table 1.) included for comparison

Full size image

Note that the cold challenge of -15Clong is arbitrarily graphed at a position on Fig. 3 at three degrees C colder than −15. The coefficient for this point converges at zero for all types of tubers tested, since all treatments had 0 % sprouting at -15Clong.

We also inspected the data and found that -15C was the temperature at which the greatest variability in sprouting among populations was observed for both greenhouse and field tubers, thus presumably providing the best opportunity to distinguish populations. But when ANOVA was calculated for sprouting at just the -15C treatment, the effect of populations was still not found to be significant. However, at this coldest non-lethal temperature (-15C) tubers from the field had highly significantly more sprouting than their counterparts from the same population grown in the greenhouse (Fig. 3). Note that this also supports the conclusion of the previous experiment that extreme cold hardiness of jamesii tubers does not depend on being grown in soil-less medium.

Figure 3. illustrates that at -10C, average sprouting of species other than jamesii was 10% or less. These Mini-core populations, when analyzed separately, were slightly (but significantly) different, with the best sprouting observed in populations of acaule, infundibuliforme, kurtzianum, and okadae, generally fitting the pattern observed in the initial Mini-core screening of 2015 (Table 1.). The uniform cold sensitivity of hybrids between the very sensitive cardiophyllum and jamesii suggests tuber cold tolerance is not a simple or dominant trait.

7. Test to assess viability of ostensibly normal-looking tubers that were firm but did not sprout after extreme cold challenge. Regardless of growing environment, 50–60% of jamesii tubers cooled to -15C survived and sprouted, while virtually no other species’ tubers did (Fig. 3). After being held for 3 mo at -15C (and the 15 weeks of accumulated time and increasingly colder temperatures preceding), no jamesii tubers sprouted, but 10% of those from the field, and 40% from the greenhouse were firm. Almost no tubers of any other species were even firm after 1 wk at -15C. Under the microscope, unsprouted eye meristems of firm jamesii tubers from the -15Clong treatment looked discolored and soft compared to those of control tubers. Tuber flesh of unfrozen jamesii is very white, but -15Clong tuber flesh appeared slightly discolored (Fig. 4a). Tetrazolium clearly stained tubers that had not been subjected to cold, but not those firm tubers from -15Clong treatment (Fig. 4b). After 6 mo on callus induction media in vitro, no -15Clong tuber tissue explants showed any growth, despite looking essentially normal.

Fig. 4
figure 4

a & b. Appearance of jamesii tuber flesh of firm but non-sprouting tubers after long exposure to -15C compared to non-frozen control tubers. Natural and after 24 h in tetrazolium stain or water control

Full size image

8. Test of the co-occurrence of ice formation and tuber death. DTA revealed that only jamesii tubers showed low temperature exotherms (LTE) (Fig. 5). No LTEs were detected in any of the tubers of Russet Burbank and Snowden. A single exotherm for cultivars at about -7C represents freezing of the tuber tissue at about the same time as the moist paper wrapping used to provide external nucleation (Fig. 5a). In contrast, S. jamesii tubers had the same exotherm when their papers froze, but the second large exotherm marking the freezing of the tuber tissue did not occur until reaching at least six degrees colder (Fig. 5b).

Fig. 5
figure 5

a & b. Examples of different traces of DTA for Solanum tuberosum (a) and Solanum jamesii tubers (b)

Full size image

The temperatures at which LTEs were detected in jamesii varied from -7C to -18C (Fig. 6). Variation in LTEs was broad among tubers, genotypes and growing environments, but the results of DTA experiments show that jamesii tubers are capable of supercooling to very cold temperatures without killing ice formation.

Fig. 6
figure 6

a & b. Summary of temperatures at which low temperature exotherms were detected in different genotypes of Solanum jamesii. Each bar represents a separate tuber produced in the greenhouse at the Genebank (a) or Biotron controlled environment facility (b) at the University of Wisconsin, Madison

Full size image

Conclusions and Future Work

This work demonstrated that S. jamesii tubers survive much colder temperatures than those of 25 other representative potato species. This hardiness was observed in all of many jamesii populations tested, and was independent of greenhouse or field growing conditions. Tubers of jamesii artificially raised to tetraploid level did not have hardiness, nor did interspecific hybrids. The mechanism of hardiness appears to be the ability to prevent ice formation even far below zero C. The limit of survival was surpassed after many weeks at -15C, when sprout meristems were killed, tuber tissue was discolored and no longer responded to tetrazolium vital stain, although it did not become flaccid as commonly observed in tubers of cultivated potato.

Wild species related to the potato crop often have extreme expression of traits with practical value for breeding. To the other remarkable traits of Solanum jamesii (Bamberg et al. 2017), we have here added evidence of this species’ unusual ability to produce tubers that remain viable after exposure to surprisingly low temperatures by supercooling to avoid ice formation.

Future work could examine artificial hybrid families segregating for the trait to determine the genetic, physiological and anatomical basis. The interaction between temperature level and exposure time could be precisely characterized, for example, determining whether tubers held at just a few degrees below zero remain undamaged and viable for a very long time. The impact of different environmental factors during tuber development could also be studied.

All germplasm described here is freely available from the Genebank for these and other studies, as well as enhancement and breeding efforts aimed at incorporation of the trait into the potato crop.

Acknowledgements and Perspectives

We thank Max Martin and Renee Sauer for technical inputs and collecting the data, as well as the UW Peninsular Agricultural Research Station program and staff for their assistance. During germplasm collecting expeditions, one Colorado landowner told us that her father intentionally collected and exposed jamesii tubers to freezing temperatures overnight and found they were unharmed. A rancher in New Mexico told us that in years past, water pipes buried deep underground sometimes froze, proof that the jamesii thriving in the area must have had tubers quite invulnerable to hard frost. It also seems likely that the ancient native inhabitants of the southwest USA who used jamesii for food knew that the tubers survived freezing. Our publishing of experimental results of cold tolerance of jamesii tubers here is “new” in terms of being systematic, conclusive and quantitative. But we acknowledge these others who had the idea earlier as casual observations and popular wisdom, and recognize the value of listening to their stories.