Frozen Mulberry Inspection: Stalks, Breakage and Natural Fruit Shape

Inspect frozen mulberries by separating three questions: whether the fleshy fruit body is complete, whether an external stalk remains attached, and whether loose plant material is present. A curved or tapered mulberry is not automatically broken. Conversely, a recognisable fruit can still have an unacceptable stalk or a torn end. Agree those classifications before quoting a whole-fruit percentage.

For buyers, the difficult samples are usually the borderline ones: a narrow tip that looks clipped, a short brown projection at the base, or a small piece that might be an entire naturally short fruit. We would resolve those examples with close photographs and a written inspection rule before approving a supply specification. The goal is a decision two inspectors can reproduce, with the result tied to the application and the supplied lot.

Start with the mulberry body, not a perfect outline

Mulberries do not have the smooth surface of a blueberry. The visible body contains many small fleshy units, or drupelets. A 2025 primary study of mulberry development describes mulberry as a multiple fruit formed from many flowers on one inflorescence; the peduncle supports that structure. This explains the compound surface and the external stalk connection. The study does not establish a frozen-product grade.

Inspect the continuity of that fruit body rather than demanding one ideal silhouette. A whole sample may include short, elongated, tapered or gently curved forms. Those forms can differ visibly without a section having broken away. If a buyer needs a narrow size or shape range for a decoration, that is an additional selection requirement. Calling every naturally short fruit “broken” makes the reported damage rate depend on an unstated appearance preference.

Turn a doubtful piece gently to view its end and sides. A closed surface covered with connected drupelets can support an intact classification under the agreed reference. An abrupt plane of exposed flesh, a missing side or a detached section presents a different question. Frost can obscure both, so record when the viewing condition prevents a reliable decision. Do not force a classification simply because the inspection form has no uncertainty field.

Length alone cannot settle the issue. A short whole fruit and the remaining portion of a longer broken fruit can measure the same. A ruler helps document scale; it does not replace examination of the surface. Keep size measurements and structural damage in separate fields. This lets a buyer accept natural shape variation while still enforcing a useful minimum size for a particular finished product.

Separate attached stalks from loose plant material

Use a small identification dictionary at the inspection station. “Stem-free” can be interpreted as no long stalks, no visible external projections, or no detectable plant material in the inspected sample. Those are different requirements. Describe which condition the order actually purchases and how small fragments are handled in the agreed method.

Attached stalk incidence and loose stalk loading answer different operational questions. Attached stalks may require handling individual fruits; loose material may be found while spreading a sample. If one attached stalk detaches during inspection, do not silently turn one observation into two defects. Record its original state and any handling change. Decide in advance whether the reporting rule follows the received condition or a specified later assessment stage.

A short projection also needs a visual reference. Inspectors should be able to distinguish external plant tissue from fleshy material at the fruit base without tearing every doubtful piece apart. Where identification is uncertain, retain the object and request a qualified assessment. A photograph may support classification, but it cannot always identify a tiny dark object conclusively.

Keep unrelated foreign material outside an appearance negotiation. A commercial tolerance for fruit stalks does not automatically authorise wood, packaging fragments or unidentified hard matter. The applicable product and safety requirements need their own disposition. Equally, a photograph showing no visible stalks in a handful cannot prove absence throughout a shipment. Write what was inspected and what was observed.

Define damage by missing structure and crushing

A practical mulberry dictionary can distinguish complete bodies, bodies with a limited damaged area, larger broken sections, small fragments and crushed material. The boundaries should follow the buyer’s intended use and the supplier’s offer. They are proposed commercial classes, not an official grading system. Include photographs of the boundary cases in the approval record.

For a partly damaged fruit, decide whether a small missing group of drupelets changes its whole-fruit classification or receives a separate surface-damage flag. Both approaches can work when explicitly agreed. Problems arise when the seller accepts a recognisable body as whole while the receiver requires every surface to be complete. A single percentage then appears to contradict another even though the physical sample has not changed.

Crushing is also different from simple separation. A broken section can retain distinct connected drupelets; crushed fruit may have lost its body and become a compressed or smeared mass. Document what you mean by crushed using actual reference material. Do not infer the cause from shape alone. Pack pressure, production handling, sampling and subsequent handling require their own evidence before responsibility can be assigned.

Colour belongs beside these observations. Frosted purple fruit, a redder patch and exposed flesh can look different in a photograph. That does not justify labelling every lighter area unripe, damaged or spoiled. Compare colour under the agreed viewing state, with the relevant product reference. Where condition raises a separate quality concern, record it explicitly rather than hiding it inside the broken fraction.

For clumped material, record the original condition first. Breaking apart a fused mass can create fragments and alter the very result being measured. Agree how separability is assessed and what manipulation is allowed. If the original material cannot be evaluated by the contracted method, preserve that finding and request instructions; do not produce a precise broken percentage from an undocumented forceful separation.

Set up a sample that preserves the evidence

Start the inspection record before emptying the package. Capture the sample code, lot identification, pack format, arrival condition and time of examination. Photograph the closed pack and the initial fruit condition where relevant. State whether the material is a development sample or was drawn under the shipment’s agreed sampling plan. These records answer different questions and should remain identifiable.

Select the packages and increments according to the agreed plan. A decorative handful chosen from the top is useful for a product photograph, but it does not represent the distribution of defects in a lot. Loose stalks and small fragments can be distributed unevenly within a package. Preserve individual package observations where the acceptance rule requires them, even if a combined result is also reported.

Use a clean, suitable contrasting tray and enough space to expose the fruit surfaces without piling the material deeply. Transfer gently with equipment appropriate to the product. Note material remaining in the liner and on the tray; it is easy to exclude fines or loose plant matter unintentionally when only large fruit is moved. Account for what was included in the evaluated portion.

Fix the observation state. A frozen appearance check, a controlled thawed check and an application trial are separate evaluations. Follow the product’s intended-use instructions and the site’s approved handling procedure. Record the actual state, timing and preparation used. This article’s visual method does not specify a safe holding period or establish ready-to-eat suitability.

If frost prevents a decision, use the agreed clarification procedure and preserve the initial photograph. Do not rinse one supplier’s sample while evaluating another directly from the bag. That can remove loose material and change surface appearance. Any preparation that changes the sample needs to be defined, applied consistently and disclosed in the result.

Keep defect counts and broken mass separate

Commercial specifications often combine several units. A public supplier mulberry sheet revised in 2015, for example, lists stalks and leaves by count per kilogram and broken or damaged fruit as a percentage. It illustrates why the unit matters; its limits and older compliance references are not adopted here. A percentage without a stated mass or count basis remains incomplete.

Consider an illustrative inspection of 1,000 g of product. Suppose 35 g meets the agreed broken-fruit definition and three detached stalks are recorded in that same original sample. Broken fruit is 3.5% of the original sample mass; detached stalk loading is three per kilogram. The results cannot be added. They measure different things, and neither figure is an acceptance limit or an actual XMG test result.

For attached stalks, a separate selected count check might find six fruits with stalks among 100 assessed fruit bodies. That is 6% of those assessed bodies. It is not six stalks per kilogram. If bodies differ in size, their count and mass contributions differ. Write exactly which bodies were eligible for the count and how fragments were handled, especially when the fruit condition makes “one fruit” ambiguous.

Do not reconstruct original fruit counts by guessing which fragments once belonged together. If the contracted measure is broken material by mass, weigh the defined fraction. If the decision requires complete decorative pieces per portion, count usable pieces under that application protocol. Each result can be useful without converting it into the other.

Also define overlapping defects. One fruit can have an attached stalk and a damaged side. Attribute flags may record both conditions, while a mass balance requires each piece to enter only one exclusive fraction. A practical record can use exclusive structural classes for the mass calculation and a separate stalk observation field. State that arrangement so a reader does not expect all reported percentages to sum to 100.

Reconcile the inspected mass, including any separately retained material and unexplained handling loss. Keep the original denominator visible. Removing fine fragments before weighing the “sample” can inflate the apparent whole-fruit proportion. A clear reconciliation lets another inspector repeat the arithmetic and see whether the difference is product condition or a different calculation.

Use the intended product to set appearance priorities

A whole-fruit topping needs a different result from a cooked fruit preparation. For a visible topping, inspect recognisable fruit bodies after the actual portioning and preparation steps. Record how much sorting is required and how many usable portions remain. An attractive frozen tray does not demonstrate the appearance after the buyer’s handling sequence.

For a preparation that deliberately breaks the fruit down, paying for a narrow silhouette may add little value. Stem control, the desired particle texture and the performance of the agreed fruit fraction can matter more. Test the offered form in the intended recipe and process. A broken-fruit allowance should be a deliberate purchasing choice, with other product requirements still in place.

Do not assume later screening will resolve every unwanted material issue. Confirm what the actual equipment removes, what passes through, and what additional product loss or operating work occurs. Run a suitable trial before assigning a yield credit to a less selected grade. A claim that the material “will be blended anyway” does not answer whether stalks are acceptable in the input specification.

Compare commercial options using the same required outcome. If one proposal needs extra manual selection for visible fruit, include observed labour and usable recovery in the evaluation. Keep the trial record separate from a shipment-wide guarantee. If the sample size or pack differs from the order, identify the remaining confirmation needed before relying on the trial’s economics.

Build a reference sheet inspectors can share

A useful reference sheet contains more than one ideal fruit photograph. Show an accepted natural taper, a naturally short whole body, an attached stalk, a loose stalk, a torn end and the agreed boundary for partial damage. Use consistent lighting, a scale where relevant and a plain background. Label the observed feature directly without relying only on a red or green mark.

For each image, state the class, observation state and any condition that changes the decision. An image approved for frozen sorting should not silently become the reference for a thawed presentation. Date the sheet, give it a version and link it to the product specification. Store the original photographs as well as the labelled sheet so details remain available when a small reproduction is unclear.

Test the dictionary before testing a supplier dispute. Ask two inspectors to classify the same group of borderline examples independently. Compare disagreements and refine the words or photographs. If one person calls a tapered intact fruit broken and another accepts it, the first repair belongs in the definition. Repeating the inspection without resolving that difference will reproduce the disagreement.

Agree how later changes enter the record. A new crop, selection requirement, form or packaging route may warrant a fresh reference or trial. Specify which changes need notification and which require renewed approval. Keep the superseded version identifiable but separate from the current receiving instruction, so an old image does not become an accidental release criterion.

Resolve a disputed lot with the original evidence

Describe the deviation using the contracted fields. “Too many bad mulberries” is difficult to investigate. “Attached stalk incidence exceeded the agreed criterion in the named sample” or “broken mass was calculated using a different denominator” gives both parties a specific issue to resolve. Attach the method, raw observations, photographs and calculation.

Preserve the affected sample and package information under the agreed procedure. Record any manipulation already performed. If the disputed material has been thawed, screened or repeatedly poured, disclose that history; it may no longer represent the original receiving condition. Use retained or replacement samples according to the agreed dispute process instead of presenting altered material as untouched evidence.

Compare definitions before debating responsibility. Confirm that the same reference version, observation state and reporting unit were used. Then examine lot and handling records relevant to the actual finding. Shape alone cannot identify when damage occurred. A documented agreement on reinspection, acceptance, rework or another disposition should follow the evidence and the responsible teams’ decision.

References

• Yang and colleagues, mulberry fruit development and abscission, Frontiers in Plant Science, 2025. See the introduction for multiple-fruit and peduncle structure.
• ISOLP, Mulberry IQF specification, revision dated 2015. Historical commercial example of separate stalk counts and damage percentages.