You Reap What You Sow: A Botanical and Economic Assessment of Wildﬂower Seed Mixes Available in Ireland

: Planting wildﬂower seed mixes has become popular with individuals and community groups aiming to promote wildlife and enhance local biodiversity. Recently, however, these seed mixes have been criticized with respect to the origin of the seeds and the species they contain. There is a growing awareness that the unintended planting of exotic species may disrupt native ecological networks, introduce aggressive weeds, or facilitate the establishment of invasive species in new localities. In this study, we purchased two packets of twelve brands of wildﬂower seeds available in Ireland from stores or online suppliers. In total, the 24 packets contained 69,409 seeds weighing 304 g, and represented 92 plant species in 23 families. Only 25% of the seed packets purchased in Ireland originated from Ireland, and only 43% of the plant species we identiﬁed are considered native to Ireland. To reinforce this point, the most frequent species, Phacelia tanacetifolia , which occurred in nine of the twelve brands, is not a native Irish plant species. Multivariate analysis identiﬁed no obvious grouping of seed mixes based on their intended target group (e.g., bees, butterﬂies, wildlife), which might be expected had manufacturers followed scientiﬁc guidance describing which plants are preferred by which pollinator group. The creation of patches of diverse ﬂoral habitats in gardens and urban settings can signiﬁcantly beneﬁt wildlife and human wellbeing. Our results, however, reinforce the need for caution before using wildﬂower mixes in attempts to restore or recreate natural or semi-natural plant communities.


Introduction
There have been multiple recent reports describing significant global declines in wild insect pollinators, with accumulating evidence of decreased abundance, shrinking geographical ranges, and reduced species diversity [1][2][3][4]. At a European level, declines in hoverflies (Syrphidae), butterflies (Lepidoptera), and bees (Apidae) have all been identified [5][6][7]. The outlook for Ireland appears similarly bleak, with 50% of bumblebee species and 45% of solitary bees showing evidence of waning population sizes and/or distribution [8]. The threats contributing to pollinator declines are numerous and inter-related, but include habitat loss, agricultural intensification, agrochemical pollution, and the introduction of pests, diseases, and invasive competitors [3,[9][10][11]. Loss and fragmentation of natural and semi-natural habitats can reduce, modify, or isolate nesting and floral resources, and result in habitat or diet specialists, or species with smaller foraging ranges, being unable to adapt to new conditions [12,13].
In order to reverse the current trends in pollinator declines, strategies involving multiple complementary approaches, including the re-establishment of semi-natural habitat and provision of dietary supplements, are generally considered to be the most effective. In Ireland, for example, the All-Ireland Pollinator Plan (AIPP) now lists almost 200 sciencebased actions that can be performed by all sectors of society (e.g., individuals, community groups, businesses, regional councils) to help maintain and enhance pollinator communities [14]. A key component of many pollination conservation initiatives is the creation of areas of habitat that contain a high abundance and diversity of flowering plants. These florally-rich patches or strips deliver a variety of nectar and pollen, provide host plants for lepidopteran larvae, and supply nesting resources such as empty stems and foliage required by stem-nesting bees.
In agricultural settings, pollinator conservation measures include the preservation of hedgerows, the creation of florally-rich meadows, and increasing the floral diversity of forage crops by the inclusion of clovers and other flowering plants in multi-species swards. The creation of wildflower strips in field margins and headlands is often specifically promoted via agri-environmental schemes, and these field edge flower plantings have been demonstrated to have clear positive effects on pollinator abundance, diversity, and reproduction [15][16][17]. Additional ecosystem service benefits may also occur because of enhanced pollination of the focal crop, and the presence of other beneficial invertebrates, such as parasitoid wasps and predatory beetles, that can reduce infestations of crop pests [18,19]. By enhancing pollinator nutrition, it has been suggested that floral plantings can also help mitigate the negative impacts of pesticide exposure encountered in agricultural landscapes [17].
In domestic or urban settings, areas of florally-rich habitat can be created using blends of flower seeds that are now readily available from commercial seed merchants, gardening centres, general stores, and online sellers. These seed packets are affordable and convenient: they usually contain a wide variety of seeds and eliminate the need to spend large amounts of time and money purchasing individual plant species. Members of the public or community groups motivated to become involved with pollinator conservation have readily embraced these mixed seed packets as simple and effective habitat-creation tools [20,21], and there is ample evidence that the florally-rich resources produced by sowing these seed blends enhances pollinator abundance and diversity. For example, Blackmore and Goulson [22] recorded 25 times more flowers, 50 times more bumblebees, and 13 times more hoverflies in urban areas planted with wildflower seeds compared with mown grassland. Similarly, Griffiths-Lee et al. [23] found that sown 'mini-meadows' in domestic gardens resulted in double the number of bumble bees and solitary bees. As an indicator of how popular wildflower seed mixes have become, packets are often given away free as a means of engaging the public with conservation initiatives [24].
More recently, however, several reports and online articles have raised concerns with the uninhibited, 'blanket' use of mixed seed blends, and the potential ecological and agricultural harm they might cause [21,[25][26][27][28][29]. The wording on these seed packets generally suggests they contain 'wildflowers', and that the plants produced are wildlife friendly, a theme reinforced by images of large, colourful flowers along with bees, butterflies, and other insects on the packaging [30]. This labelling can, however, be misleading, and is likely designed to entice well-intentioned consumers aiming to create wildlife-friendly habitat in their gardens [21]. In an Irish setting, it has been highlighted that many mixed seed packets are imported, contain species that are not native to Ireland, and/or the seed is not sourced in Ireland [21,26,28]. This situation can result in problems for both insects and plants: native populations of pollinators have generally evolved to interact with native plant species, and seeds with local provenance will tend be better adapted to local conditions and cope better with local pests and diseases [21,25,31]. Even when the packets contain native plants, sowing these species in areas where they do not normally occur can obscure natural habitat associations and biogeographic patterns. Similarly, although the species might be native, the seed might be of ornamental or agricultural cultivars, and bred for vibrant or unusual flower colour, flower size, or seed productivity, compared with native wild forms [21,29]. Finally, even when all the seeds are of native (or naturalized) plants, the subsequent species assemblage they produce might never be found under natural conditions [27,29,30]. In the extreme case, seed packets have been recalled due to the discovery of highly invasive weeds, such as blackgrass (Alopecurus myosuroides), that can be both a serious agricultural pest and threaten native flora [25,27,32].
There are now several publications listing species of flowering plants that are attractive to pollinating insects, including plant species used in agri-environment schemes and those intended for smaller-scale garden or community use; many of these lists still include both native and non-native plant species [33][34][35][36]. Often these plant lists specify which flowers are attractive to different pollinator taxa, usually based on what plant species or families appear to induce higher visitation rates, or increase the abundance of honey bees, bumble bees, hover flies, butterflies, and so on [23,[37][38][39][40]. It might be hypothesized, therefore, that if seed mix producers were following this scientific evidence, then different brands of seeds aimed at producing habitat that is attractive to the same pollinator group would have seed blends of similar species compositions.
The primary objective of this study was to provide an assessment of the contents of different brands of wildflower seed mixes readily available in Ireland. The different brands were compared in terms of seed weight and seed count, the number of species present, and, subsequently, their value for money. By determining the species composition of each packet, we were able to determine which flowering species occur most frequently, and the degree to which these wildflower mixes contain species that are native to Ireland. Finally, to assess whether there was any evidence that seed packagers were producing blends based on scientific evidence, we used multivariate analysis to assess whether seed mixes that indicated similar conservation aims on the packaging contained seed blends with similar species compositions.

Seed Packets
To locate different brands of commercially-produced wildflower seed mixes available in Ireland, we visited garden centres, supermarkets, general retail stores, and online sellers. The aim was to buy small, individual packets of seed intended to cover small patches of ground in private or community gardens. Therefore, large boxes of seeds, or quantities aimed at agricultural or landscape use, or for the planting of large 'meadows' or wildlife areas in parks or amenity situations, were avoided. In total, twelve brands of seeds were obtained, with two packets of each brand being purchased ( Table 1). The name of the brand, where the seed packets were purchased, and the price of the seed packets were recorded. Seed packets were classified according to key words present in the name of the seed mix that gave some indication of their intended function, such as 'bee', 'butterfly', 'wildlife', and 'wildflowers'.
The seeds were sorted and the number and weight of seeds of each species in each packet were recorded. The weight of any waste material in each packet (e.g., empty seed husks, seed debris, dirt) was also recorded. As many seeds as possible were identified directly to species, which was guided by the species listed on the packaging or provided by online sellers. Seeds that could not be identified directly were germinated (see below) and allowed to produce seedlings, and these seedlings were then identified using the PlantNet app, Collin's Wildflower Guide [41], Webb's an Irish Flora [42], and The Wildflowers of Ireland [43,44]. These seedling identifications were then cross referenced with images of the seeds for verification. Species that could be identified to species level were classed as native or non-native to Ireland by reference to Seawright [45] and Devlin [44]. The proportion of native plant species in each brand of seeds was then calculated using the total number of identifiable species in the two packets.

Seed Germination
Germination trials were carried out to determine the quality of seeds in each brand. Ten seeds of each species from each seed packet were placed in a plastic Petri dish (9 cm diameter) lined with moist tissue paper. If there were fewer than 10 seeds available, then all the seeds were placed in the Petri dish minus one, which was retained for identification should none of the seeds germinate. The Petri dishes were sealed with Parafilm and placed in a dark cupboard for 7 d. On Day 7, the Petri dishes were removed from the cupboard and the number of seeds that had germinated was counted: seeds were classed as germinated if there were visible roots and/or shoots present. Table 1. Summary of twelve wildflower seed mixes purchased in Ireland in 2022. Name of seed brand, where purchased (ol-online purchase), country where produced, mean number of species per packet, mean number of seeds per packet, mean weight of seed per packet (n = 2 packets per brand). The percentage of species native to Ireland was calculated using the total species in both packets, and only considered those seeds that could be identified to species level. When seeds could not be identified directly, germinated seeds were transferred into 7 × 7 cm plastic pots containing a mixture of Westland Garden Soil and vermiculite. These pots were maintained in a glass house at Rosemount Environmental Research Station, University College Dublin, Ireland, and allowed to grow to a sufficient stage that allowed the seedlings to be identified.

Data Analysis
Microsoft Excel was used to collate and organize data, perform simple calculations (e.g., grams of seed per unit cost), and produce all figures, whereas statistical tests were performed using Genstat v21 (VSN International Ltd., Hemel Hempstead, UK). Relationships between price of each packet with total seed count, total seed weight, number of species, and proportion of native plant species were assessed using Spearman's rank correlation coefficient. Similarly, the relationships between species number with total seed count and total seed weight were also examined using Spearman's rank correlation. Differences in seed count, seed weight, species number, and proportion of native species among different countries of origin (Ireland, UK, Netherlands) were assessed using one-way ANOVA, with response variables first being log 10 transformed.
Multivariate analyses, using only the data for seeds that could be identified to species level, were performed using Community Analysis Package v4 software (Pisces Conservation Ltd., Hampshire, UK). Non-metric multi-dimensional scaling (NMDS) analyses were carried out to determine whether seed packets marketed as bee, butterfly, wildlife, or wildflower mixes showed similarities based on plant species composition. Differences between these groupings were also assessed using analysis of similarity (ANOSIM; [46,47]), which produces an indication of statistical significance by comparing the relative withinand between-group similarity with that obtained by 1000 random permutations of the raw data. Similarity percentages analysis (SIMPER) was then performed to see which plant species were primarily responsible for within-group similarity (and between-group dissimilarity). These multivariate analyses were carried out on four versions of the raw data: simple presence or absence of each plant species within each brand, the number of seeds of each species per packet (square-root transformed), the weight of seeds of each species per packet, and the relative germination success, where this was equal to the product of the seed count and germination rate of each species in each brand.

Seed Packet Contents and Value-for-Money
The 24 seed packets contained a total of 69,409 seeds which weighed a total of 304.2 g. The average number of seeds per packet ranged from 230 (BEE3) to 10,724 (BEE1), and the seed weight ranged from 3.9 g (BUT1) to 43.6 g (BEE1; Table 1). Only one brand, WFL3, contained no waste material in the packet, whereas 12.4% of the packet contents of BUT1 (by weight) consisted of empty seed husks, seed debris, and dirt.
The most species-rich brand was WFL5 which contained a mean of 25 species, whereas WFL4 contained an average of only 8 species (Table 1). Apart from WFL4, all the remaining brands contained an average of over 10 species per packet, with six of the twelve brands containing an average of ≥20 species (Table 1). There was no relationship between the number of species a brand contained and the total seed count (r S = 0.17, n = 12, p = 0.588) or the weight of seed (r S = 0.18, n = 12, p = 0.589). The overall germination rate was relatively high, with ten of the twelve brands having a mean germination success higher than 60%, and six of these having a germination success higher than 70% (Table 1). Only WFL2 and WDL1 had an overall germination rate lower than 60%, with 41.0% and 36.6%, respectively ( Table 1).
In terms of cost, there were no significant relationships between price and seed count, seed weight, number of species, and proportion of native plant species (r S < |0.35|, n = 12, p > 0.275 in all cases). At the time of purchase, WDL2 was the cheapest brand at EUR 1.99 per packet, whereas BUT1 was the most expensive at EUR 5.50 per packet (Table 2). BEE1, which contained the most seeds and the greatest seed weight, was also the best value for money, costing EUR 3.50 per packet and working out at 3064 seeds per euro and 12 g of seed per euro. In terms of seed count, BEE3 was the worst value for money at 69.9 seeds per euro, although this was offset somewhat by the high germination rate (Tables 1 and 2). In terms of seed weight, BUT1, which only contained 3.9 g of seed and was the most expensive brand, was, consequently, the worst value for money with only 0.7 g of seed per euro (Tables 1 and 2). On average, WFL3 provided the most (10.3) species per euro, and was followed closely by WDL2 with 10.1 species per euro. The fewest species per euro was 1.7, which occurred in brand WFL4 which also contained the fewest number of species on average (Tables 1 and 2).

Taxonomic Composition of Seed Mixes
In terms of higher taxonomic levels, the seed packets contained 23 plant families, with the commonest five families (Apiaceae, Asteraceae, Fabaceae, Boraginaceae, Brassicaceae) occurring in ten or more of the twelve brands (Table 3). Of these common families, Asteraceae was the most frequently occurring in all twelve seed brands, and was also the family represented by the largest number of species (18) (Tables 3 and 4). The Fabaceae was also well represented in the seed mixes, with 14 species occurring in ten brands (Tables 3 and 4). There were five families that occurred in only one brand each: Amaryllidaceae, Limanthacaea, Solanaceae, Primulaceae and Rosaceae (Table 3). Table 2. Summary of value-for-money measures for 12 mixed wildflower seed mixes purchased in Ireland in 2022. Price is per packet at time and place of purchase (see Table 1).

Code
Price ( Asteraceae  3  5  2  4  7  7  4  5  3  7  5  4  Boraginaceae  2  3  2  2  4  1  1  2  1  3  4  Brassicaceae  2  1  2  1  2  1  3  2  2  3  Caprifoliaceae  1  1  1  Caryophyllaceae  1  1  1  1  1  2  1  1  Fabaceae  3  2  7  1  1  4  5  3  2  1  Lamiaceae  1  3  Limnanthaceae  1  Linaceae  1  1  1  1  Malvaceae  1  1  1  1  1  1  1  1  1  Orobanchaceae  1  1  Papaveraceae  2  1  1  1  2  2  Plantaginaceae  1  1  1  1  1  Polygonaceae  2  1  1  1  1  1  1  Primulaceae  1  Ranunculaceae  2  1  1  1  1  Rosaceae  1  Rubiaceae  1  1  Scrophulariaceae  1  1  1  Solanaceae  1 There was a total of 92 distinct plant species in the 24 wildflower seed packets, although 12 of these species could not be identified from seed or by attempting to germinate and produce seedlings ( Table 4). Three of the ten most commonly occurring plant species belonged to the family Asteraceae (Centaurea cyanus, Cichorium intybus, and Calendula officinalis), two belonged to the family Boraginaceae (Phacelia tanacetifolia and Echium vulgare) and one belonged to each of the families Polygonaceae (Fagopyrum esculentum), Malvaceae (Malva sylvestris), Caprifoliaceae (Agrostemma githago), Brassicaceae (Sinapus alba), and Apiaceae (Anethum graveolens) ( Table 5). Phacelia tanacetifolia, an exotic species in Ireland, was the most common species in terms of the number of packets in which it occurred, the total number of seeds, and the weight of seeds. Malva sylvestris, Centaurea cyanus and Anethum graveolens were also prevalent, occurring in seven or eight of the twelve brands (Table 5). Anethum graveolens, Papaver somniferum, Pulicaria dysenterica, and Trifolium incarnatum were common in terms of seed number with an average of over 2000 seeds of each species across the twelve brands. Trifolium incarnatum, Fagopyrum esculentum, Coriandrum sativum, and Anethum graveolens were also considered common in terms of the weight of the seeds with an average of over 10 g of seed of each species across the twelve brands ( Table 4). The three species that occurred in the top 10 based on all criteria (frequency, seed number, and seed weight) were Phacelia tanacetifolia, Centaurea cyanus, and Anethum graveolens. Table 5. The most common plant species occurring in wildflower seed mixes available in Ireland based on the number of brands in which the species occurred, the total number of seeds per packet, and the total seed weight (g) per packet. Names in bold represent native Irish plant species.

Rank
Brands (

Multivarate Analysis of Seed Species Composition
The suite of NMDS analyses did not indicate there were any particularly distinct clusters of brands based on plant species composition and the proposed function of the seed blend (e.g., to promote bees, butterflies, wildlife, wildflowers; Figure 1). In support of this, the ANOSIM procedures also indicated there were no significant differences in species composition among the four classes of seeds (p > 0.6 for all four analyses).

Seed Origins
Information was available on the packaging for all brands indicating where the seeds  Over all four NMDS analyses, the more generic, 'wildflower' and 'wildlife' seed mixes were well spread in terms of the NMDS Axis and Axis 2 scores (Figure 1). The NMDS plots (Figure 1) suggested there could be some separation of the seed mixes aimed at 'bees' and those aimed at 'butterflies', and pairwise comparisons of these two groups performed as part of the ANOSIM procedures provided moderate evidence that this might be the case (0.3 > p > 0.1). SIMPER analyses suggested that brands aimed for 'bees' were dominated by Phacelia tanacetifolia, Fagopyrum esculentum, Centaurea cyanus, Sinapus alba, and Anethum graveolens. Conversely, brands aimed for butterflies were dominated by Trifolium repens, Centaurea nigra, Centaurea cyanus, Malva sylvestris, and Echium vulgare. Although there was some inconsistency among the SIMPER analyses on the different data sets, the separation of the bee and butterfly groups generally arose because of the prevalence of Phacelia tanacetifolia and Trifolium incarnatum in the bee brands, and prevalence of Trifolium repens and Papaver rhoeas in the butterfly brands.

Seed Origins
Information was available on the packaging for all brands indicating where the seeds were packaged but not the provenance of the actual seeds. More than half (7/12) of the brands purchased in Ireland originated from the UK, two were from the Netherlands, and only three brands were packaged in Ireland (BEE1, BUT1, WFL2). There were no differences among countries of origin in terms of seed weight, species number, and proportion of native species (F 2,9 < 1.7, p > 0.24 in all cases). There was, however, a difference among countries in terms of seed number, with the brands packaged in Ireland containing significantly more seeds than those packaged in the UK (F 2,9 = 10.7, p = 0.004).
Of the 80 plant species that were identified, only 35 (~43%) were judged to be native to Ireland (Table 4). Of the ten most frequent species, only two, Centaurea cyanus and Echium vulgare, are considered native to Ireland by Seawright [45] and Devlin [44] (Table 4). In terms of seed number and seed weight, five native Irish species were ranked in the top ten: Pulicaria dysenterica, Trifolium incarnatum, Trifolium repens, Leucanthemum vulgare, and Centaurea cyanus ( Table 5).
None of the seed brands contained 100% native Irish plant species, with the proportion of species considered native ranging between 10% and 73%. Brand BUT1, which is packaged in Ireland and advertised as a 'native wildflower seed mix', contained the highest number of native plants with 15 species, although this was still only 67% of the 24 total species that were identified in that brand. WLF1 had the largest proportion of native species with 73% of species considered Irish natives ( Table 1). The three brands whose labelling suggested they were aimed at producing flowers for attracting bees contained 25% or less native Irish plant species. WFL4, purchased from a website called Seeds Ireland but where the packaging of the seeds we purchased was carried out in the Netherlands, contained the lowest proportion (12.5%) of native Irish plant species.

Discussion
The results of this study illustrate the large variation in seed number, seed weight, species diversity, and value for money among wildflower seed mixes available to consumers in Ireland. Price ranged from EUR 2.00 to EUR 5.50 per packet, and they contained from 230 to over 10,000 seeds and between 8 and 25 species. We concede that our sample of only twelve brands is only a small proportion of the many brands available, and that we have not included other formats, such as large boxes of seeds used to create mini-meadows, landscape mixes, and 'bee bombs'. Additionally, obtaining data from only two packets of seeds per brand did not allow us to fully examine 'within-brand' variation. Nevertheless, the results add further impetus to growing concerns regarding the provenance and biogeographic status of the species contained in these wildflower mixes, and the ecological 'naturalness' of the plant assemblages they create.
Customers buying these seed mixes can generally obtain information on the weight or number of species included in the packet, and therefore gauge value for money. Information on how many species were contained in the packet, which species were present, and whether these species are native to Ireland was less available. In some cases, it may have been possible to determine which species were present using images on the packaging, although these images were not always available [27]. The germination rate of seeds also varied considerably, highlighting that the success of wildflower mixes is a function of both the species composition and plant establishment. Consequently, even when planting wildflower seed mixes with the same seed composition, the final floral assemblages may differ, and may not contain all the species listed or produce plants in the same proportions as stated in the seed blend [35,39].
Another issue of concern was the difficulty in finding information relating to seed origin, as, although the majority of seed packets stated the country in which the seeds were packaged, none specifically stated the country of seed origin. For Irish consumers attempting to obtain seeds of Irish origin, as recommend by several conservation organizations, this situation could be perplexing. This problem can sometimes be further exacerbated when buying seeds from online suppliers, where website names such as Seeds Ireland and Irish Plants Direct might suggest seeds are of Irish origin when they are actually supplied from other European countries such as the Netherlands and the UK.
The most common plant families included in the seed mixes were the Asteraceae, Apiaceae, and Fabaceae. These families frequently dominate seed mixes aimed at pollinators, with different pollinator groups showing preferences for flowers within these three families [48] For example, bumble bees often show strong preferences for clovers and other legumes in the Fabaceae, whereas hover flies (Syrphidae) have been shown to exhibit preferences for members of the Apiaceae [22,35,36,39]. Phacelia tanacetifolia, a species in the borage family native to Mexico and south-western USA, was the most common plant species across all twelve brands. Phacelia is often referred to as a 'bee plant' [45], and included in seed mixes for agri-environmental schemes as well as community use. Warzecha et al. [49] found that P. tanacetifolia was one of the top four key species in wildflower seed mixes intended for pollinators, and supported 80% of flower visitors in sown wildflower patches. Other key flowering species for pollinators, as determined by Warzecha et al. [49], that were also found in the current study are Achillea millefolium, Fagopyrum esculentum, Malva sylvestris, Daucus carota, Echium vulgare, Calendula officinalis, Linum usitatissimum, and Centaurea cyanus [49]. Nevertheless, many plant species known to be attractive to pollinators in Ireland, such as Taraxacum spp., Lamium spp., and Leontodon spp., were not identified in any of the packets we examined [38].
Given that several previous studies have demonstrated a disparity among floral preferences of different pollinator taxa, any generic pollinator seed mix would benefit from containing a high diversity of plant species, belonging to a range of families [37]. Additionally, it could be hypothesized that, based on the same evidence, pollinator-specific seed mixes would have floral compositions based on the preferences identified for each taxon [23,35,39]. The results of our NMDS and ANOSIM analyses, however, provided no evidence that seed mixes that stated similar conservation aims on the packaging contained seed mixes with similar species compositions. For example, despite there being evidence of negative associations between butterflies and Asteraceae, ten different Asteraceae species were present in the two butterfly-targeted brands [39]. Similarly, although bumblebees often show clear preferences for Fabaceae, the bee-targeted brands in this study contained very few species in this family. It appears, therefore, that the targeted seed mixes we examined offered little indication of being designed based on scientific evidence of pollinator-plant interactions, and this may represent another issue of misrepresentation to consumers who wish to create habitat for a specific pollinator group.
Overall, only 43% of the 80 species we identified in the wildflower seed packets sold in Ireland were actually native to Ireland, and no brand contained 100% native species. In our seed packets, three species, Phacelia tanacetifolia, Anethum graveolents, and Centaurea cyanus, were all considered 'common' based on all three measurements we used (frequency, seed count, seed weight) but of these, only C. cyanus is native to Ireland. In the extreme, one brand of seeds, described as a 'wildflower mix' and obtained from an online seller called Seeds Ireland, contained only one native Irish species. These findings, although disappointing, were not surprising, as several previous commentators on the use of wildflower seed mixes in Ireland have highlighted the prevalence of non-native species (e.g., [21,26,27]. Even when species native to Ireland were present in the packets, as no information on seed origin was provided, we feel it highly unlikely that the seeds were of Irish provenance. In effect, any labelling or packaging describing these seed blends as 'native' or 'wildflowers' could be construed as being highly misleading, and one of the strongest recommendations we would make at this time is that at least the word "wild" is removed from labelling [27]. If the seed blends are described simply as "mixed flowers", there are no longer any suggestions that the species they contain are native wildflowers, or that they will re-create natural plant communities. An alternative approach would be to establish a regional or national Irish certification system, similar to the 'vegetal local' initiative in France, that guarantees the region of origin of seeds and provides a means of traceability to help conserve genetic diversity and provide locally-adapted plants for restoration of local ecosystems (www.vegetal-local.fr; accessed on 19 January 2023).
Whether the plant species are native or not, these seed mixes provide a convenient method of rapidly creating areas of flowers of different taxa, colours, and sizes, that are available over the course of the year and are attractive to a diverse community of pollinator groups [50][51][52]. Thus, the value of the seed mixes to different stakeholders (e.g., farmers, community groups, individuals, landscape gardeners, restoration ecologists) also requires some consideration. In farmland floral strips, the aim is generally not to create or re-establish a native plant community but to provide resources for pollinating insects and promote ecosystem services. Consequently, many of these agricultural seed blends contain non-native species (e.g., Phacelia tanacetifolia, buckwheat, borage) that are known to attract pollinators [18]. Additionally, wildflower mixes based on legumes, even if these are non-native or agricultural cultivars, would still provide agricultural benefits by fixing nitrogen as part of fallow rotations [21,36]. For conservationists, the primary aim might be restoration of a florally-rich semi-natural habitat, and to create a mix of native species in a self-sustaining plant community [53]. The seed blends used to create these seminatural habitats, such as traditional hay meadows, will, however, often contain agricultural cultivars and non-native archeaophytes such as cornflowers and corn marigolds [36].
In many areas of western Europe, many people already live in highly modified landscapes where few natural plant communities still exist, and it must be remembered that the creation of wildflower areas or 'mini meadows' on waste land or road verges is not there only to benefit wildlife but also serves to improve the wellbeing of the people who interact with that space [29,54]. Where conservationists might aim to create a 'wild' habitat that enhances the diversity of native flora and fauna, community groups may be more focused on developing a site that is aesthetically attractive, not over-grown, and represents a space that can be used safely for wider community participation, socializing, recreation, and education [55]. In the latter case, the overall aim of creating such sites might not be to exactly emulate a natural ecological plant assemblage, but provide an acceptable alternative that benefits both wildlife and the local human community. The presence of ancient introductions, such as corncockles, cornflowers, and poppies, or the inclusion of non-native species with long-lasting, high visual impact, should not detract from the benefits to human wellbeing that these florally-rich spaces provide [29,30,33,56]. Given that small urban vegetation patches can make significant contributions to local floral species richness [57][58][59], and that non-native plant species often integrate into local plant-pollinator networks, in many cases a balance could be found that meets these various social and ecological objectives [55,60]. Ultimately, even though a 'wildflower' mix might contain non-native plant species this does not necessarily imply the floral assemblage produced has no value for wildlife conservation, and there is a risk that over-demonizing these products might deter members of the public from creating wildflower spaces at all.