**1. Introduction**

Blueberries (*Vaccinium* spp.) are becoming increasingly popular due to the rising awareness of the health benefits of consuming blueberry fruit, which include decreased risk of cardiovascular diseases, improved cognitive performance, and decrease in aging-related damage [1,2]. Commercially important blueberry species include lowbush (*Vaccinium angustifolium* Ait.) and northern highbush (*Vaccinium corymbosum* L.) mainly cultivated in the northern parts of the United States, and rabbiteye (*V. virgatum* Ait.) and southern highbush (hybrids of *V. corymbosum*, *V. virgatum*, and *V. darrowii* Camp.) grown mostly in the southern states [3,4]. Recently, production of blueberries has expanded to 27 countries (in 2011) compared with only ten countries in 1990 [5]. The United States is the largest producer of blueberries globally [5], supplying 347.7 million kg of cultivated and wild blueberries in 2016 [6]. The United States also plays an important role in the import and export trade of blueberries [7]. In 2016, the United States exported 31.7 million kg of fresh and 25.4 million kg of frozen blueberries and imported 149 million kg of fresh and 75.6 million kg of frozen fruit [8].

As global production and trade continues to rise, it becomes increasingly important to maintain fruit quality, nutrient content, phytosanitary safety, and eliminate pests and diseases in blueberries during storage to ensure that this fast-growing export and import market is not negatively impacted. Postharvest losses in fruits can vary from 10 to 40% [9]. After harvest, blueberries have a shelf-life of approximately 7 to 40 days depending on the genotype, method of harvest, and storage regime [9,10]. During postharvest storage, blueberry fruit quality can decline due to fruit softening [11]. Other contributing factors in loss of fruit quality are postharvest diseases caused primarily by fungal plant pathogens such as *Colletotrichum* spp. (ripe rot), *Alternaria* spp. (Alternaria fruit rot), and *Botrytis cinerea* (gray mold), among others [12–15]. In addition to postharvest disease-causing organisms, it is important to eliminate foodborne pathogens or associated indicator organisms [16–18]. Although outbreaks of foodborne illnesses associated with consumption of blueberry fruit have been relatively rare, produce brokers and buyers have begun to apply rigid (and typically proprietary) microbial standards to frozen blueberries destined for the processing market [19]. Although similar standards currently are not in place for the fresh-market, reducing microbial risk remains a key consideration for fresh-market production as well [20]. Finally, in order to export blueberries to other countries, they are required to be certified free of certain insect pests such as Mediterranean fruit fly (*Ceratitis capitata*), South American fruit fly (*Anastrepha fraterculus*), European grapevine moth (*Lobesia botrana*), blueberry maggo<sup>t</sup> (*Rhagoletis mendax*), and plum curculio (*Conotrachelus nenuphar*) [21,22].

Fumigation of export goods with methyl bromide was the most commonly used phytosanitary treatment for elimination of pests, but has been phased out in the United States, with the exception of a few critical uses [23,24]. Methyl bromide also requires the produce temperature to be increased in order to be effective, thereby breaking the cold-chain and potentially having an adverse effect on quality. Interruption of cold-chain can decrease shelf-life considerably by increasing undesirable fruit metabolism [25]. Irradiation using gamma rays, X-rays, or electron beams could be an alternative to fumigation in eliminating pests and in preserving quality by reducing decay organisms and plant and human pathogens [23,24,26]. Previous work supported the use of electron beam and gamma irradiation to maintain shelf-life and fruit quality attributes in blueberry fruit [27–30]. In the United States, regulatory approval has been obtained for the use of irradiation on fresh fruits and vegetables up to 1 kGy [31]. Previous studies suggested an irradiation dose of 0.4 kGy to be effective against most insect pests, 0.2–0.8 kGy to cause a 1-log reduction in surface bacterial pathogens causing foodborne illness, and higher doses of 1–3 kGy for postharvest disease-causing fungi [22,32–34].

The objective of this study was to determine the effect of irradiating postharvest blueberry fruit using a new form of electron beam technology, Electronic Cold-PasteurizationTM (ECPTM) developed by ScanTech Sciences (Norcross, GA, USA) at their Research and Development (R&D) facility at Idaho State University (ISU). This R&D facility is a small-scale version of a commercial ECPTM food treatment facility, which is currently being constructed by ScanTech in McAllen, TX and will be operational in the fourth quarter of 2018. This technology employs a highly focused beam of electrons, treating samples for only milliseconds on a high-speed conveyor while maintaining cold-chain integrity. A key advantage of electron beam irradiation over gamma rays (from nuclear sources such as Cobalt-60) or X-rays is the ability to deliver extremely high dose rates with improved accuracy since the beam dynamics can be more precisely controlled. These high dose rates equate to significantly less time for treatment and, consequently, potential for higher quality produce. The ECPTM treatment can treat an entire truckload (around 60,000 clamshells) of blueberries in a little over 30 min, whereas gamma rays can take several hours for the same quantity (C. Starns, unpublished observations). This is the first study to investigate the effect of irradiation on fruit quality attributes, postharvest disease incidence, and surface microbes of food safety concern in two southern highbush blueberry cultivars treated with ECPTM prior to cold storage.

#### **2. Materials and Methods**

#### *2.1. Fruit Collection and Irradiation*

Two trials were conducted with hand-harvested fruit from southern highbush blueberry cultivars 'Farthing' and 'Rebel' in Alma, GA. In trial 1 (April 2016), 'Farthing' fruit were obtained from a commercial packing facility, where fruit had already been prepacked into pint-size clamshell containers (473 mL). In trial 2 (May 2016), 'Rebel' fruit were obtained from a different packing facility, also already prepacked in pint-size clamshells. In addition, trial 2 included 'Farthing' fruit hand-harvested by the investigators from a commercial blueberry farm and packed into pint-size clamshells.

A subsample of clamshells in each trial was taken directly to the University of Georgia, Athens, GA, USA (330-km transit in refrigerated cooler) to serve as an unshipped control (not transported to and from the irradiation facility). Initial fruit quality attributes and postharvest disease incidence were recorded from this unshipped control. The remaining fruit in clamshells were arranged on standard flats (12 clamshells/flat), placed in a styrofoam cooler with ice packs, and shipped overnight from Alma, GA to ISU, Pocatello, ID. A foam sheet was placed on the inner side of the lid of each clamshell and in between clamshells to minimize fruit injury during shipment.

At ISU, fruit in clamshells were subjected to electron beam irradiation treatment at ScanTech's R&D facility using proprietary ECPTM technology. A 10-MeV electron beam, driven by an advanced high-energy electron accelerator, is magnetically focused through a scanning horn which delivers precision dose control. At the R&D facility, clamshells containing fruit were subjected to four levels of irradiation, 0, 0.15, 0.5, and 1.0 kGy; the treatments were completed in less than a second per clamshell. The respective doses were achieved using the National Institute of Standards and Technology (NIST)-traceable alanine pellets with extensive dose mapping on various blueberry configurations prior to the experimental fruit being shipped to the facility. Hundreds of data points were obtained and measured on a Bruker Bio-spin Electron Paramagnetic Resonance spectrometer, all of which are NIST traceable and International Organization for Standardization/American Section of the International Association for Testing Materials compliant. Treatments were replicated four times (i.e., four clamshells/irradiation level/postharvest storage period/cultivar), with a few exceptions where fewer replicate clamshells were available. The 0-kGy treatment served as an untreated control wherein fruit were shipped but not irradiated. After irradiation, fruit were shipped back by overnight courier to the University of Georgia where they were placed in a walk-in cooler at 2 to 4 ◦C under high relative humidity (>85%) until further assessment. The entire shipping and treatment process (from Alma to the treatment facility at ISU and to Athens for cold-storage and evaluation) took between 6 to 7 days. The unshipped control clamshells were stored in a 2 to 4 ◦C walk-in cooler until further evaluation. Fruit were removed from cold storage and evaluated for postharvest fruit quality attributes at 6, 13, and 25 days after irradiation treatment; microbial load on the fruit surface at 6 days after treatment; and postharvest disease incidence at 6 and 13 days after treatment followed by 4 days at room temperature. Fruit quality, microbial load and postharvest disease incidence analyses at a given time-point were performed using four replicates; for every replicate, fruit from a separate clamshell were used and divided for the above analyses.

#### *2.2. Evaluation of Fruit Quality Attributes*

For evaluation of fruit quality, visual assessment as well as measurement of fruit weight, texture, titratable acidity (TA), and total soluble solids (TSS) content were performed. For visual assessment, 30 fruit per replicate were scored for symptoms of bruising such as tears, dents, leakiness, or signs of mold. Fruit were examined by eye for visual defects and percent sound fruit were calculated. For fruit texture, two variables, fruit compression and skin puncture force, were measured on 12 fruit per replicate using a fruit texture analyzer (GS-15, Güss Manufacturing, Strand, South Africa); fruit were oriented on the equatorial plane for this assessment. For compression measurements, a 1.5-cm diameter plate was used with parameters set at forward speed 6 mm/s, measure speed 5 mm/s, and measure distance 1.00 mm. For skin puncture force measurements, a 1.5-mm flat-tip probe was used with parameters set at a forward speed 10 mm/s, measure speed 5 mm/s, and measure distance 3.00 mm. Fruit weight was recorded on 20 individual fruit per replicate using a balance (Quintix Precision Balance, Sartorius, Bohemia, NY, USA).

For TA and TSS measurement, juice was extracted from ~40 g of fruit per replicate using a household blender and centrifuged for 10 min at 3901X *g* using a benchtop centrifuge (Allegra X-22, Beckman Coulter Life Sciences, Indianapolis, IN, USA). The resulting supernatant was filtered through two layers of cheesecloth. To measure TSS, 300 μL of supernatant was tested using a digital handheld refractometer (Atago USA, Belleveue, WA, USA). For TA, the supernatant was titrated using an automatic mini titrator (Hanna Instruments, Woonsocket, RI, USA) and values were reported as percent citric acid (CA). Statistical analysis (one-way analysis of variance for a completely randomized design) was performed separately for each trial and cultivar using JMP Pro 12 (SAS Institute, Cary, NC, USA). Means were separated using Tukey's Honest Significant Difference (HSD) test ( α = 0.05).

#### *2.3. Evaluation of Fruit Surface Contaminants*

Microbial loads on the fruit surface were determined 6 days after treatment following the protocol described in Mehra et al. [35]. One 50-g fruit sample (~30 berries) per replicate was placed in a 0.5-L flask with 50 mL of sterile phosphate buffer (pH 7.2), and the flask was agitated on a wrist action shaker (Burrell, Pittsburg, PA, USA) at medium speed for 15 min. Aliquots of the wash buffer and 1:20 and 1:100 dilutions were plated in triplicate onto plate count agar (PCA), dichloran rose bengal chloramphenicol agar (DRBC), and Petrifilms (3M Microbiology, St. Paul, MN, USA) for enumeration of aerobic bacteria, total yeasts and molds, and *E. coli* and coliforms, respectively. PCA and DRBC dishes were incubated at room temperature and evaluated after 3 and 5 days, respectively. Petrifilms were incubated at 35 ◦C and colony counts made after 2 days. Colony-forming units (CFU) per gram of fruit were log-transformed and subjected to one-way analysis of variance using PROC GLM in SAS version 9.4 (SAS Institute, Cary, NC, UAS) followed by means separation using Tukey's test.

#### *2.4. Assessment of Postharvest Disease*

An initial postharvest disease assessment was made on the unshipped control following 4 days of storage at room temperature (23–25 ◦C) to allow latent infections to manifest themselves [35]. Subsequently, on fruit subjected to ECPTM treatment, fruit samples (60 berries per replicate) were removed from postharvest storage 6 days (trials 1 and 2) and 13 days (trial 1 only) after treatment, and similarly incubated at room temperature for 4 days. The 13-day assessment was not included in trial 2 as poor fruit quality of 'Rebel' in that trial resulted in near 100% decay after cold storage and subsequent room temperature incubation. For each assessment date and replicate, the number of fruit with symptoms and signs of postharvest decay was counted following examination of fruit samples with a stereo microscope. Fungal pathogens associated with diseased fruit were identified macroscopically and microscopically (utilizing both stereo- and compound microscopes) based on characteristic symptoms and signs [36,37]. Based on the number of fruit with disease symptoms and pathogen signs, postharvest disease incidence was calculated and arcsine-square root transformed for analysis by one-way analysis of variance using PROC GLM followed by means separation using Tukey's test.
