Ninety-eight taxa of wild edible plants were identified in Questionnaire 1 and Questionnaire 2, seventy-one of these species were confirmed by herbarium specimens (see Appendix). The category of fruits and seeds was the most highly represented (35 species and 45% of all the species × locality records). Green vegetables (any species whose leaves, shoots or unripe fruits are consumed in quantities larger than for flavouring) were represented by 31 species (26% of species × locality records), beverages (infusions, sap and wine) by 26 species (9% records), flowers eaten for their sweet taste by 8 species (2%), spices/flavourings (any parts of plants) – 6 species (5%), and underground parts used for food – 6 species (3%).

The most commonly listed plants were two green vegetables: Rumex spp., appearing in 75 localities, and Chenopodium album, in 50 (Table 3). Next in the ranking were three taxa bearing fleshy fruits: Rubus subgenus Rubus ('blackberries'), Fragaria vesca and Vaccinium myrtillus, followed by another green vegetable, Urtica spp., two species bearing fruits: Vaccinium vitis-idaea and Rubus idaeus and then a children's snack, Oxalis spp. (Table 3).

The sour-tasting leaves and shoots of Rumex spp. were commonly collected throughout the country to make soup. They were sometimes stored for winter, pressed tightly with salt in sealed bottles. They were also the commonest children's snack. Chenopodium album, Urtica dioica and U. urens were described, in most questionnaires, as poor people's or famine food used only until the beginning of the 20^th century or during World War Two. Still used in several localities at the time of study, they were, however, associated with old-fashioned habits or poverty and collected mainly by elderly women. Chenopodium leaves were usually briefly boiled, drained, fried and then garnished with cream or milk. Urtica leaves were used in a similar way (often together with the former) but they tended to be used in soups.

The fruits used in at least a quarter of the studied localities were: Rubus subgebus Rubus, Rubus idaeus, Vaccinium myrtillus, V. vitis-idaea, Fragaria vesca, Corylus avellana, Rosa spp., Prunus spinosa and Sambucus nigra. The following fruits were used in fewer localities, but their use must still have been widespread throughout the country (they were recorded in at least nine localities): Crataegus spp., Prunus padus, Ribes spp., Vaccinium uliginosum, V. oxycoccos, Fagus sylvatica and Sorbus aucuparia. Most wild fruits were collected by women and/or children (Table 4), eaten raw or made into juice, jam or wine. Fruits were often stored for winter with added sugar (informers often complained about the high price of sugar, particularly when describing the use of Vaccinium vitis-idaea). Vaccinium myrtillus fruits were also dried. Corylus, Fagus and Prunus padus were mainly eaten raw.

A very limited range of wild plants was used as spices and flavouring. Carum carvi seeds were used to flavour bread and sauerkraut, and grated Armoracia rusticana roots were added to meat, boiled beetroot or other salads. In the southern and eastern part of Poland mint leaves (Mentha spp.) were used to flavour dumpling filling or soups.

Apart from foods, the questionnaires contain information on herbal infusions drunk by healthy people on an everyday basis, these included two taxa used commonly throughout the country: infusions of Tilia cordata flowers and roasted Quercus robur acorns (use of the latter was nearly obsolete in 1948), and taxa used rarely and locally on an everyday basis (used more often medicinally): Rubus subgenus Rubus, Rubus idaeus and Fragaria vesca leaves, Mentha arvensis, Chamomilla recutita, Galium odoratum, Viola tricolor, Thymus pulegioides and Thymus serpyllum flowering shoots. In Maciowakrze, near Opole, Rumex obtusifolius leaves were added to compotes.

The studied questionnaires contain information on the use of species not listed in the review of edible plants of Poland 1. These are: Rumex thyrsiflorus (leaves), R. obtusifolius (leaves), Ribes spicatum (fruits), Cirsium arvense (leaves), Lamium album (flowers and leaves), Stratiotes aloides (whole plants), Galium odoratum (flowering shoots), Sedum maximum (underground parts), Parthenocissus sp. (fruits), Stachys palustris (rhizomes), Tragopogon pratensis (shoots), Viola hirta (roots), V. tricolor (flowering shoots for tea), Viscum album (fruits), which were confirmed by herbarium specimens, whereas Chamomilla recutita (flowering shoots for tea), Papaver rhoeas (seeds), Primula elatior (flowers) and Ulmus sp. (fruits and bark) were identified on the basis of folk names and additional information in the questionnaires.

According to the results of this study it was mainly women and children who collected wild plants. For the 12 most commonly collected taxa, the following categories scored around 30% in question number 10 (who collects the plants): 'children and women' (32%), 'everyone' (32%) and 'only children' (24%) (Table 4). When we look at particular species the above mentioned ratio is usually similar, with a few exceptions: Oxalis spp. collected mainly by children, Corylus avellana collected mainly by men and children, and Urtica spp. which were not collected by children, but by adults, mainly women (Table 4).

When rarer taxa are included, the proportion of answers 'collected only by children' is similar (21%, 81 out of 389). It must be stressed however that although children were the most important plant collectors, they also collected plants for the adults, as the proportion of answers 'eaten only by children' is lower, at only 5% (20 out of 389).

The commonest children's snack, Oxalis spp., ranked just between such important wild crops as Rubus idaeus and Corylus avellana. Other commonly collected species, which were eaten mainly by children were Prunus spinosa, Crataegus spp. and Rosa spp. (these three taxa were, to a lesser extent, also used to make preserves). Other species of predominantly children's snacks included: young shoots of Acorus calamus, a few species of flowers, whose nectar was sucked, particularly Trifolium pratense, T. repens, Robinia pseudoacacia, Lamium album, Symphytum officinale, fruits of Prunus padus, Rubus saxatilis, Maianthemum bifolium, Frangula alnus, Malva spp., Capsella bursa-pastoris and the sweet rhizomes of Polypodium vulgare. Surprisingly, in a country where children have always been discouraged from drinking alcohol, several respondents wrote about children independently making a kind of "wine", particularly with the petals of Centaurea cyanus, which were fermented for some time with water and sugar (12 reports from various regions). The inventory of children's snacks seems to be uniform across the country with very few regional differences.

The mean number of species used in a single locality (calculated from lists longer than 2 species) was 11.3, the modal value was 10. There was high variation in the number of species used per locality (SD = 7.5). The longest list, 39 species, was recorded in Barwino, near Słupsk (Pm). This village used to be a part of Germany before 1939, so most of the inhabitants were removed after World War II, and the list was based on interviews with four newcomers from the area of Łuck (present Belarus). The second longest list, 31 species, was recorded in eastern Poland, in Wola Osowińska, near Łuków. Lists ranging from 20 to 30 species were recorded in localities from a few regions of Poland (mainly E and S, but also NE and NW), with the exceptions of central, western and south-western Poland (Fig. 2).

The mean number of species per locality was highest in the Carpathians (12.5 ± 1.7 SE), lower in lowland Poland (11.4 ± 1.0 SE) and lowest in Ziemie Odzyskane, the areas reclaimed from Germany (10.4 ± 2.1 SE). The difference was not significant between any of the pairs of the three above mentioned parts of Poland (Mann Whitney U test, P > 0.05) Differences were larger if we look at modal values, which were 10 both for the Carpathians and lowland Poland and only 7 for Ziemie Odzyskane, since the mean species number per locality for this area was elevated by a few very species-rich lists.

There was no correlation between latitude and the number of species listed (Spearman rank correlation coefficient, rho = 0.01, P = 0.89), however there was a small but significant correlation between the longitude and the species number (Spearman rank correlation coefficient, rho = 0.29, P = 0.007). Almost identical results were obtained when localities from Ziemie Odzyskane, dominated by immigrant populations, were not included in the calculations, although, in this case, neither of the correlations were significant (rho = -0.004, P = 0.98 for latitude and rho = 0.23, P = 0.07 for longitude).

The use of more common species did not show any strong geographical patterns. The use of some species recorded rarely (1–3 localities) was probably more regionalised. A strong regional pattern can be noticed only in the distribution of a few species. Ribes fruits were gathered from the wild only in eastern and northern Poland, Polypodium vulgare rhizomes were used only in the Carpathians, Stratiotes aloides was used only in central Poland (Fig. 3) and Malva spp. were used predominantly in western Poland.

The herbarium specimens made available help to answer questions which were raised by Łuczaj & Szymański's review 1: they reveal the exact proportions of particular species contained in folk taxa, which were impossible to estimate from descriptive ethnographic works.

Within the folk taxon szczaw the proportion of Rumex acetosa, R. acetosella and R. thyrsiflorus is 16:4:1. The latter species has never been mentioned in ethnobotanical literature before. It is easily confused with R. acetosa, it is equally large, however it flowers later and grows in dry, sandy soils where R. acetosa is not found.

Some publications 112 suggest that within the folk taxon lebioda/łoboda a variety of Chenopodium and Atriplex species were used. However all 16 herbarium specimens from this taxon belong to Chenopodium album.

The herbarium specimens confirm that both nettle species (Urtica dioica and U. urens) were used (ratio 9:4).

All previous ethnographic publications put an equation mark between szczaw zajęczy and Oxalis acetosella, however specimens of Oxalis stricta s.l. were also found in the analyzed material (ratio 9:2). O. stricta is an alien species, which occurs mainly on arable land, so it may have been, in some villages, more available to children than the woodland species, O. acetosella.

Most mint specimens were unidentifiable, however both Mentha longifolia and M. arvensis were found in the material. Their use had been reported before 1, but only in single locations. The majority of wild mints used as cooking herbs or for infusions must have belonged to these two species.

The use of both species of birch, Betula pendula and B. pubescens was confirmed (ratio 4:1).

Crataegus and Rosa are difficult to distinguish in ethnographic materials, as they are often called głóg, and were used in a similar fashion. All the identified herbarium specimens from these genera belong to Crataegus monogyna (6), whereas the five specimens of Rosa are probably R. canina, but full identification is impossible due to the lack of fruits.

All the specimens called mlecz (the scientific name for Sonchus) belong to the genus Taraxacum, and not Sonchus, which creates a suspicion that previous reports on the use of Sonchus are botanical mistakes.

Both Malva sylvestris and M. neglecta were called the same folk names, the ratio of their specimens is 3:1.

Due to the lack of flowers and fruits in the preserved Rubus specimens it was not possible to identify particular species, except for Rubus caesius, which constitutes a surprisingly large proportion of specimens (ratio 6:7). Its popularity as a food plant, although its berries are some of the smallest and sourest of the blackberry species, can be attributed to the fact that in many areas of Poland (especially NE) other Rubus subgenus Rubus species are rare 13.

Out of the newly recorded species, probably the most valuable finding is the discovery of the use of Stratiotes aloides. This water plant, although it occurs over a large area of Poland, was an important famine plant only in the Łódź area (Glinnik, Zamość and Zarzęcin), where, until the turn of the 19^th and 20^th century, it was commonly collected from the bottom of lakes and cooked.

In Mszana Dolna (W Carpathians) Viola hirta roots were eaten as a children's snack, and called słodkie korzenie ('sweet roots'). Such properties of V. hirta have never been reported before, and as this information is based on one herbarium specimen, it needs further confirmation.

Stachys palustris, recorded in the village of Biały Dunajec – Stołowe (W Carpathians) is another famine plant previously under-recorded by Polish ethnographers. It was not listed in Łuczaj and Szymański's review 1, it was, however, recorded by Rostafiński 14 in 1888.

The studied questionnaires also contain information on a third locality in the Western Carpathians (Rabka) where Heracleum sphondylium was used to make soup, until the 20^th century.

A few of the presented plants have not been considered as edible in Europe, but were consumed in other parts of the world: related Maianthemum species fruits and Oxalis stricta leaves by Native Americans 15, Convolvulus arvensis s.l. shoots and Ulmus fruits by the Chinese 16 and Viscum fruits in Nepal 17.

In a few dozen records neither scientific species or genus name could be matched, sometimes due to the fact that a folk name is commonly used to describe two genera, e.g. 'babka' for Plantago and Malva and 'głóg' for Crataegus and Rosa. The use of the seeds of a plant called anyżek (literally 'little anis') was reported in a few places. This may be Pimpinella saxifraga or some other aromatic Apiaceae plant. Near Nowy Targ (Mp) the use of some underground bulbs called orzechy ziemne ('earth nuts') in Szaflary and ziemne jabłka ('earth apples') in Frydman was recorded. This is most likely Lathyrus tuberosus, which was earlier recorded in SE Poland in the Rożnów area 18 or Helianthus tuberosus. In Zelczyna near Kraków, koniczyna wodna ('water clover') was eaten. This folk name may refer to either Menyanthes trifoliata or Marsilea quadrifolia.

It is practically a cliché that in hunter-gatherer societies it was mainly women who gathered food, while men hunted 19. As far as the children's contribution to subsistence effort is concerned, it varies. In some hunter-gatherer societies, such as Hadza, children's participation is important, whereas among !Kung they contribute little to gathering 2021. In the mid-20th century Polish countryside, where both men and women were strongly involved in farming practices, children, outside school hours, served as shepherds, and were the group in society which had the most contact with wild plants, often even replacing women as the main gatherers. The transmission of traditional ethnobotanical knowledge in such circumstances is an interesting issue. We can presume that traditional knowledge transfer was horizontal, with older children showing new plants to the younger and this conclusion can be supported by the authors' experiences from field interviews with older people. As children usually ate plants raw, this horizontal transfer mainly concerned plant recognition, whereas traditional knowledge on the preparation of cooked and fried dishes (jams, juices, soups) was probably passed vertically from mothers to daughters, as in many traditional societies 22. Such a mixture of horizontal and vertical transfer of knowledge has been well documented recently in Thailand 23.

A large proportion of plants, eaten around 1948 mainly as children's snacks, are probably forgotten articles of adult food, e.g. Oxalis leaves, which were used to make soup, and Trifolium flowers, which were eaten as famine food 1. Thus children's interest in snacking on wild plants had an adaptive value: these plants could be used in larger quantities in times of food shortages. The repertoire of children's snacks constituted a reserve list of edible plants for the community. Sometimes they were just plants which did not occur in large quantities or were time-consuming to collect (e.g. Equisetum arvense and Lathyrus tuberosus bulbils, Oxalis leaves, Polypodium rhizomes) but could be used in emergencies or if they became more abundant. The part of plants used by children may differ from the most nutritious part (e.g. Malva fruits used by children versus leaves cooked by adults, Lamium album flowers versus leaves) but the notion that a given species was edible was preserved.

The knowledge that certain plants are edible and tasty, even if they remain exclusively a children's snack and no famine occurs for decades, could probably have been maintained through a few generations in traditional rural communities. Children spent a lot of time outdoors, served as shepherds and helped parents in collecting wild plants. Trying different tastes must have been exciting for people raised on the bland staples of potatoes, cereals and dairy. However within the last two decades since the fall of Communism in 1989, a larger drop in the knowledge of wild edible snacks has occurred among Polish children, according to the author's preliminary observations. This process can be easily explained by migrations from rural areas, the growing choice of food articles available (e.g. exotic fruits) and the increasing length of time spent by children indoors (e.g. when watching television or using computers).

Assuming that the lists in particular localities come close to representing the total traditional knowledge of a village community, they can be compared with parallel studies from other countries. In one area in Italy with over seven thousand inhabitants, 44 species are known 24 and in another village of southern Italy, Castelmezzano, with less than a thousand inhabitants, the use of 60 species of edible plants was recorded 25. In a small traditional community of Mapuche Indians in temperate parts of Argentina (with climate similar to this of Poland), 24 wild edible plants species are known 26. The results of this study, although they extend the list of edible plants of Poland considerably, still document relatively low levels of traditional knowledge in 1948, as on average only 11 species of plants (mainly common edible fruits) were listed, with the longest list consisting of 39 species. This avoidance of wild plants in the Polish diet, except for fruits, was extensively discussed in a previous article, where it was attributed mainly to cultural factors 1. This study supports the cultural hypothesis (rather than putting blame on past researchers' neglect), as it is fruits and not green vegetables that make up the largest category, in sharp contrast to southern European countries, e.g. Spain 27, Bosnia and Herzegovina 28, or the central and southern part of Italy 2930. Poland is more similar to northern Italy, where the eating of bitter green vegetables is not popular 29. Moreover the recorded use of spices and cooking herbs was extremely limited (5% of records) and incomparably lower than in southern Europe. The modern gathering habits analysed in another study 31 just a few years ago show a drastic reduction of wild food plants collected, particularly the non-fruit component, when compared to the results of the study carried out in the 1960s 31 and the results presented here. There seems to be a strong pattern of avoidance of using the green parts of plants, particularly strong tasting ones, as either vegetables or flavouring, during periods when food is not scarce. It is a matter of discussion whether this pattern could be referred to as 'herbophobia', a term analogous to Wasson's 'mycophobia' which commonly refers to the almost total absence of fungi in traditional English and German cuisine 3233. Although the use of green vegetables constituted 26% of all records, half of them concerned plants characterised by respondents as famine or poverty food and three quarters of all the records for green vegetables were made up by only four genera (Rumex, Chenopodium, Urtica and Oxalis). Among the twenty most frequently used taxa (Table 3), there are only five taxa whose green parts were used: two famine vegetables, two raw children's snacks and only one taxon – Rumex, perceived as a normal, non-famine cooked vegetable. In contrast to this, in the list of twenty culturally most significant wild plants in Garfagnana, Italy, even a few decades after 1948, as many as seventeen taxa were green vegetables or aromatic herbs. The term 'herbophilia' could apply to such cultures as that of Garfagnana and other parts of central and southern Italy, as well as China and Japan, in which the green parts of plants of numerous species are often used and highly prized 162934. On the other hand the term 'herbophobia' may not be fully adequate as it implies a phobia (an irrational, intense, persistent fear of something), whereas the phenomenon described is more of a disappearance of wild green vegetables from the diet, linked to their low cultural significance and associations with poverty (but not the devil, as was the case in some countries for mushrooms 33). So maybe a term 'culinary a-herbia' (or 'aherbia') would be more appropriate for cultures which display no interest in using larger numbers of species of wild vegetables and culinary herbs? The adjective 'culinary' is needed as the green parts of plants have been used in Poland widely, mainly as infusions, in a medicinal context 6. It must be stressed that the occurrence of culinary aherbia/herbophilia and mycophobia/mycophilia is not necessarily correlated. Poles are strongly mycophilous 33 and 'fructophilous' but have historically largely neglected the use of green vegetables and culinary herbs, whereas other national cultures exert different patterns, an issue which needs further study. Utilisation of a large number of species of green vegetables is a characteristic feature of agricultural communities, particularly those in which food shortages are frequent. In such societies the utilization of weeds as food provided extra calories and made space for the growth of main crops. Once the danger of famine is removed, some societies reject green food as a symbol of famine, others preserve at least some of the "famine" vegetables as traditional foods or food additives.

The possible reasons why Poles have used few green parts of wild species have already been pointed out by Łuczaj & Szymański 1. Most edible plant species in the Mediterranean are used as spices, salads or appetisers, not as staple foods 2730. Many of those which are found as common plants in Poland (e.g. Thymus spp.), have hardly ever been used in Polish cooking as spices, although they are often listed by ethnographic sources as medicinal plants throughout Poland 6. The primary reason for the difference in attitudes towards herbs between Poland and the Mediterranean may be climate. In warmer climates the addition of herbs to meats, dairy and sauces kept them from going off, whereas in the Polish temperate climate there is less need for this. Hence "pure", refined foods like white sugar, white bread and pure good quality meat were most highly prized, and wild plants, apart from fruits and mushrooms, were associated only with times of famine and seasonal spring food shortages. Another reason is that Poland, a mostly flat country with reliable rainfall, has been a thoroughly agricultural country with a large proportion of arable land, where vegetables could be easily cultivated, whereas the countries of the Mediterranean Basin are very mountainous, with a large proportion of land covered by stony semi-arid pastures where cultivation of vegetables is difficult and wild plants could have been a valuable addition to the pastoral economy.

The results of this study undermine the conclusions of a recent review of edible plants of Poland 1, which stated that many more wild edible plants have been collected in the Carpathians, particularly in their western part, than in lowland Poland. Now it can be clearly seen that such results were caused by the considerably larger number of ethnographic studies undertaken in the Carpathians, since this study proves that places where large numbers of edible plants were collected existed in a few parts of Poland in the mid-20^th century (Fig. 2), not only in the Carpathians, but along the eastern border as well as in Polish communities originating from the expanded Soviet Union (Lithuania, Ukraine and particularly Belarus), which had been resettled to the west of Poland. We can imagine the enclaves in which traditional knowledge about edible plants persisted as less-developed rural areas within biodiverse regions, or as families particularly oriented towards gathering. The high heterogeneity of species richness of the plant inventories obtained may have been caused by individual differences among informers, differences between locations and differences in the effort put in by the researchers to obtain information.

The longitudinal pattern observed in species richness of the studied questionnaires is typical of many ethnographic phenomena in Poland 23. The west of Poland was strongly influenced by Germany, its industry and modernisation of farming practices, remaining under Prussian occupation from the end of the 18^th century until 1918. The east and the south constituted the outskirts of the Russian and the Austro-Hungarian Empires and were less affected by modernisation. Hence more traditional folk culture was preserved in the eastern part of Poland.

Since 1948, a gradual decrease in the use of wild food plants has occurred in Poland and the data presented here are of historical character. At the time of data collection (1948) people still remembered the use of some famine plants, the soup of Rumex acetosa leaves was made in nearly every village and wild fruit preserves for winter were made in a large proportion of households in the countryside. As recent studies by the team of Polish Ethnographic Atlas showed 31, at the beginning of the 21st century the gathering of wild food plants has become restricted to a few individuals particularly interested in this kind of activity. In Jędrusik's thesis 31, data on the gathering of wild plants and mushrooms from 82 villages, in 1964–68 and in 2000–2003, obtained using similar questionnaires, were compared. The number of villages where they were collected in 2000–2003 had decreased to a fraction of the number in the 1960s (e.g. for Vaccinium myrtillus from 77 to 21, Armoracia rusticana from 67 to 17, Rumex acetosa from 59 to 19). The gathering of some species had stopped entirely (e.g. Prunus spinosa or Vaccinium uliginosum). In contrast to these data the frequency of mushroom collection has not changed much in the last few decades, apart from a change in the way they are stored for winter (a shift from drying towards pickling and freezing).

Each entry has the following format:

Latin name 'H' + number of identified herbarium specimens (number of recorded localities + 'loc.') – parts used, preparation methods; list of localities for some interesting species; LN: all recorded local names (starting from the commonest).

Each locality is expressed as: "village name (nearest town name, region code)". Regions (województwo) were coded as follows:

Sl – śląskie, Mp – małopolskie, Pk – podkarpackie, Lu – lubelskie, Sw – świętokrzyskie, Łd – łódzkie, Mz – mazowieckie, Ps – podlaskie, Wm – warmińsko-mazurskie, Pm – pomorskie, Kp – kujawsko-pomorskie, Wp – wielkopolskie, Zp – zachodniopomorskie, Ls – lubuskie, Ds – dolnośląskie, Op – opolskie.