Le terme « agriculture urbaine » renvoie à une diversité de « pratiques agricoles individuelles ou collectives qui se déroulent au sein même de la ville » (Mundler et coll., 2014). Bien que le sujet ait déjà généré beaucoup d’intérêt académique, ce mémoire vise à offrir une perspective novatrice, centrée sur les changements ontologiques qui se produisent chez les jardiniers tels que rapportés dans leurs témoignages alors qu’ils s’engagent pratiquement et affectivement avec les plantes qu’ils et elles cultivent. De cet engagement résulte la participation des jardiniers.ères à un réseau de relations multiespèces impliquant tous les insectes, animaux, champignons et microbes qui interagissent avec leurs plantes. Je décris ainsi comment les jardiniers.ères en viennent à développer des liens avec toutes ces espèces fourmillantes et à s’engager pour qu’elles prospèrent. Finalement, cet engagement les fait réévaluer comment ils désirent prendre une place comme humains dans ce collectif multiespèces et ils et elles en viennent à repenser le social sur le modèle permaculturel inspiré du jardin pour imaginer une « permaculture humaine », une nouvelle manière d’envisager le lien social et l’existence humaine et urbaine. / The term “urban agriculture” refers to a diversity of “individual and collective agricultural practices taking place within a city” (Mundler et al., 2014, free translation). This subject has been the interest of much discourse in the academic as well as the public sphere, as we hear more and more about a “greening of cities” that is coming about with increasing temperatures, drought, fresh food scarcity and loss of biodiversity in and around cities. This dissertation aims to offer a novel perspective on the subject of urban gardening, inspired by literature around ontologies and multispecies sociability. In my interviews of different urban gardeners involved in the production of food in cities around the province of Quebec, I highlight how these gardeners’ perspectives—and, more fundamentally, their world vision—become transformed as they entangle themselves in the network of multispecies living taking place in and around the garden. As their understanding of other species in the garden are transformed, so are their perspective of themselves as humans and their knowing of their place in the garden-and more generally, on our planet. Their practice of care, attention and responsibility for their other-than-human counterparts in the garden also allows them to rethink the politics of their occupation of urban space and food production more generally, as they propose we move towards a “human permaculture”.

<p dir="ltr">Urban agriculture has grown in popularity in recent decades, due to its ability to provide access to healthy fruits and vegetables in urban zones, as well as its importance in fostering knowledge of agriculture within communities. However, urban agriculture may struggle with unique challenges due to its proximity to urban and industrial activities, such as food safety risks due to toxic heavy metals and metalloids which may be present in urban soils in high concentrations. Heavy metals and metalloids (HM) like arsenic, cadmium, and lead are absorbed by plants from the soil, and may accumulate in the plants’ edible tissues, which are consumed by humans. Carrot (<i>Daucus carota</i> L.), in particular, hyperaccumulates these toxic heavy metals in its edible taproots, leading to food safety risks on urban farms.</p><p dir="ltr">One potential way to help address this challenge is to breed carrot varieties with low uptake of HM. In recent years, researchers have identified lines with high and low uptake in greenhouse trials and single location breeding nurseries. However, to be viable, these lines must consistently vary in HM across sites despite differences in environmental and management factors that can also greatly influence HM bioavailability and uptake. Moreover, screening for differences in HM uptake is time-consuming and expensive, and breeders need new tools to select among segregating breeding populations. By using on-farm participatory research as well as advanced phenotyping technologies, we investigate the viability of breeding carrots for HM uptake and the potential of new tools to advance these efforts in order to mitigate the risks on urban farms.</p><p dir="ltr">In the summers of 2021 and 2022, participatory on-farm trials were conducted to determine the HM risks on Indiana urban farms and to investigate the consistency of differences in HM uptake between carrot breeding lines taken from breeding trials (Chapter 2). Results of these trials indicated that while carrot genotype had an effect, there was still significant variability in carrot uptake of arsenic, cadmium and lead between farm sites and years. Results indicated significant differences between site-years, and carrot HM concentrations that correlated strongly with soil concentrations for that particular element. However, there were some site-years with low soil HM content and other soil factors expected to reduce uptake such as pH and phytoavailable zinc concentrations (such as site-year H), that had high carrot HM content. There were significant differences in carrot cadmium (Cd) and arsenic (As) content between carrot breeding lines. For instance, breeding line 3271 had a high As average concentration but low Cd average concentration, while breeding lines 6220 and 2327 had low As and high Cd concentrations. We identify the possibility of other mediating factors, such as uptake of antagonistic micronutrients, or microbe-assisted HM uptake and amelioration that need further attention.</p><p dir="ltr">In the fall of 2022, a study was conducted to investigate the possibility of using phenotyping technologies such as RGB and hyperspectral imaging to detect Cd stress in carrot and attempt to predict uptake (Chapter 3). RGB (red green blue) is a digital color model in which cameras can capture important visual cues compiled from information about each pixel. Hyperspectral imaging uses cameras to capture wavelengths beyond the visible spectrum, which can detect plant stress indicators like increased anthocyanin content for specific environmental stresses. Results of this trial were useful, with some time points and indices noting differences between carrot lines. For instance, RGB factors hue and fluorescence as well as hyperspectral reflectance plots and vegetative indices swirNDVI and ANTH were the most diagnostic. Breeding lines 6636 and 8503 showed the greatest separation between Cd treated and control carrots in imaging indices. However, further studies will be needed to optimize this approach for breeding programs.</p><p dir="ltr">This research demonstrates that growing carrots on most urban farms in Indiana is safe. The studies also provide further evidence that it will be possible to help lower food safety risks by selecting carrot varieties with low HM uptake, and phenotyping can help to advance these efforts. At the same time, new research to understand how soil factors such as microbiomes influence HM bioavailability and uptake on urban farms are also needed to further reduce potential risks. In the meantime, farmers should continue to test their soil for HM and take appropriate actions to reduce risks such as using raised beds and soil amendments that can bind metals like biochar. Consumers should also continue to wash and peel their carrots before consumption, as well as consume a balanced diet with a diverse set of vegetables and other crops.</p>