Storability of Irish Potato (Solanum tuberosum L.) Varieties Grown in Kenya, Under Different Storage Conditions

Long-term use of fresh potatoes is constrained due to post-harvest losses and limited or poor storage systems. Loss of potato quality during storage is mostly due to weight loss, excessive sprouting, decaying, greening, pest attack and changes in sugar content which is a major concern for processors. The storability of three potato varieties namely Shangi, Unica, and Dutch Robijn was evaluated under four storage conditions: (1) room temperature (RT) (21.7 ± 5 °C) and ambient relative humidity (RH) (73.5 ± 6.7%); (2) 10 °C/75% RH; (3) 10 °C/ambient RH and (4) 7 °C/75% RH for 3 months. Parameters analysed included weight loss, sprouting, greening, rotting incidence and changes in simple sugars (sucrose, glucose and fructose) content. Shangi variety had the poorest storability having exhibited the highest weight loss, sprouting, greening and rotting rates among the varieties and in all storage conditions. In all the varieties, weight loss was highest at room temperature/RH (6.9–35.13%) and lowest at 7 °C/75% RH (1.6–3.7%). Sprouting rate was highest at RT/RH (100%) and lowest at 7 °C/75% RH (2.0–45.2%). Greening incidence was highest at RT/RH (10.3–24.0%) and lowest at 7 °C/75% RH (2.0–16.1%). The tubers accumulated simple sugars during storage with the highest relative accumulation recorded at 7 °C/75% RH. Total reducing sugars increased by 3.8–9.3-fold at RT/RH to 26.0–68.5-fold at 7 °C/75% RH. Low-temperature storage minimized the rate of physical tuber deterioration with 7 °C/75% RH best preserving the potatoes’ physical properties while it led to the highest accumulation of simple sugars.


Introduction
Potato (Solanum tuberosum L.) is currently the world's largest non-cereal crop and ranks fourth as the most important crop in the world after maize, wheat and rice in terms of area covered and production (Murigi 2016). The average potato production globally is estimated at 359 million metric tonnes per annum (FAOSTAT 2022). Potatoes are cultivated in more than 80% of countries across the world, with more than 1.3 billion people worldwide consuming them as a staple food (Devaux et al. 2021). Global statistics show that the production of potatoes is shifting towards developing countries, with significant growth in production observed mainly in Asia and Africa (especially in East Africa) (Devaux et al. 2021). This growth attests to the significance of potatoes in meeting the nutritional and economic needs of the population in developing countries. In Kenya, potato is the second most important crop after maize in terms of production and consumption and plays an important role as a food and cash crop (International Potato Center 2019). The crop has a great potential for food security given its high productivity per unit area and time (Hussain 2016). It also plays a big role in poverty alleviation by providing a source of livelihood to farmers and people employed in the potato value chain (Murigi 2016). Post-harvest management of potatoes is therefore critical in not only enhancing food and nutrition security but also preserving livelihoods. Globally, 70% of potatoes after harvest are stored for different periods of time depending on the demand from processors and domestic consumers (Benkeblia et al. 2008). Usually, potatoes are stored as fresh produce in a perishable form.
Average post-harvest potato losses across the world are estimated to be at 10-15% annually. However, in developing countries, losses can be as high as 30% due to insufficient adoption of efficient storage practices (Benkeblia et al. 2008). According to Kaguongo et al. (2014), post-harvest loss of potatoes in Kenya is estimated at 12.8% at the farm, 24.4% at the open market, 12% at the processing level and 25% in supermarkets. Approximately 19% of potato is lost per hectare production during each harvest season, which corresponds to 42,824 Kenyan shillings (KES) loss per hectare. This translates to approximately 815,000 t of damaged produce annually when extrapolated to the national level, which represents a value of about KES 12.9 billion (Kaguongo et al. 2014). Some of the major causes of potato quality loss include tuber weight loss, sprouting, greening and rotting.
Potatoes in Kenya are not commonly stored for future use, thus leading to seasonal fluctuations in stocks (Food and Agriculture Organization 2013). Due to a lack of knowledge and efficient storage systems, most farmers in Kenya sell their potato produce immediately after harvest or harvest their potatoes only upon identifying buyers (Nyankanga et al. 2018a, b). Smallholder farmers store potatoes in poor storage facilities, resulting in losses through decay, sprouting and greening among others (Food and Agriculture Organization 2013). Processing industries and restaurants also usually need to store potatoes for short periods prior to processing (Food and Agriculture Organization 2013). Almost a quarter of the potatoes placed in Kenyan markets are damaged or turned green (Musita et al. 2019). This raises concerns over food losses and also the safety of the potatoes to consumers due to the possible presence of glycoalkaloids, compounds that cause toxicity in humans if consumed in large amounts (Musita et al. 2019). Sprouting contributes to weight loss, reduces marketability of tubers and diminishes the nutritional and processing quality of the tubers (Gumbo et al. 2021). Consumers largely assess the quality of ware potatoes by visual appearance which includes tuber firmness and absence of rot, greening and visible sprouts (Murigi 2016). The main factors that directly affect the quality and storage life of potatoes include temperature, relative humidity (RH), light and potato variety (Pinhero and Yada 2016).
Post-harvest potato storage practices across the world include on-farm methods such as the use of clamps and pit storage, use of sprout suppressants, controlled atmosphere storage, cold storage (2-6 °C) and evaporative cooling storage (Eltawil et al. 2006;Kibar 2012;Marwaha et al. 2005;Wasukira et al. 2017). Potatoes stored at temperatures higher than 8 °C such as those in pit storage and clamps have to be treated with a sprout suppressant such as isopropyl N-(3-chlorophenyl)carbamate (CIPC), 1,4-dimethylnaphthalene and peppermint oil to minimize sprouting. It has been reported, however, that the efficacy of such sprout suppressants is highly influenced by storage temperatures (Murigi 2016;Paul et al. 2016). Additionally, in recent years, there have been increasing health and environmental concerns over the use of chemical sprout inhibitors, spurring an increase in the use of safer nonchemical options (Paul et al. 2016;Pinhero et al. 2009). For instance, the European Union Commission banned the use of CIPC in all EU member countries following concerns over chronic consumer risk (Arici, 2019; CERTIS/ACETO/UPL 2019). Given the increasing rate of potato consumption in developing countries such as Kenya, this measure is also likely to be imposed. Controlled atmosphere storage is not well accepted yet as conflicting findings have been reported on its effects on the sprouting and sugar levels of potatoes. This technique is also expensive to install and maintain in practical terms (Daniels-lake 2013; Gottschalk 2011; Thompson 2010). Although cold storage has been promoted as one of the most effective techniques for inhibiting tuber deterioration during storage, some varieties accumulate simple sugars in a process called cold-induced sweetening (CIS). However, storage at temperatures above 6 °C has been reported to minimize the accumulation of simple sugars. Additionally, reconditioning potatoes intended for processing is reported to reduce simple sugar content prior to processing (Marwaha et al. 2005;Pereira et al. 2021;Wayumba et al. 2019).
Although a lot of research has been done in the temperate regions on the effect of different storage practices on potatoes, there is limited information on the storability of specific potato varieties particularly under cold storage in tropical environments like Kenya. This study aimed to evaluate the storability of three potato varieties (Shangi, Dutch Robijn and Unica) grown in Kenya kept under different storage conditions. Unica is a fairly new variety with promising processing qualities while Shangi and Dutch Robijn are the most popular varieties for French fry and crisp processing in Kenya (National Potato Council of Kenya 2017). Information on the storability of these varieties under different conditions will be valuable to farmers, processors and primary consumers in terms of selecting appropriate storage depending on the intended end use.

Sample Acquisition and Preparation
Shangi, Unica and Dutch Robijn varieties were cultivated under uniform agronomic conditions and cultural practices on a single farm in Nyandarua County. The potatoes were grown between October and January 2020 during the short rains, following standard practices including ridging, fertilization, weeding and pest control. Diammonium phosphate fertilizer (DAP) was applied during planting. The potatoes were hand-harvested after 3 months, a period after which they are considered mature and ready for the market in the region. The tubers were transported in nylon sacks at room temperature to Jomo Kenyatta University of Agriculture and Technology. The tubers were kept in a dark room and left to cure for 7 days at room temperature. Medium-sized (length 55-95 mm), undamaged and apparently healthy tubers were selected for experimentation. In triplicates, the tubers were apportioned into approximately 3-kg lots and placed in incubators (Low Constant Incubator Model SHP-150, China) at different temperatures. The storage conditions were as follows: RT/RH (21.7 ± 5 °C/73.5 ± 6.7%), 10 °C/75% RH, 10 °C/ ambient RH and 7 °C/75% RH. The physico-chemical properties of the tubers were evaluated at 2-week intervals for Unica and Dutch Robijn and weekly intervals for the Shangi variety (because it was observed to deteriorate more rapidly than the rest) up to 13 weeks (91 days). Parameters analysed in storage included physical characteristics (weight loss, sprouting, greening and rotting incidence) and changes in simple sugar concentration (glucose, fructose and sucrose).

Weight Loss Determination
Tuber weight loss was determined as described by Abong et al. (2015). Tubers were weighed at the beginning of the experiment and at subsequent intervals of 1 week for Shangi and every 2 weeks for Unica and Dutch Robijn using an electronic scale (KERN Pls 1200-3A, China). Weight loss was calculated as the change in the subsequent weights at the evaluation intervals relative to the initial weight of tubers. The change in weight was expressed as a percentage of the original weight.

Sprouting Determination
Tuber sprouting was evaluated as described by Murigi (2016). Tubers were considered sprouted if they had at least one visible sprout of at least 3-mm length. The number of sprouted tubers was counted per sample batch at 1-week intervals for Shangi and 2-week intervals for Unica and Dutch Robijn for 91 days. Percentage sprouting was calculated as the number of sprouted tubers relative to the total number of tubers per sample.

Greening Determination
Evaluation of tuber greening was conducted as described by Murigi (2016). Tubers were evaluated visually by observation, and those displaying any signs of a green coloration were considered to have greened. Greening was evaluated at 1-week intervals for Shangi and 2-week intervals for Unica and Dutch Robijn for 91 days and expressed as a percentage of all the greened tubers per sample batch.

Rotting Determination
Rotting was evaluated by visual observation as described by Nyankanga et al. (2018b) for 91 days. Each tuber was observed for any sign of dry and soft rots every 2 weeks for Unica and Dutch Robijn and weekly for Shangi variety. The number of affected tubers per sample batch was recorded, and rotting incidence was calculated as the percentage of all the rotten tubers per sample batch.

Evaluation of Simple Sugars
Glucose, fructose and sucrose concentrations were analysed following the method described by Abong' and Kabira (2011). Five grams of potato tubers freshly homogenized with a blender was weighed into 50-ml pear-shaped flasks. Ethanol (20 ml) was added to the flasks containing samples and swirled to mix. The mixture was refluxed at 100 °C for 1 h after which the contents were filtered to obtain the filtrate. The filtrate was evaporated at 80 °C to dryness on a rotary evaporator after which a 2-ml solution of acetonitrile and distilled water in the ratio of 1:1 was used to reconstitute the dried sample. The solution was passed through 0.45-μm pore size microfilters to eliminate any debris before injecting 20 µl into a Shimadzu Nexera Ultra-Flow Liquid Chromatography machine (UFLC, Japan) fitted with a SIL-20A HT prominence autosampler, a 20-A refractive index detector and an LC-20AD pump. The chromatographic conditions entailed an isocratic elution of water and acetonitrile (25:75) pumped through a normal-phase Ultisil NH 2 column with a 6 × 250 mm internal diameter at a flow rate of 1.8 ml/min. The column oven (CTO-10ASvp) temperature was set at 40 °C. Glucose, fructose and sucrose were quantified by comparing against their corresponding standard solutions and expressed as mg/100 g fresh weight. The relative sugar concentration was calculated as follows: RC relative sugar concentration (concentration of sugar in proportion to the initial concentration). RC = S1 − S2 S1 S1 initial sugar concentration (before storage).
S2 subsequent sugar concentrations during storage.

Data Analysis
All analyses were conducted in triplicate, and data was expressed as means and standard error of the means. One-way analysis of variance was carried out using STATA software version 12. Mean differences among the treatments were separated by Bonferroni's means separation test at 5% level of significance.

Weight Loss
The rate and extent of weight loss were highest in potatoes stored at RT/RH followed by those stored at 10 °C/ambient RH, 10 °C/75% RH then 7 °C/75% RH as shown in Fig. 1. At RT/RH, there was a significant varietal difference (p < 0.0001) with

Fig. 1
Relative weight loss in three potato tuber varieties stored in different temperature and RH conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: relative humidity of ambient air (not determined) Shangi recording the highest weight loss (35.13%) followed by Dutch Robijn (7.8%) then Unica variety (6.9%) after 3 months of storage. Storage at 10 °C/ambient RH led to a 9.0% weight loss in Shangi, 4.4% loss in Unica and 3.7% in Dutch Robijn with a significant varietal difference (p < 0.0001) being observed in the weight loss. At 10 °C/75% RH, tubers exhibited a lower rate of weight loss with Shangi portraying a 5.2% loss while Dutch Robijn and Unica both exhibited a 1.8% weight loss. At 7 °C/75% RH, the varieties differed significantly (p < 0.0001) with only 3.7% loss observed in Shangi, 1.8% in Unica and 1.6% in Dutch Robijn at the end of the experiment. Generally, the Shangi variety exhibited the highest and fastest loss of weight across all the storage conditions which indicates a short dormancy period. Tuber weight loss reduced with reduction in storage temperature, and storage at 7 °C/75% RH preserved the tuber weights best. Weight loss at 10 °C/75% RH and 10 °C/ambient RH differed significantly (p < 0.0001) suggesting that RH might have a substantial effect on tuber weight loss. High RH has been reported to minimize wilting, shrivelling and weight loss in fruits and vegetables by reducing their rate of transpiration (Aharoni et al. 2007). Transpiration and sprouting account for almost 90% of the decrease in physiological weight of potato tubers while respiration accounts for 3-10% of the weight loss (Murigi 2016). The high rate of weight loss at room temperature and RH was attributed to evaporative losses through the tuber surface and sprouts and was exacerbated by tuber rotting. Similar observations were reported by Valencia Flórez et al. (2019) whereby tubers stored at 18 °C presented the highest weight loss compared to those stored at 4 °C.
Variations in weight loss across varieties could be attributed to differences in varietal physiological traits and genotypic differences. Moreover, varieties with longer dormancy periods have been reported to maintain weight better in storage under non-cold storage conditions since sprouted tubers lose more weight than unsprouted ones (Azad et al. 2017). Other factors such as periderm thickness, quantity of lenticels on the tuber surface and the number of cell layers in the periderm may have also influenced tuber weight loss (Ezekiel et al. 2004;Makani et al. 2017).

Tuber Sprouting
The trend in sprouting rate was observed to be highest in tubers stored at RT/RH followed by those stored at 10 °C/ambient RH, 10 °C/75% RH and 7 °C/75% RH in that order as shown in Fig. 2.
At RT/RH, tubers began sprouting on day 14 for Shangi and day 34 for both Unica and Dutch Robijn. The Shangi variety displayed 100% sprouting by day 30 while Unica and Dutch Robijn showed 100% sprouting on days 77 and 91 respectively. A significant difference in sprouting rate was observed between Shangi variety and the rest from day 14 to day 63 (p < 0.0001-0.027). There was no difference in the rate of sprouting among the varieties on day 77 (p = 0.4219) and the last day of storage when all the varieties were 100% sprouted.
At 10 °C/ambient RH, sprouting commenced on day 24 for Shangi and day 63 for both Unica and Dutch Robijn. Dutch Robijn exhibited the least amount of sprouting by the last day of storage (7.1%) followed by Unica (96%) then Shangi (100%).
From day 27, a significant difference (highest p-value < 0.0001) in sprouting rate across the varieties was observed. At 10 °C/75% RH, tubers started sprouting on day 31 for Shangi, day 63 for Unica and day 91 for Dutch Robijn. Only 2% of Dutch Robijn had sprouted compared to 87.5% and 100% of Unica and Shangi respectively. There was a significant difference in sprouting rate among the varieties beginning day 31 of storage when Shangi started sprouting to the last day of storage (highest p-value = 0.0002). Generally, Shangi variety sprouted the highest, while Dutch Robijn sprouted the least across all storage conditions. Similar findings were reported by Abong et al. (2015 andMurigi (2016) whereby the variety Shangi exhibited a significantly higher sprouting rate in different storage treatments than other varieties such as Dutch Robijn, Asante and Kenya Mpya, among others.
At 7 °C/75% RH, tubers began sprouting on day 63 for Shangi, day 77 for Unica and day 91 for Dutch Robijn varieties. A significant difference in sprouting rate among the varieties was observed for both day 77 (p = 0.0001) and the last day of storage (p = 0.0001) when the sprouting rate was 45.2% for Shangi, 32.1% for Unica Fig. 2 Sprouting rate in three potato tuber varieties stored at different temperatures and RH conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: relative humidity of ambient air (not determined) and 2% for Dutch Robijn. Compared to storage at room temperature/RH, the rate of sprouting at 7 °C/75% RH was less by 54.8% for Shangi, 67.9% for Unica and 98% for Dutch Robijn.
Maintaining potato tuber quality during storage is vital to preventing economic losses. In addition to damaging tuber appearance and diminishing marketability, sprouting also contributes to loss of moisture in tubers, weight loss due to respiration, physiological aging and accelerated starch breakdown (Nyankanga et al. 2018a). Furthermore, increased glycoalkaloid levels and reduction in vitamin content have been associated with tuber sprouting (Pinhero et al. 2009). Temperature significantly influences sprouting. Within the temperature range of 3-20 °C, potato dormancy is inversely proportional to temperature (Pinhero and Yada 2016). The variations in the sprouting rate of the varieties in all storage conditions could be due to their genotypic differences. Figure 3 shows the physical appearance of the tubers used in the present study after 91 days of storage.

Tuber Greening
Tuber greening was highest in the Shangi variety followed by Dutch Robijn then Unica in all the storage conditions except at 10 °C/ambient RH as shown in Fig. 4. There was no significant difference in the Greening incidence of Shangi across the storage conditions (p = 0.2869). The highest trend, however, was observed in tubers stored at RT/RH (24.14%) and the lowest at 7 °C/75% RH (16.13%). A significant difference across the storage conditions was observed for Dutch Robijn (p = 0.0018) and Unica (p = 0.0092). Greening in Dutch Robijn was highest at RT/RH (10.29%) and lowest at 7 °C/75% RH 2%. Unica on the other hand showed the highest greening rate (12%) at 10 °C/ambient RH and the lowest (3.61%) at RT/RH storage. Compared to RT/RH, the greening rate in tubers stored at 7 °C/75% RH was lower by 7.9% in Shangi and 8.3% in Dutch Robijn. At 10 °C/75% RH, the rate of tuber greening was less than that in RT/RH conditions by 4.7% in Shangi and 5.7% in Dutch Robijn. This was however not the case for Unica which exhibited a higher greening rate compared to room conditions by 0.73%. For Shangi and Dutch Robijn varieties, there was no significant difference (p = 0.4401 and p = 0.3373 respectively) in the rate of greening between tubers stored at 10 °C/75% RH and those at 10 °C/ ambient RH.  Potato greening has been reported to increase with storage temperature and is lower in cold storage (Tanios et al. 2018). Shangi and Dutch Robijn exhibited this trend while Unica had a converse behaviour. It is unclear why Unica displayed the different trend, and further investigation is necessary to understand this behaviour. Differences in the greening behaviour of the three varieties across the different temperature treatments could be attributed to genotypic variations. Nyankanga et al. (2018b) also reported varietal differences in greening levels in potatoes stored using different packaging materials.
Greening diminishes the quality, marketability and palatability of tubers. Greening occurs mainly because of exposure of potatoes to light resulting in the accumulation of chlorophyll in the amyloplasts (Chang 2013). Chlorophyll accumulation in potatoes due to exposure to light occurs concomitantly with an increase in glycoalkaloid levels (α-solanine and α-chaconine). Glycoalkaloids are toxic compounds that when consumed in excess (> 1 mg/kg body weight) can cause mild to severe toxicity and might negatively affect the taste of potatoes (Olsen et al. 2017;Tsikrika et al. 2019). Although greening and glycoalkaloid formation occur simultaneously in response to light, they are formed through different biochemical processes (Bamberg et al. 2015;Chang 2013). Tuber greening is dependent on pre-and post-harvest factors such as temperature, wounding, soil nitrogen, light exposure and genotype (Grunenfelder 2005). Tuber skin thickness and the presence of accessory pigments are suggested to be linked to resistance of tubers to greening as they influence the amount and quality of light penetrating the periderm (Tanios et al. 2018).

Rotting Incidence
A significant difference in rotting (p = 0.0002) was observed among the varieties in tubers stored at RT/RH conditions. At the end of the experiment, 13.79% of Shangi tubers under RT/RH storage were rotten whereas Unica and Dutch Robijn exhibited no rotting incidence (Fig. 5). There was no rotting incidence observed in tubers stored at 10 °C/ambient RH 10 °C/75% RH and 7 °C/75% RH in all the varieties.
Tuber rotting in Shangi was primarily attributed to the high room temperature and uncontrolled RH. It is recognized that as much as high RH is beneficial in maintaining tuber weight during storage, it increases disease incidence, particularly when accompanied by high temperature (Nyankanga et al. 2018a, b). In addition, rotting was speculated to have been due to soft rot and blackleg, the most common and destructive diseases in potatoes caused by Pectobacterium carotovorum spp. (Pc) and Dickeya spp. (Kamau et al. 2019;Pavlista 2013). The bacteria can survive for long periods in plant debris such as potato tissue without causing disease and become active when the environmental conditions are favourable (Rosenzweig et al. 2016). The rotten Shangi tubers in this study exhibited symptoms of soft rot including water-soaked lesions on tubers that become soft, mushy, disintegrated, black in colour and/or creamy and slimy (Czajkowski et al. 2011;Rosenzweig et al. 2016).
Factors influencing resistance to soft rot include tuber dry matter and sugar content, electrolyte composition, cell turgidity, oxygen levels during storage, water loss, membrane permeability enlarged lenticels, starch content and calcium content. These factors, however, interact in very complex ways depending on the genotypic characteristics of every variety (Chung et al. 2013). Rotting of Shangi compared to the other varieties could be attributed to its high rate of physical and chemical degradation which altered the tuber physiology such as cell turgidity and membrane permeability. The changes may have reduced Shangi's resistance to the soft-rot and blackleg-causing bacteria compared to the other varieties. Nyankanga et al. (2018a, b) reported similar characteristics in potato tubers stored at ambient tropical temperatures in Kenya, and decay was speculated to be due to soft rot. Similarly, a potato disease surveillance in Kenya showed that Shangi variety displayed the highest prevalence of infection by Dickeya and Pectobacterium at (89%) compared to other varieties like Dutch Robijn (9%), Asante (1%), Destiny (1%) and Kabale (1%) (Mulema et al. 2021).
Tuber rotting not only renders potatoes unfit for consumption but also spreads the infection to the adjacent tubers in storage. Storage at 7 °C/75% RH, 10 °C/75% RH and 10 °C/ambient RH completely inhibited rotting in Shangi whereas Unica and Dutch Robijn did not rot at all even at RT/RH, demonstrating varietal influence on tuber rotting during storage.

Fig. 5
Rotting incidence in three Kenyan potato varieties stored under different temperature and RH conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: Relative humidity of ambient air (not determined)

Simple Sugars
In addition to physical tuber characteristics, sugar levels largely determine the processing quality of potatoes during storage. Potato tubers contain simple sugars including sucrose which is the main disaccharide and fructose and glucose which are reducing sugars (Bonierbale et al. 2010). Very high levels of reducing sugars in potato tubers cause excessive browning in fried and baked potato products in a process called the Maillard reaction (Wayumba et al. 2019). Fructose and glucose react with free amino acids at high temperatures leading to the formation of darkpigmented bitter products. This reaction also leads to the formation of a compound called acrylamide, a probable carcinogen Tilahun, 2020). Genetic variations have been reported to have a major influence in the reducing sugar content as well as changes in sugar levels during storage (Abong et al. 2009). In this study, an increase in the concentration of simple sugars in the stored tubers at different rates in the different storage conditions was observed. Tubers stored under low temperature exhibited an increase in fructose, glucose and sucrose concentrations over time, a process known as cold-induced sweetening (CIS) (Bhaskar et al. 2010).

Sucrose
The sucrose concentrations in Shangi, Unica and Dutch Robijn increased in all storage conditions by 1.72-3.32-fold at RT/RH, 6.47-10.33-fold at 10 °C/ambient RH, 6.26-12.63-fold at 10 °C/75% RH and 9.72-15.27-fold at 7 °C/75% RH. There was a significant difference in the relative accumulation of sucrose across the different storage regimes (p < 0.0001 for all the varieties) with the highest rate being observed at 7 °C/75% RH and the lowest rate at RT/RH (Fig. 6). There was also a significant difference in sucrose build-up among the varieties (p < 0.0001 for all storage conditions). However, accumulation of sucrose was 5.8-12.1-fold lower than that of fructose and glucose. This might be because sucrose is broken down to glucose and fructose enzymatically. Our findings for RT/RH storage were in alignment with those reported by Galani et al. (2016) in Indian varieties whereby sucrose content increased by 1.5-5.0-fold at room temperature. Converse to our findings, however, Galani et al. (2016) reported a decrease in sucrose content of different Indian varieties by 1.5-8.9-fold from days 15-30 to day 105 in storage at 4 °C. This disparity could be attributed to differences in varieties studied, agronomic conditions or differences in storage temperature which might influence the activity of enzymes involved in sucrose synthesis and breakdown during cold storage.
During CIS, the carbon fluxes connecting starch with sucrose are affected leading to an elevated rate of sucrose synthesis. The enzymes involved in this process include sucrose phosphate phosphatase, UDP-glucose pyrophosphorylase and sucrose phosphate synthase (Bhaskar et al. 2010). Due to an imbalance between the breakdown of starch and metabolism of sucrose, some sucrose enters the cells' vacuoles. When sucrose reaches the vacuole, the enzyme vacuolar acid invertase (VINV) is upregulated leading to the cleaving of sucrose into reducing sugars (fructose and glucose) (Bhaskar et al. 2010).

Fructose
Fructose levels in the tubers increased in concentration across all storage conditions with the highest rate being recorded in tubers stored at 7 °C/75% RH (43.76to 184.85-fold) followed by storage at 10 °C/ambient RH (27.99-to 120.79-fold), 10 °C/75% RH (36.41-to 117.43-fold) and RT/RH (6.12 to 19.36) in that order. There was a significant difference (p = 0.0016) in the rate of accumulation between Shangi and the other two varieties in tubers stored at RT/RH. Significant differences were also observed in relative fructose accumulation among the three varieties stored at 7 °C/75% RH (p = 0.0046), and 10 °C/ambient RH (p = 0.0243). Shangi generally displayed the highest fructose accumulation rate followed by Dutch Robijn and then Unica across all storage conditions (Fig. 7).
There was no significant difference in the relative accumulation of fructose between the tubers stored at 10 °C/ambient RH and those at 10 °C/75% RH for all the varieties; Shangi (p = 0.6789), Unica (p = 0.5418) and Dutch Robijn p = (0.5317) implying that RH does not have any influence on changes in sugar concentration of potato tubers during storage. Galani et al. (2016) reported a similar trend in the behaviour of fructose: a 0.8-to 2.8-fold and 21.4-to 85.0-fold increase in tubers stored at room temperature and 4 °C, respectively.

Fig. 6
Relative change in sucrose concentration of three potato tuber varieties stored under different temperature and RH storage conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: relative humidity of ambient air (not determined)
Storing potatoes at temperatures below 10 °C has been reported to cause a buildup of sucrose, some of which is broken down to fructose and glucose by the vacuolar acid invertase enzyme (Wiberley-Bradford et al. 2014). Fructose and glucose accumulation, also known as cold-induced or low-temperature sweetening, is attributed to starch breakdown into simple sugars hydrolytically or phosphorytically. It is suggested that CIS occurs due to the restriction of hexose phosphates (products of starch Fig. 7 Rate of change in fructose concentration of three potato tuber varieties stored under different temperature and RH storage conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: relative humidity of ambient air (not determined) degradation from the amyloplasts) from entering the glycolytic pathway. This restriction is due to the inactivation of glycolytic enzymes such as phosphofructokinase and fructose-6-phosphatase at low temperatures (Malone et al. 2006). The hexose phosphates and other starch metabolites are diverted into the sucrose synthesis pathway where they are converted to sucrose by the enzyme sucrose phosphate synthase. Sucrose is subsequently hydrolysed into glucose and fructose (Wiberley-Bradford et al. 2014).
The differences in fructose accumulation could be attributed to genotypic variations of the tubers influencing the rate and ratio to which enzyme sucrose synthase converted sucrose to uridine diphosphate glucose (UDP-glucose) and fructose. On the other hand, the genotypic differences of the varieties could have led to differential activities and affinities of hexokinases and fructokinases responsible for the irreversible phosphorylation of fructose for metabolic processes.
Since tuber sugar level is important in influencing the processing quality of tubers for frying, low sugar content is an essential feature with significant commercial value. Fructose and glucose react with free amino acids in potatoes during processing at high temperatures, leading to the development of undesirable brown-coloured and bitter-tasting potato products through a process called Maillard reaction (Kumar 2011). Maillard reaction is a persistent and costly challenge for the potato processing industry. Furthermore, this reaction generates acrylamide, a compound known to be a neurotoxin and a potential carcinogen (Bhaskar et al. 2010).

Fig. 8
Relative change in glucose concentration of three potato tuber varieties stored under different temperature and RH storage conditions. Values are means of triplicate determinations ± standard error. Bonferroni's means separation test used at 5% level of significance. Key: RT/RH: room temperature and relative humidity (21.7 ± 5 °C/73.5 ± 6.7%). Ambient RH: relative humidity of ambient air (not determined)

Total Reducing Sugars
The total reducing sugar concentration (total content of fructose and glucose) of the three potato varieties overtime during different storage conditions is presented in Table 1. There were significant differences (p < 0.05) in total reducing sugar content across tuber varieties in all the storage conditions. The recommended total reducing sugar concentration limit for potato tubers intended for high heat processing is 250-500 mg/100 g (Mareček et al. 2013). In this study, tubers of all three varieties stored at room temperature were still in good processing condition up to the 91st day (week 13) of storage with respect to total reducing sugar content. At 7 °C/75% RH, crisp and French fry processing is possible up to day 77 of storage for Shangi tubers, day 91 for Unica tubers and day 63 Overall, in this study, storage caused an increase in total reducing sugars by 26.0-to 68.5-fold at 7 °C/75% RH, 25.5-to 61.7-fold at 10 °C/ambient RH, 23.7to 58.8-fold at 10 °C/75% RH and 3.8-to 9.3-fold at room temperature. These findings are in agreement with the results of Galani et al. (2016) who reported the lowest accumulation of reducing sugars at room temperature, the highest increase in cold storage (4 °C) and a significant variation in the degree of accumulation among the Indian potato varieties. In terms of storage at RT/RH, our findings differed from those of Abong et al. (2009) who reported that ambient storage conditions did not trigger a significant increase or decrease in reducing sugar content for Kenyan varieties such as Dutch Robijn, Tigoni, Kenya Karibu and Kenya Sifa. On the other hand, Abong et al. (2015) reported a similar trend to the findings of this study for cold storage. The Shangi variety exhibited an increase in reducing sugar concentration by 1.24-fold in ambient conditions, 1.82-fold at 12-14 °C, fourfold at 8-10 °C and 5.25-fold at 4-6 °C .
Tubers stored at RT/RH experienced the least sugar accumulation and would be best for processing except for the fact that they had deteriorated in other aspects such as weight loss, rotting, sprouting and greening. Sugar accumulation at RT/RH may have been due to sugar mobilization stimulated by sprouting and physiological aging. Sprouting is accompanied by high rates of starch hydrolysis which is concomitant with the accumulation of reducing sugars to provide carbon and energy for sprout growth and development (Freitas et al. 2012;Pinhero et al. 2009). It was also noted that although Shangi had the lowest reducing sugar content among the other varieties at the beginning of the storage experiments, it did not maintain the least sugar concentration over the storage period. This suggests that varieties with low reducing sugar content at harvest will not necessarily have low sugar content after storage compared to their counterparts which might have had higher reducing sugar content before storage.

Conclusion
From this study, it was established that Shangi variety has the lowest storability while Dutch Robijn generally exhibited the best storability in terms of physical traits under all the storage conditions in this study. Storage at 7 °C/75% RH best preserved potato tubers in terms of weight, and inhibition of tuber sprouting, greening and rotting. However, it resulted in the highest rate of accumulation of sugars in the potatoes compared to the other storage conditions. This is undesirable in the potato processing industry. The sugar accumulation, however, can possibly be overcome through reconditioning at room temperature and RH for a certain duration as suggested in various studies. It is recommended that the effect of reconditioning post cold storage be determined for the three potato varieties under this study as well as the enzyme activity leading to cold-induced sweetening (CIS) in those varieties. Other solutions for CIS-sensitive varieties can be explored such as the use of natural sprout suppressants to inhibit sprouting and breeding for CIS-resistant varieties to improve potato storage in Kenya. Further studies are also recommended to determine the accumulation of glycoalkaloids in potato tubers during storage.