Belén Market lies on the banks of the Itaya River, close to the edge of the Belén District in the city of Iquitos (see Figure 1 ; Supplementary Data Sheet 1 ). Iquitos city in the Department of Loreto, north-east Peru (73.2′ W, 3.7° S), is situated ~120 m above sea level at the confluence of the Nanay, Itaya, and Amazon Rivers in the Amazon River Basin. Most of the population is mixed Spanish and American Indian (Mestizo). The urban area of Iquitos merges with peri urban and rural populations and is surrounded by secondary rainforest, with transportation into and out of the city confined to boats and aeroplanes. The climate in Iquitos is hot, humid, and rainy throughout the year. Mean daily precipitation ranges from 155 mm in August to 350 mm in March, and rainfall is heaviest between November and May. The mean daily temperature ranges from 26.3°C in July to 27.6°C in November.

Iquitos comprises four principal districts with a combined population of 413,556. The districts are Iquitos (population 146,853), San Jaun Bautista (127,005), Punchana (75,210), and Belén (64,488) (INEI, 2023). Belén Market is located on several streets in Belén District. In 2023, Belén Market contained approximately 61 permanent stalls selling wildlife along with a number of temporary stalls and sellers which varied weekly. Modelo Market is situated ~3 km north-east of Belén Market on the edge of the district of Iquitos ( Figure 1 ). The central part of the market is located between the streets Nanay, Callao, Arequipa and Celendín, and contains approximately 29 permanent stalls along with several temporary stalls.

Inventories of all stalls selling wildlife and/or wildlife products at Belén and Modelo markets were conducted on an approximately weekly basis (alternating between weekday and weekend visits) over a 12-month period between September 2022 and September 2023. Two survey methods were used – open (cameras on open display and no restrictions on verbal engagement with vendors) and discreet (cameras not on public display and verbal engagements with vendors kept to an absolute minimum) – each of which was carried out by independent local field teams employed by in-country project partners including Instituto de Investigaciones de la Amazonia Peruana (IIAP). For both methods, inventories of individual stalls were limited to products offered for sale that were openly and publicly displayed; there was no attempt to search beyond that which was on display (with the exception of animals that could be seen underneath the stall). Surveys were carried out by two investigators who followed a delineated transect through the market, recorded observations in notebooks, and took images of species/products on mobile phones. Discreet surveys differed from open surveys only in that observations were recorded via hidden body-worn cameras to reduce the possibility of open recording affecting market activity. Species names were recorded as local, common, and scientific names (identification was verified by a taxonomist with expertise in Peruvian wildlife). Global and national threat status was noted for each species identified, in accordance with the IUCN Red List of Threatened Species (Version 2023-1, IUCN, 2023, Table 1 ) and the Peruvian Red List (published by SERFOR in 2018), respectively. Prices were recorded in Peruvian sols (PEN) and converted to USD using 1 PEN=0.26635933 USD (xe.com, 06.11.23). Full details of the field methods and data recorded are in Supplementary Data Sheet 1 .

No personal or identifying vendor data were collected. Stall locations were recorded only for assessing variations over time, individual stalls were coded in the database, and no data were linked to individual vendors to protect vendors from harm or discrimination (John et al., 2016). The database collated is entirely anonymous and is stored in a password-protected cloud storage platform, with access restricted to immediate project staff.

First, to provide an overview of the nature of the two markets, and the differences between them, data were summarized as ‘observations’ (the observed presence of a particular product of a particular species on one market stall, each unique species-product recorded only once per stall per survey). Observations reflect the availability, prevalence, and salience of species-products over the period of the study and allow broad descriptions and comparisons but do not necessarily relate – or relate consistently - to either the overall, or relative, numbers of individual animals involved.

Second, using rankings based on the number of observations of individual species, we identified species and species ‘groups’ that were most frequently observed. To provide baseline data that would allow detection of change in market dynamics we then quantified, for each of the most frequently observed species or species ‘group’, a number of variables (market metrics, Table 1 ) that collectively provide an index of the temporal availability of species-products and any seasonal changes over the year, their prevalence, an index of the number of individuals involved, and their market value.

Third, to provide as complete a picture as possible of the diversity of species traded at the market, and their uses, and, in particular, the prevalence of threatened species at the market, we provide a qualitative summary of threatened species (and products derived from threatened species) observed at the markets that were observed relatively rarely (i.e. all threatened species that were not included in the most frequently observed species, above).

All statistical analyses were carried out in R (R Core Team, 2023). Chi-squared tests were used to test for statistical associations between taxa, product type, and market, using simulated p values (based on 2000 replicates) for tests with low expected values, and effect sizes calculated using Cohen’s ω (in the “rcompanion” package) where a value of < 0.3 is considered ‘small’, and > 0.5 considered ‘large’ (Mangiafico, 2016). Rolling window correlations (estimated using Pearson’s correlation coefficient) were used to test for correlations between the number of individuals observed (as an index of trade volume) of each of the most frequently observed species groups at Belén Market and river water level (data obtained from the National Meteorology and Hydrology Service (SENAMHI), Peru; https://www.senamhi.gob.pe). Correlation coefficients were estimated in the R package “NonParRolCor” (Polanco-Martínez and López-Martínez, 2023), which takes account of multiple testing, for a window length of 55 to test for an overall association between the two time series, statistical significance was accepted at p < 0.05 and approximated using Monte Carlo simulations with 1000 replicates. Linear models were used to test the effect of species, type of meat (fresh or smoked), and market, on price, with post-hoc pairwise comparisons carried out using the “grafify” package (Shenoy, 2021) and p-values adjusted for multiple tests using Tukey’s method. Additional post-hoc tests are as stated in the text. Graphs were drawn using “ggplot2” (Wickham, 2016) and “gridExtra” (Auguie, 2017).

Over the duration of the study, we carried out 48 open and 49 discreet surveys of Belén market (hereafter Belén) and 50 open surveys of Modelo market (hereafter Modelo) ( Supplementary Data Sheet 1 ). The resulting dataset comprises 4,355 observations (3,498 at Belén, and 855 at Modelo) resulting from open surveys and 1,973 from discreet surveys (at Belén). The following analysis is based on data derived from open surveys; a comparison of species detected using the two methods is in section 3.5.

At least 66 unique species (27 Mammalia, 20 Reptilia, 17 Aves, and two Insecta; Supplementary Data Sheet 2 ) were observed by open surveys across the two markets (63 at Belén, 21 at Modelo). We did not detect any amphibians. All species observed are native to the Amazon. In total, 27.3% (n = 18) of species observed are threatened or Near Threatened at either a global or national level (28% [n = 18] at Belén, 33% [n = 7] at Modelo).

Across both markets, observations were primarily of mammals or mammal-origin products (hereafter mammals) (43.8% and 71.3% at Belén and Modelo, respectively) and reptiles or reptile-origin products (hereafter reptiles) (46.9% and 24.5%, Figure 2 ). Birds or bird-origin products (hereafter birds) comprised only 6.6% and 3.2% observations, and invertebrates 2.7% and 0.9%, respectively ( Figure 2 ). Observations were most often (67% at both markets) of species classified on the IUCN Red List as Least Concern; 24.2% and 28.5% (at Belén and Modelo respectively) were of threatened or Near Threatened species ( Figure 2 ). The products observed were predominantly food items (96.7% and 59.1% at Belén and Modelo, respectively); pets comprised 4.9% and 3.0% (respectively), and ‘other’ products 35.8% and 0.2% (respectively, Figure 2 ). Where ‘other’ included spiritual items (20% observations), ornaments or decorative items (10%) and musical instruments (0.5%; combined for analysis as “accessories’), and medicine (5%).

Across both markets mammals (89.6%) and insects (92.9%) were sold primarily for food, reptiles primarily for food (49.2%) and spiritual purposes (34.5%), and birds primarily for ornamental purposes (40%), and as pets (55.7%) (X² = 3074.3, df=4, p<0.001, Cohen’s ω~0.85; Figure 2 ). Effect sizes were small (Cohen’s ω≤0.32) for all other differences detected (details in Supplementary Data Sheet 2 ).

Animals sold live comprised 6.4% of all observations and involved 38.1% (n = 24) of the species identified. Most (71.8%) observations of live animals were for pets; 27.6% were for food. Pets were most commonly parrots/parakeets (69.8%) or turtles (24.1%) but also occasionally other species (including primates). Live animals sold for food were exclusively turtles/tortoises, with the exception of one observation of live weevil larvae. A preliminary assessment of the welfare conditions of the live animals sold as pets is in Supplementary Data Sheet 3 .

The nine most frequently observed species were (in order of occurrence) lowland paca Cuniculus paca, collared peccary Pecari tajacu, common caiman Caiman crocodilus, yellow-spotted river turtle Podocnemis unifilis, green anaconda Eunectes murinus, white-lipped peccary Tayassu pecari, yellow footed tortoise Chelonoidis denticulatus, red brocket deer Mazama americana, and the giant South American turtle P. expansa. The following quantitative summaries are based on eight species ‘groups’ (lowland paca, peccaries [Tayassuidae], caiman [Alligatoridae], river turtles [Podocnemis spp.], boas and anacondas [Boidae], yellow-footed tortoise, brocket deer [Mazama spp.], and parrots [Psittacidae]) that include these nine species, and with the inclusion of the parrots, collectively comprise 82% of all market observations (both markets combined). Comparable metrics are given separately for each species-product in Supplementary Data Sheet 4 .

Threatened and Near Threatened species that were observed infrequently (by open and discreet surveys) included globally Endangered spider monkey Ateles spp. (n=1 observation) and Amazon river dolphin Inia geoffrensis (n=16), Vulnerable giant armadillo Priodontes maximus (n=14), Amazon tapir Tapirus terrestris (n=33), and common woolly monkey Lagothrix lagotricha (n=19), and Near Threatened jaguar Panthera onca (n=52), Neotropical otter Lontra longicaudis (n=10), Illiger’s saddle-back tamarin Leontocebus illigeri (n=1), and ornate eagle-hawk Spizaetus ornatus (n=1), as well as puma Puma concolor (n=6) that are Near Threatened nationally, and Colombian red howler monkey Alouatta seniculus (n=6) that is Vulnerable nationally. The products that these species were used in, their prevalence, total number of items observed, price, and purpose, are in Table 3 (see also Figure 4).

Individual items cost between 3.19 USD for a jaguar skin bracelet and > 200 USD for a whole jaguar skin ( Table 2 ). Primate meat (fresh and smoked combined) cost 3.19-7.45 USD per kg, and smoked tapir meat cost 4.00-7.45 USD per kg ( Table 3 ). Fresh primate meat (woolly and red howler monkeys combined) cost statistically significantly less than the cheapest frequently observed fresh meat (common caiman, above) at Belén (one-tailed t-test: t_47.77=-3.91, p < 0.001) and both smoked primate meat (woolly monkey) and smoked tapir meat cost statistically significantly less than the cheapest frequently observed smoked meat (red brocket deer) (one-tailed t-test: primate - t_12.07=-4.96, p < 0.001; tapir - t_51.73=-7.66, p < 0.001).

Species-product observations equated to an estimated 2 – 3 individual jaguars, 4 pumas, a single Neotropical river otter, 1 – 6 individual giant armadillos, 1 ornate eagle-owl, 1 to 33 Amazon tapirs, 1 Illiger’s saddle-back tamarin, 2 spider monkeys, 10 red howler monkeys, and between 12 and 34 individual common woolly monkeys (depending on turnover and longevity of smoked meat), as well as an unknown number of river dolphins (in an estimated 63 bottles of oil) recorded over the duration of the study ( Table 3 ).

Forty unique species (17 Mammalia, 12 Reptilia, 10 Aves, and one aquatic invertebrate, Table 1 ) were identified by discreet survey methods. The taxonomic composition of species detected did not differ from that recorded during open surveys (X² = 0.060, df=2, p=0.970). Discreet surveys detected five genera that were not detected by open surveys; three of these were identified to species, one of which (the ornate hawk-eagle Spizaetus ornatus) is categorised on the IUCN Red List as Near Threatened (others are categorised as Least Concern, Supplementary Data Sheet 2 ). Discreet surveys missed 25 species (and 24 genera) that were detected by open surveys, including spider monkeys that are Endangered and Illiger’s saddle-back tamarin that is Near Threatened.

In this analysis we aimed to portray a picture of the nature of two markets in Iquitos that are heavily engaged in illegal wildlife trade. We did not attempt to summarize or quantify all species-products observed, instead we focused on threatened species because these are of most immediate conservation concern, and the most frequently observed species because these had sufficient sample size to allow precise estimates of market metrics. The latter was intended to establish a baseline for follow-on monitoring (see, e.g., Table 2 ), against which a number of different changes could be detected in response to legal enforcement or other conservation-oriented interventions.

Modelo is, as described, primarily a food market. Belén, however, was more diverse, with more than a third of observations of wild animal-origin products sold for ‘other’ (e.g. spiritual, ornamental, or medicinal) purposes. Across both markets, and both open and discreet surveys, approximately a quarter (26.8%) of the 71 unique species identified were threatened or Near Threatened, either nationally or internationally ( Supplementary Data Sheet 2 ). This was true of both markets despite species richness at Modelo being a third that of Belén. Four of the threatened species identified were amongst those most frequently observed: the globally Vulnerable white-lipped peccary, yellow-spotted river turtle, and yellow-footed tortoise, and the nationally Endangered giant South American river turtle, all of which were traded primarily for food. Trade, however, appeared to take place openly as surveyors did not see any sign of legal enforcement during market surveys, and there was little or no evidence that discreet surveys provided market data that could not be obtained from open surveys. Below, we outline some of the key findings associated with the different product purpose types traded – food, traditional medicine/belief-based and ornamental products, and live pets.

Six of the eight most frequently observed species groups (including rodents, ungulates, and reptiles), and four of the rarely observed threatened species (primates and tapirs), were sold either exclusively or primarily for food. This is not unexpected given that food is reported as the most common use of wildlife in Peru (Bodmer et al., 2004) and the most frequently stated purpose of the sale of wildlife by vendors at Belén and nearby markets in Iquitos (D’Cruze et al., 2021). Amongst the ‘meat’ products, smoked meat was more common (and was observed in higher volumes) at both markets than was fresh meat for all species except reptiles (caiman were sold primarily as fresh meat, and turtles and tortoises were predominantly sold alive). Smoked meat was presumably prevalent because it has an extended shelf life and keeps better during transport (Bolton, 2012; Buck et al., 2017) but perhaps also due to national or local preference (van Vliet et al., 2014).

Amongst the most frequently observed species groups sold for bushmeat, two - lowland paca and yellow-footed tortoise - were previously identified by Belén vendors as the species to be most profitable but that are also becoming increasingly rare (D’Cruze et al., 2021; see also Morcatty and Valsecchi, 2015). Lowland paca can occur at high densities (Emmons, 2016) and hunting of this species is considered to have been carried out at a sustainable level in some parts of the Peruvian Amazon (e.g. in the Itaya river basin, Aquino, 2009). However, they are one of the most consumed subsistence foods in Peru (Gallina et al., 2012), and elsewhere in the Amazonia region catch-per-unit-effort appears to be declining (Valsecchi et al., 2014). These observations suggest that the current level of use of lowland paca at Belén and Modelo market cannot automatically be assumed to be sustainable. Yellow-footed tortoises are, similarly, one of the most commonly hunted species across the Amazon (Tavares et al., 2020 and references therein) and a preferred species for both rural and urban communities (Tavares et al., 2020; Morcatty and Valsecchi, 2015). Peccaries are also considered to be threatened by overhunting (combined with habitat destruction), especially white-lipped peccaries that appear to be particularly vulnerable to hunting pressure (Keuroghlian et al., 2013). Peccaries are also widespread and traditionally a preferred source of bush meat in the Amazonian region (Gongora et al., 2011; Keuroghlian et al., 2013). Subsistence hunting of peccaries is legally permitted in Peru (Gongora et al., 2011). However, local extinctions of white-lipped peccary have been recorded in pristine habitats and in large, contiguous protected areas (Peres, 1996; Keuroghlian et al., 2013) and, in the Argentine Chaco, Altrichter (2005) found that white-lipped peccary density was three times lower in areas closer to villages than in protected areas. Our preliminary observations of the relative numbers of these two species at the markets suggest that in terms of availability of peccary meat the two species may be interchangeable. This suggests little incentive to ensure sustainable offtake of the more vulnerable species, and although price differences were small compared with price differences among species, the higher price commanded by white-lipped peccaries could lead to the more vulnerable species being targeted by hunters (cf. the anthropogenic allee effect, Courchamp et al., 2006). Whether the small premium in price is associated with the relative rarity of the species or a genuine preference for white-lipped over collared peccaries is unknown.

For lowland paca, yellow-footed tortoise, and caiman (but not peccaries or brocket deer), the numbers observed at the markets appeared to show some seasonality, with peaks (most pronounced for lowland paca) coinciding with peak river water levels. Similar seasonal patterns were noted by van Vliet et al. (2014), who also referred to a relative scarcity of bushmeat during the dry season. For lowland paca, which showed the greatest increase in numbers observed during high water levels, the seasonal ‘excess’ appeared to be sold primarily as smoked meat. It is not clear to what extent this might be driven by factors such as consumer preference, or lack of access to refrigeration. In contrast with these patterns, the numbers of river turtle eggs offered for sale appeared to be lowest when river water levels were high, which presumably reflects seasonal nesting behaviour of these species (Norris et al., 2020).

Primate meat (spider monkey, common woolly monkey, and Colombian red howler monkey) was observed only occasionally. All of these species were traditionally preferred species for subsistence hunting in the Amazon basin but all are now thought to have disappeared from some local areas, particularly those close to human settlements (Alves et al., 2021; Aquino et al., 2016; Stevenson et al., 2021; Link et al., 2021a, Link et al., 2021b; SERFOR, 2018). The low numbers observed in this study suggest that primates are rarely encountered by hunters and/or that they are taken opportunistically (Buck et al., 2017), and are in keeping with reports that hunted populations of spider and woolly monkeys in the Amazon basin have declined in recent decades (Peres and Palacios, 2007). Indeed, the illegal trade in wild primates for bushmeat and pets is considered one of the largest threats to this fauna in Peru (Shanee et al., 2017). In a recent consumer survey, none of 265 regular shoppers at Belén Market who stated that they frequently purchased wild meat at the market, mentioned purchase of primate meat (Moorhouse et al., 2023). There was also no evidence that the apparent rarity of these species equated to higher commercial value; however, bush meat vendors reported that woolly monkey meat is “in demand” and that it “reached the market less often than collared peccary or lowland paca meat” (Elwin et al., unpub. data).

In addition to food, medicinal, “belief-based”, and “decorative” uses for wildlife products are also considered profitable by vendors (D’Cruze et al., 2021). D’Cruze et al. (2021) referred to the use of parrot feathers as being particularly profitable amongst decorative products. The use of brightly coloured feathers has a long history in the Amazon region (Giuntini, 2006) and in this study macaw feathers were also commonly observed for sale as jewellery and other decorative items. Our observations of jaguar “accessories” is also in keeping with studies elsewhere within the jaguars’ range that found these products to be openly sold at local markets (e.g. Elwin et al., 2024), and domestic ownership of jaguar body parts for decorative, medicinal, and cultural purposes to be common (Arias et al., 2021). In contrast with Arias et al. (2021) we did not observe medicinal jaguar products. These types of products (and similar, e.g. necklaces adorned with an armadillo claw) may be bought by foreign tourists as souvenirs as well as by regular market customers (D’Cruze et al., 2021; Bodmer and Lozano, 2001; Braczkowski et al., 2019). For jaguar bones, it has been suggested that they may be used by the increasing Asian community in Latin America as a replacement for tiger bones in traditional medicines (Quigley et al., 2017; Morcatty et al., 2020). Observations of some other objects, such as otter, puma and caiman heads, suggested that actual sales were infrequent; these items were reported by some vendors “to attract customers to the business” and appeared to serve primarily as “shop-windows” ( Table 2 ).

Some potential species-products notably appeared to be absent, presumably because alternative (e.g. international) markets exist. For example, although considerable numbers of peccaries were hunted (evidenced by observations of meat for sale), no peccary (or tapir) pelts or peccary leather products were observed (although leather products of other species such as green anaconda were). Peccary skins can be traded legally by local communities and also exported legally under CITES permits for high-end products in the European leather industry (Keuroghlian et al., 2013). Tapir leather products are also sold internationally (Varela et al., 2019).

The impacts of trade in these types of products on wild populations are difficult to assess. In part, because it is difficult to equate products observed to the number of individuals killed. For jaguars, at least some body parts are most likely obtained from animals killed due to conflict with humans rather than being targeted for their use per se (Arias et al., 2021; UNODC, 2020). For other products, the species may be mislabelled (deliberately or not): recent molecular studies have for example revealed that products labelled as river dolphin are actually often from domestic pigs or sheep (Gravena et al., 2008 in da Silva et al., 2018).

Pets comprised a relatively minor component of wildlife trade at Belén and Modelo markets but involved a number of different wild species that varied in both prevalence and price. Parrots and turtles, for example, were observed relatively frequently with up to 23 individual parrots and 15 individual turtles per stall, whereas squirrel monkeys and Illiger’s saddle-back tamarin were observed only once and each as a single individual. Similarly, whilst some species (notably the green anaconda) commanded prices of up to 79 USD or more (equivalent to more than three day’s salary for an urban worker in Peru, based on average monthly income in December 2023 of 1,581.50 Peruvian soles [INEI, 2023] and 19-22 working days in a month), others (e.g. hatchling turtles and yellow wing parakeet) cost less than one USD, considerably less than a kilogram of meat or a single serving of a cooked turtle meat meal at the market ( Supplementary Data Sheet 4 ). A “baby” brown-throated sloth could be purchased for 5 USD. Regueira and Bernard (2012) similarly reported birds sold at urban street markets in northeastern Brazil for one US dollar. Both extremes have implications in that at the upper end they suggest significant financial incentives for hunters and selective pressure on particular species (cf. Tella and Hiraldo, 2014; Romero-Vidal et al., 2023) whilst at the lower end they highlight the lack of ‘value’ attached to wild animals in this region.

The pet trade is considered a serious threat to many Neotropical parrot species (Berkunsky et al., 2017) with the potential impacts of market trade and internal exports compounded by local poaching for self-supply or local scale rural sales (Romero-Vidal et al., 2023). An apparent decline in numbers of Illiger’s saddle-back tamarin is also attributed in part to hunting for the pet trade, which is believed to have increased in the last decade (Heymann et al., 2020). We did not attempt to estimate turnover rates but note that the lack of repeat observations of the squirrel monkey and the Illiger’s saddle-back tamarin (and, similarly, of two toucans observed on two consecutive occasions) suggest that the animals were sold. Others have observed significantly larger trade volumes in wild animal pets in Peru – particularly primates (Shanee et al., 2017), although, notably, none of the vendors in D’Cruze et al (2021) study named primates for sale live as pets as a profitable endeavour. Many of the animals sold as pets were very young – hatchling turtles, “baby” sloths and parrots – presumably because they are both easier to keep and to capture (and are more appealing) than adult animals. For turtles, individuals that were too small to be sold for food appeared to be sold as pets. In contrast with patterns observed for some of the species traded as meat, the numbers of live parrots observed for sale were lowest during peak river water levels; it is not known whether this was due to seasonal changes in the abundance of parrots in floodplain forests (e.g. Lee and Marsden, 2012) or changes in the behaviour or activity area of the hunters during the wet season.

With regards to animal welfare, whilst hunting and killing animals for food or other products almost certainly involves some level of animal suffering, the suffering experienced by animals traded live may be greatest because there is potential for poor welfare impacts during all stages of the trade chain (Baker et al., 2013). Our welfare assessments were preliminary and pertain only to the period while the animals were on display at the market but nevertheless suggest that conditions, in most cases, were poor. Most of the live animals on sale as pets were held in metal cages or plastic containers with bare floors and although the majority appeared to be healthy and in “acceptable” hygiene conditions, most were overcrowded, lacked shelter or enrichment and did not appear to have food or water available ( Supplementary Data Sheet 3 ).

In terms of generating a robust baseline for future comparative market assessments, it is important to note that our survey methods may have failed to detect a ‘hidden’ component of market trade (Bušina et al., 2020), and it is possible that our dataset is biased towards less threatened species. The openness with which wildlife is traded in urban areas of Peru, much of which takes place illegally, suggests that trade, in this case, is not generally hidden but it is possible that high value items are hidden to avoid confiscation by the authorities (Daut et al., 2015; Romero-Vidal et al., 2023 and references therein) or conducted through alternative channels such as social media (Siriwat and Nijman, 2020). We were also unable to identify many of the bottles of oils observed, and so may have underestimated the diversity (and perhaps threat status) of the species used for medicinal purposes in particular. This may explain why, for example, we did not record medicinal jaguar products, in contrast with Arias et al. (2021). Similarly, for all but a few of the least frequently observed species we can only reliably quantify trade volume as it was observed during each survey. We do not know, for example, to what extent fresh meat is restocked through the day (cf. Bušina et al., 2020). Although the large number of repeat surveys over an entire year and across season lends weight to the accuracy and precision of the various market metrics defined (and provides a measure of annual variation in metrics that might be observed in the absence of any external change) it is important that these data are recognized as providing a ‘snapshot’ (or index) of the trade that presumably occurred over a full week.