Three disclosed facts separate a grower from a manufacturer: owned land carrying a named country and a named crop, owned post-harvest capacity counted by type rather than by total, and revenue that comes from fruit the company packed and cooled rather than fruit it merely moved.
Owned land is the first test, and the disclosure is either specific or it is not. Dole plc farms about 110,000 acres of its own production, roughly 100,000 acres of it Fresh Fruit land in Central and South America, which is specific enough to check against a filing. Hortifrut S.A. held 4,210 hectares of planting at 31 December 2025, of which 2,355 hectares were owned and 1,854 leased, with 4,043 in production, and the figure of more than 10,000 hectares that circulates widely is roughly two and a half times the disclosure. Unifrutti reports 60 owned farms and states its land as 14,000 hectares in its headline numbers against more than 16,000 hectares managed after its 2024 acquisitions. A buyer should ask for a country, a crop and a hectare count together, and treat a claim carrying only one of the three as unverified.
Post-harvest capacity has to be counted by type, because one number can hide two entirely different businesses. Dole reports 250-plus facilities worldwide and the filing splits them: about 75 are pack houses, cold storage and ripening facilities, while about 160 are marketplace and distribution operations that process no fruit at all. Del Monte Corporation runs 18 fresh-cut processing plants against 31 global distribution centres with cold storage and banana ripening, four US port facilities and roughly 11,000 refrigerated containers. Hortifrut operates 17 company-owned fresh operations and also 31 third-party operations run on site by its own teams, which is operational control without ownership. Zespri's 39 pack houses and 64 cool stores in the 2025/26 season are partner or contracted facilities rather than owned plants.
A grower sells a crop while a manufacturer sells a finished, cooled, scheduled product, so the customer list is the third test. Joy Wing Mau is the clearest manufacturer by function and the weakest by ownership: it holds essentially no production, yet reports 30-plus cold-chain logistics centres, distributes more than 2,000 tonnes of fruit a day on its own website figure, serves more than 10,000 supermarkets and 25,000 fruit specialty stores across 300-plus cities, and reaches more than 100 domestic planting bases through planting, acquisition, cooperation and minority equity. Chiquita shows why ownership alone settles nothing: it owns around 70 banana plantations across five Latin American countries, but about 50 third-party suppliers provide roughly 60% of its volume, so the label and the ripening discipline are its own while most of the fruit is not. Costa's most modern post-harvest asset is a 10,000 sq m packhouse at Larache in Morocco with a 12-lane packing line.
Where a company publishes nothing, the silence is a finding rather than a gap to be filled. Driscoll's releases no revenue, no headcount and no facility count, and no primary source confirms the totals that circulate for it. Chiquita's last public financial statement is FY2014 net sales of US$3,090.2 million, which is historical and should never be presented as current, and Costa's last public full year is FY2023, A$1,458.55 million of revenue against a A$24.32 million loss, because it left the ASX in February 2024. Licensed plantings and member-grower networks are real production capacity and never owned capacity, and sites that publish estimates of unknown methodology are not a substitute for a filing.
The honest answer is that the share runs from almost nothing to most of the volume, and the published numbers cannot be lined up side by side because companies are answering three different questions with what looks like a single figure.
The unit, the scope and the crop are never the same from company to company. Dole plc reports about 110,000 acres of its own production and narrows it to roughly 100,000 acres of Fresh Fruit land in Central and South America, so the number arrives with a crop and a region attached. Mission Produce reports 4,000-plus hectares across California, Peru, Colombia, Guatemala and South Africa growing avocados and mangoes, and that crop list is the part most often wrong in circulated summaries, which describe the same hectares as blueberry farms in one country. Hortifrut S.A. reports 4,210 hectares of planting with an explicit split of 2,355 hectares owned against 1,854 leased and 4,043 hectares in production. Unifrutti states its land two ways, as 14,000 hectares in its headline numbers and more than 16,000 hectares managed after its 2024 acquisitions.
Ownership, lease and management are three different claims that companies rarely keep apart. A long lease behaves like ownership on an income statement and not at all like ownership on a balance sheet, and a management contract delivers operational control with neither. Zespri's 16,115 production hectares, 3,404 registered orchards and 2,786 producers belong to its growers rather than to Zespri, which is exactly why the cooperative's power sits in a plant variety licence instead of a land title. Hortifrut's 31 third-party operations run on site by its own teams are a management claim dressed in operational language, and Costa Group's 8,532 hectares include 2,565 hectares under protected cropping, a different cost base from open-field hectares that should not be averaged with them. Where a hectare is leased, managed, contracted or licensed, the exposure is smaller and the balance sheet is thinner, so a reader who counts all four as owned land will overstate the company every time.
The right denominator is volume rather than land, and integration also runs the other way, into processing. Chiquita owns around 70 banana plantations across five Latin American countries, yet about 50 third-party suppliers account for roughly 60% of its volume, so a hectare-based reading of Chiquita describes a company that barely exists. Hortifrut sold 189 million kilograms in 2025 on 4,043 hectares in production and still reported FY2025 income from operating activities of US$1,215.0 million with a net loss of US$77.7 million, which shows how much a volume figure can conceal. Del Monte Corporation converts its own harvest and purchased fruit through 18 fresh-cut plants into a higher-margin prepared product, and Hortifrut runs six company-owned Vitafoods frozen plants with 33,500 metric tonnes of combined annual capacity.
Where a company contradicts itself, the lower figure is the one to use and the contradiction should be reported rather than smoothed over. Joy Wing Mau's own materials give its daily throughput as 8,000 tonnes in one profile and 10,000 tonnes in another against a website claim of more than 2,000, and its warehouse area as 200,000 sq m in one profile and 300,000 sq m in another. Chiquita's own site states approximately 20,000 people in 70 countries in one place and 18,000 people in 25 countries in another, with nothing reconciling the two. The practical rule is to take the conservative figure, state the range, and treat any producer whose operating numbers move by a quarter between two of its own documents as one whose unverified claims should stay unverified.
Ownership moves the loss into the producer's own income statement and onto its own fixed cost base, so a missed harvest stops being a sourcing problem and becomes a solvency question, and the disclosed results of the companies on this page show the shape of that transfer.
A shortfall on owned acreage lands as revenue decline or margin compression rather than as a supplier relationship that simply lapses. Hortifrut sold 189 million kilograms in 2025, up 25% on the prior year, and still reported FY2025 income from operating activities of US$1,215.0 million with a net loss attributable to owners of US$77.7 million, because volume growth and a loss can arrive together and only the detailed statements show which cost line absorbed the difference. Costa Group's last public full year, FY2023, was A$1,458.55 million of revenue against a A$24.32 million net loss, following net income of A$33.63 million in FY2022, a swing that had nothing to do with its market position. Del Monte Corporation turned US$4,322.3 million of FY2025 net sales into a 9.2% gross margin and US$90.7 million of net income attributable to the company.
The owner carries labour and facility costs that a buyer does not, and this industry discloses them with unusual clarity. Hortifrut employed 3,038 permanent collaborators at 31 December 2025 together with a further 24,532 seasonal and temporary workers across 11 countries, a total 2025 workforce of 27,570 that peaked at 28,033. Del Monte's workforce of roughly 40,000 people consists of 8,562 full-time employees and more than 31,466 daily, seasonal or temporary workers, about 84% of them based at production locations. Costa reports 11,156 full-time equivalents for 2025. When a harvest fails the seasonal block can be cut and the permanent block cannot, which is why the ratio of permanent to seasonal labour is the best available proxy for how much of a bad season a company is carrying.
Spreading production across hemispheres is the cheapest insurance available, and it is bought in geography rather than in financial instruments. Costa grows berries at nine sites in Yunnan, operates farms and a packhouse in Morocco and runs glasshouses at Guyra in New South Wales. Unifrutti spans Chile, Peru, Argentina, Ecuador, South Africa, Italy, Spain and the Philippines, assembled partly by acquisition, including more than 7,500 hectares added through Verfrut, roughly 1,400 acres of Peruvian table grapes through Safco and about 990 acres of Sicilian citrus acquired in April 2026. Zespri supplements New Zealand supply with northern hemisphere production and chartered 66 vessels in its 2025/26 shipping programme, which turns a single-season catastrophe into a single-region problem rather than a company-wide one.
Owners also lose on things that have nothing to do with weather, and the disclosure matters as much as the result. Hortifrut exited the cherry business in June 2026, having built it to US$21.3 million of FY2025 revenue inside a US$1,227.1 million income base, which is a capital allocation decision rather than a climate one. Zespri's NZ$280.1 million net profit for 2025/26 falls to NZ$123.8 million once plant variety licensing revenue is stripped out, and its mainland China sales fell to NZ$1,089.0 million from NZ$1,106.6 million, which is demand exposure rather than supply exposure. Costa has published nothing since FY2023, Unifrutti's last reported turnover is US$720 million for 2021, and Driscoll's publishes no financial statements at all. A loss reported under audit is worth more than a smooth story told without one.
Because in a perishable trade the container allocation and the cool room are the schedule, and the schedule is the product, so owning the lane buys the ability to decide when fruit arrives rather than a cheaper way to move it.
The disclosed fleets are tiny in number and enormous in consequence, and only two companies here own one. Dole plc owns 13 refrigerated vessels, nine refrigerated container carriers and four conventional refrigerated ships, with one further vessel on charter, and describes the fleet as the largest dedicated refrigerated containerised fleet in the world. Del Monte Corporation owns six vessels and charters one, and runs four US port facilities, roughly 11,000 refrigerated containers, about 419 trucks and refrigerated trailers in the United States and a further 241 in the Middle East. Between them these two companies dictate the sailing schedule their own fruit travels on. Everyone else on this list depends on third-party container lines and chartered tonnage, and where that is disclosed it should be read as a stated limit rather than a hidden weakness.
Renting is the normal condition of this industry, and the disclosure about it is refreshingly explicit. Zespri's shipping programme for 2025/26 included 66 chartered vessels and its 39 pack houses and 64 cool stores are partner or contracted facilities, so a single-crop cooperative moves a record 248.1 million trays through infrastructure it mostly does not own. Mission Produce owns four pack houses and relies on the open container market for ocean freight. Hortifrut owns 17 fresh facilities and manages a further 31 third-party facilities on site with its own teams, buying operational control without the capital cost. Unifrutti reports 43 facilities, which the company describes as processing, commercial and logistics facilities rather than itemised cold stores or packing lines.
Ripening and pre-cooling are the other half of the asset, and they are where a harvested commodity is finished into a sellable product. Del Monte's 31 distribution centres include cold storage and banana ripening, and its 18 fresh-cut plants turn whole fruit into a prepared product at a price a grower selling field-run fruit cannot reach. Dole's roughly 75 pack houses, cold storage and ripening facilities sit inside a network of more than 250 sites, of which about 160 are distribution and marketplace operations, so the industrial footprint is roughly a third of the site count and should never be quoted as the whole of it. Mission Produce ripens under its Mission Control technology and reports nine ripening centres across the United States. Hortifrut's six Vitafoods frozen plants, with 33,500 metric tonnes of combined capacity, convert a perishable surplus into inventory that keeps.
Ownership stops paying when the asset is geographic rather than operational, and the test is whether a count appears in a filing. Chiquita owns around 70 banana plantations, yet about 50 third-party suppliers provide roughly 60% of its volume, so the owned plantations do not correspond to the fruit that moves. Joy Wing Mau owns essentially no production and no vessels, yet reports 30-plus cold-chain logistics centres, because somebody else owns the fruit. The widely repeated claim that Del Monte owns 13 refrigerated ships is a misplaced Dole figure, and consistency between a company's own documents is the cheapest available test of whether its operating numbers have been reconciled at all.
Because the filing obligation follows the listing rather than the company, so a take-private removes the disclosure and not the business, and the result is that two of the ten producers on this page have no current audited financial statements in the public record at all.
A take-private ends the reporting requirement outright, and Costa Group is the clearest case here. Costa traded on the ASX as CGC from 24 July 2015, had its shares suspended from the close of trading on 8 February 2024 and was removed from the official list on 27 February 2024 at its own request. The scheme was led by Paine Schwartz Partners together with Driscoll's and British Columbia Investment Management Corporation as co-investors, implemented through Chilli Buyer Pty Ltd at A$3.20 per share, for an equity value of about A$1,496 million. Its last published full year is therefore FY2023, with revenue of A$1,458.55 million and a net loss of A$24.32 million. No FY2024 or FY2025 results exist publicly, and any figure presented as one has been invented rather than found.
Delisting does not always mean silence, and the difference between these two cases is worth understanding. Hortifrut deregistered its shares from the Chilean securities registry, a step the regulator accepted in May 2024 after a shareholders' meeting agreed to it in October 2023, and its shares no longer trade on the Santiago Stock Exchange. It nonetheless continues to publish audited quarterly and annual financial statements, because its corporate bonds remain registered. That is why a reader can still see FY2025 income from operating activities of US$1,215.0 million, a net loss of US$77.7 million, 189 million kilograms sold and 4,210 hectares of planting, and can also see the season-basis figure of US$1,280.1 million for the twelve months to June 2026. Costa and Hortifrut were both delisted, and only one of them still tells you what it earns.
Family and cooperative ownership create the same gap for different reasons, and it is not always a refusal to disclose. Driscoll's is a fourth-generation family business that releases no revenue, no headcount and no facility count, and no primary source substantiates the totals that circulate for it. Chiquita has been privately held since 2015 by the Cutrale and Safra families on a 50/50 basis, and its last public financial statement remains FY2014 net sales of US$3,090.2 million, which is historical rather than current. Unifrutti is majority owned by ADQ of Abu Dhabi with no disclosed shareholding percentage, and its most recent published turnover is US$720 million for 2021 with EBITDA of US$78 million. Zespri is the counter-example that matters, because a grower-owned cooperative with no main-board listing still publishes 2,786 producers, 3,404 registered orchards and 16,115 production hectares.
Treat a disclosure gap as a statement about verifiability rather than about size, and never fill it with an aggregator estimate. Joy Wing Mau is unlisted, and its revenue figure of more than RMB 20 billion for 2024 reaches the public through district industry and commerce data and Chinese media rather than filed accounts, so no figure for it is presented here as audited. The published record for this sector is full of numbers that were plausible and wrong: land at more than 10,000 hectares against 4,210 disclosed, a workforce above 55,000 against 3,038 permanent collaborators, and Del Monte credited with 13 ships that belong to Dole. Fiscal-year dates matter for the same reason, since Mission Produce's year ends on 31 October and Del Monte's FY2025 closed on 26 December, and where nothing is disclosed the entry says so rather than substituting an estimate.
