Adapting the Cool Farm Tool for Achieving Net-Zero Emissions in Agriculture in Atlantic Canada

Abstract

Agriculture is responsible for nearly one-quarter of global greenhouse gas (GHG) emissions, with livestock and poultry systems contributing significantly through methane (CH₄), nitrous oxide (N₂O), and carbon dioxide (CO₂). Achieving net-zero agriculture demands tools that not only quantify emissions but also guide management decisions and foster behavioral change. The Cool Farm Tool (CFT)—a science-based calculator for farm-level carbon footprints, water use, and biodiversity—has been widely adopted across Europe and parts of the United States. Yet, despite its proven potential, no Canadian studies have tested or adapted CFT, leaving a major gap in the country’s progress toward climate-smart farming. This paper addresses that gap by presenting the first surveys of poultry and dairy producers in Atlantic Canada as a foundation for tailoring and localizing CFT. Our mixed-methods surveys examined farm practices, feed, manure, energy use, waste management, sustainability perceptions, and openness to digital tools. Results on 23 responses (20 for poultry, 3 for dairy) revealed limited awareness but moderate interest in emission tracking: dairy farmers, already accustomed to digital systems such as robotic milking and herd software, were receptive and confident about adopting CFT. Poultry farmers, by contrast, voiced greater concerns over cost, complexity, and uncertain benefits, signaling higher adoption barriers in this sector. These findings highlight both the opportunity and the challenge: while dairy farms appear ready for rapid uptake, poultry requires stronger incentives, clearer value demonstration, and sector-specific customization. We conclude that adapting CFT with regionally relevant data, AI-driven decision support, and supportive policy frameworks could make it a cornerstone for achieving net-zero agriculture in Atlantic Canada.

1. Introduction

Sustainable agriculture requires reducing greenhouse gas (GHG) emissions, which in turn demands accurate tools for computing farm-level emissions [1,2]. Farmers and producers require scientifically accurate, robust, universal, and user-friendly tools for assessing their day-to-day farming practices. The Cool Farm Tool (CFT) [3] is a web application that enables farmers and producers to calculate the carbon footprint of their farms and operations. CFT was developed at University of Aberdeen with the help of Unilever during 2008, and since 2012, CFT has been maintained by The Cool Farm Alliance made up of stakeholders joining from various academic and industry domains [4]. The Cool Farm Alliance’s main goal is to create farmer-friendly tools based on Intergovernmental Panel on Climate Change (IPCC) methodologies and recent research. CFT allows farmers to identify high emission sectors and apply modifications to farming practices to decrease negative environmental impacts while allowing them to customize it to their needs [5,6,7]. To date, CFT has supported over 125,000 online assessments and 29,000 users, with more than 160 members worldwide, providing a scientifically robust method for quantifying GHG emissions, water use, and biodiversity effects.

Although CFT is widely popular in Europe, the application of CFT in North America remains noticeably restricted, with only a few farmers utilizing CFT, mostly for crops. While the exact reason was not confirmed, it could be due to unfamiliarity of the tool, limited technological proficiency, reliance on traditional farming methods, the use of very few local tools (i.e., NESTT [8], Holos [9], COMET-FARM [10] etc.), differences in crop types, environment, and farming practices, or CFT itself being more tailored to European contexts and requiring customization. As part of the long-term goal of achieving net-zero agriculture in Atlantic Canada, adopting CFT locally would represent a concrete step toward Canada’s climate targets. The complete process of the adoption in poultry, dairy, and crop farms will be executed in four phases: (i) Phase 1: Data Collection (i.e., data on animal/crop type, animal feed, manure, farm energy usage, etc.); (ii) Phase 2: Customization (i.e., working with farmers, producers, and industry experts to refine CFT for Atlantic Canada); (iii) Phase 3: AI-Integration (i.e., development of AI-based CFT for real-time GHG emission tracking and predictive modeling with optimization); and (iv) Phase 4: Pilot and Scale (i.e., trials on small and medium-sized farms for evaluation leading to broader adoption).

Local adoption of CFT first requires understanding its current functions, scope, and limitations. Our literature review showed that while CFT performs well in carbon emissions, water, and biodiversity, it must be adjusted to local realities, as many internal and external factors are not reflected in its default settings [4,11]. Hence, the requirement arises to specify the gaps due to the local environment and farming practices. The originality and priority of our research is underscored by CFT’s limited use in North America (only in the United States so far). This paper lays the foundation for adapting CFT with AI to provide farmers with a decision-support tool for GHG tracking, hotspot identification, and optimization. It serves as the initial point for the aforementioned “Phase 1: Data Collection”, which includes two surveys to understand local requirements in poultry and dairy farms. Both surveys were based on CFT input requirements and modified with questions addressing local farming practices to capture region-specific insights.

To the best of our knowledge, this is the pioneering approach toward requirement analysis for CFT adoption in Atlantic Canada and development of a complete AI-powered decision support system for low-carbon, high-efficiency farming to optimize feed, manure, and energy use. In this paper, we address “Phase 1: Data Collection” through (i) designing two user-friendly, customized, and anonymous surveys for poultry and dairy farmers to collect data on farm information, practices, sustainability, waste and plastic management, and CFT adoption; (ii) distributing the surveys and interviewing farmers both online and offline, applying data preprocessing, and conducting qualitative and quantitative analyses; and (iii) interpreting survey results to identify regional agricultural requirements (with a focus on feed, manure, and energy use) for AI-driven CFT adoption.

In summary, this study has the following three objectives.

• Establish a baseline for poultry and dairy farming practices in Atlantic Canada;
• Highlight primary challenges for CFT adoption in local agriculture; and
• Outline priority areas for tailoring CFT modules considering both farming practices and external assistance from government and non-government organizations.

The organization for the rest of this paper is as follows: Section 2 summarizes the basic modules of CFT; Section 3 presents a comprehensive literature review on CFT applications and key findings; Section 4 represents the survey design and methodology; Section 5 shows the survey responses analysis; Section 6 discusses the insights gained from the analysis; Section 7 mentions the limitations of the survey and results; Section 8 lists the possible future research directions; and finally, Section 9 concludes the paper.

2. CFT Modules

CFT includes and discusses three major modules for quantifying carbon footprints of a farm: GHSs, biodiversity, and water use. These focus modules were selected with extensive empirical research to create value in the agricultural system as well as to enhance the sustainability practices in agriculture all over the world.

2.1. Greenhouse Gases

The GHG module offers farmers and supply chain partners an accessible, science-based platform to quantify on-farm GHG emissions and soil carbon sequestration. It aligns with the GHG Protocol and Science Based Targets Initiative (SBTi) to ensure consistency with recognized climate action frameworks. The tool promotes sustainable, regenerative farming by enabling users to assess how management practices (e.g., tillage, cover cropping, fertilizer use) affect carbon emissions and soil health. Despite the technical nature of carbon accounting, farmers can complete an assessment in as little as 30 min. This quick process enables users to track emissions regularly and test alternative management scenarios. The tool strikes a balance between scientific rigor and practical usability, making it a helpful resource for day-to-day decision making as well as long-term planning. Crop GHG assessments require inputs such as harvested yield, land area, fertilizer type and amount, crop protection inputs, and energy use. Data on transportation like mode, distance, and load can also be included if available. For livestock, relevant inputs include herd or flock size, feed composition, manure management practices, and associated energy or transport use. With options for both crop and livestock systems, CFT is flexible and applicable across diverse farm types.

Additionally, CFT’s emissions hotspots identification and visualization support more informed, data-driven action plans tailored to individual fields. As the results are context-specific, each farm’s response to changes is modeled individually, accounting for unique conditions. CFT is widely recognized in the food and beverage sector, with many global companies incorporating it into their sustainability strategies. By using the tool, these companies work closely with their agricultural suppliers to better understand on-farm emissions and identify opportunities for improvement, supporting a shared approach in reducing climate impact across the supply chain. The GHG assessments considers several pathways, climate zones, and methods for six categories: general, annual crops, perennials crops, dairy, beef, and other livestock. Three different scopes are defined by the GHG protocol: scope 1 (direct emission at farm from field and cattle), scope 2 (indirect emission from purchased energy or electricity), and scope 3 (other indirect emissions from products or services used in operations). The in-built conversion is applied to the inputs for unit conversions in distance, time, weight, and area. Carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) are reported as kilograms of CO₂, kg(CO₂e), as per the global warming potentials (GWPs). The emission calculations combines IPCC Tier 1 and Tier 2 methodologies, sixth assessment report (AR6) of IPCC, refined emission factors, and empirical models. The computations include many factors, including but not limited to fertilizer production and application, energy use, transport of materials, seed emissions, biomass changes, soil organic carbon (SOC) changes, pesticide production, grazing, bedding, feed production, etc. The CFT live project keeps adding and updating methods to maintain consistency with global GHG protocols and standards.

2.2. Water

Given that agriculture accounts for approximately 70% of global freshwater consumption, managing water resources sustainably has become an increasingly important responsibility. CFT’s Water Module helps farmers and supply chain partners assess water use on farms, compare applied fertilizer amount with the required amount, making it easier to spot problems, reduce waste, and make fine adjustments to increase yields. Additionally, it helps users identify measures that can improve efficiency by allowing them to test different management measures, such as changing irrigation type or increasing soil organic matter to improve water retention, helping farmers make the most of every drop of water. CFT needs inputs like farm area, planting/harvesting timeframes, soil moisture, water usage, and irrigation information to provide practical and optimized suggestions.

2.3. Biodiversity

The biodiversity module converts everyday management decisions into an intuitive score reflecting the potential of the farm to support local ecosystems. This module provides insight into which species groups—pollinators, birds, or other beneficial organisms—are most likely to benefit from their current strategies. The scoring system is designed to award credit for on-farm biodiversity practices such as maintaining flowering strips, leaving hedgerows intact, or selecting pest control options that are less disruptive to beneficial species. The inputs include total farm area (both productive and non-productive), farm management practices, cultivation, and provision of small and large habitats. Currently, the module covers multiple biomes, including temperate forests, Mediterranean, and semi-arid regions, with support for tropical forests making the tool increasingly relevant to a variety of agricultural systems.

Being a user-friendly tool, all these modules require simple inputs from the farmers as answers to some guided questions on environment features, farm information, animal/crop characteristics, and regular farming practices. CFT then computes the corresponding GHG emissions and provides a comprehensive report with scores and visualizations on both general and specific group-wise GHG assessments while identifying GHG hotspots.

3. Literature Review

An extensive literature review was conducted to better understand the effectiveness of CFT and similar tools to provide an overview on CFT usage by farmers and industry stakeholders around the world.

3.1. Article Selection

Publications from 2021 to 2025 were searched and collected from Google Scholar [12], Springer Nature [13], and Science Direct [14]. We also included a few publications from 2011 to 2020 due to their relevance. The search terms focused on “Cool Farm Tool" and its variants—such as “CFT”, “Cool Farm Tool livestock”, and combinations with keywords like “greenhouse gases”, “water footprint”, and “biodiversity scoring”. Most of the publications involved case studies, tool validation, or comparisons with other sustainability frameworks, whereas a few reported user feedback from farmers or supply chain actors. In total, 41 publications were included in the review to explore the current state of CFT and find the scopes—particularly in the context of Atlantic Canadian agriculture.

3.2. Comparing CFT to Similar Tools

In several studies where CFT and similar tools were compared, CFT stood out due to its publicly accessible methodology, visualizations, ease of use, and large number of inputs [15,16,17]. However, it lacked certain features like output for CO₂, CH₄, N₂O, CO₂-eq, mitigation measures, and emissions per source. Other tools assessed included features for environmental impact category indicators (EICIs), NH₃, scoring, improvement suggestions, and feed consultations [17]. Furthermore, CFT did not include social or economic dimensions, which were included in tools like SAFA, FRST, and Dairy Farms Plus [18,19,20,21].

Table 1. GHG calculation tools from various countries and farm types. The references in ‘Name’ and ‘Literature Mentions’ are for the tools and publications that mentioned the tool, respectively.

Name	Country Mentioned	Farm Type	Features	Literature Mentions
Farming Enterprise GHG Emissions Calculator [22]	-	Crop, Dairy, Beef, Sheep	Takes simple numerical inputs on fuel, soils, and animals. Shows annual emission scores and graph for fuel, soil, fertilizer, animal, and others.	[15]
GLEAMi [23]	-	Livestock	Takes herd, feed, manure inputs for livestock. Lists default emissions for different types of farms. Provides scores and graphs on total and source-wise emissions.	[16,17,24]
AgRECalc [25]	UK	Crop, Livestock	Takes animal, crop, soil, fuel, geospatial, and management practices as inputs. Provides total emission with sources breakdown, emission hotspots, carbon sequestration, and comparisons to peer farms and historic data.	[16,26,27]
Farm Carbon Calculator [28]	UK	-	Takes livestock, crops, soil, fuel, structures as direct or spreadsheet data. Provides report customization options on farm boundary, partial or whole farm, category, location to show source-wise live emissions, offsets, and carbon balance scores.	[15,16,26,27]
Greenhouse Accounting Frameworks [29]	Australia	-	Creates carbon footprints for whole enterprises for Australian agriculture, fisheries, and forestry. Shows direct and indirect annual emissions. Provides interactive map for inputs and team sharing access.	[16]
FarmGAS [30]	Australia	Beef, Sheep, Pig	Uses Australia’s National Greenhouse Gas Inventory (NGGI) for base value references. Provides multiple management scenarios for cost and benefit predictions.	[16]
COMET-FARM [10]	USA	-	Provides carbon sequestration and emissions for whole farms and ranches using farm activities and animal data. Provides annual emissions with interactive visualizations.	[16,20,27]
Farm Carbon Footprint Calculator [31]	New Zealand	-	Takes stock, production, fuel, feed, contractors data as inputs. Shows carbon sequestration, direct and indirect emissions. Generates comments on emission equivalence, comparing the impact to environmental effects.	[16]
Holos [9]	Canada	Crop, Livestock	Generates soil features from map based on farm location. Provides user friendly interface with additional options for crop rotation, custom diet creation, background operations, etc.	[16,17,27]
DairyGEM [32]	USA	Dairy	Takes farm and weather parameters as inputs in an offline desktop application. Provides four emission output files: summary output, full report, optional output, and parameter tables.	[16]
Poultry Carbon Footprint Calculation Tool (PCFCT) [33]	USA	Poultry	Provides simple and interactive excel file for current and projected emissions calculations and comparisons. Shows on-farm and off-farm sources-based computations.	[16]
FarmAC [34]	-	-	Shows carbon and nitrogen flows, farm commodity balance using direct or uploaded inputs. Provides GHG warming potentials and indicators for farming practices.	[17]
Overseer [35]	-	-	Provides enterprise and on-farm solutions by leveraging data from over 15000 farms. Monitors emissions, compares with peers for efficient farming practices.	[17]
Carbon Navigator [36]	-	Dairy, Beef	Provides a simple graphical interface for farm, environment and economic data. Monitors emissions, compares with peers, analyzes scenarios, and transfers knowledge.	[17]
RISE [37]	-	-	Checks farm performance based on ecological, economic, and social sustainability practices. Provides details on scopes of improvements for sustainable farming.	[17,20]
EX-ACT [38]	-	-	Provides 17 modules including inputs, outputs, farming, economy, environment, etc. Generates both numerical and graphical outputs of value chain assessment from project implementation.	[17,24,26]
SAFA [18]	-	-	Checks environmental, social, and economic inputs for sustainability assessment. Provides ratings from best to worst on products, water, biodiversity, accountability etc.	[20]
FSA [19]	-	-	Inputs farm management, environmental and economic factors, labor, goals, and risks. Provides performance level, identifies weaknesses, generates comparisons and improvement plans.	[20]
SENSE Tool [39]	-	-	Provides lifecycle impact assessments with climate change, human toxicity, eutrophication, acidification, land use, ecotoxicity, abiotic resource depletion, and water depletion. Generates category-based scores for climate change potentials due to farming practices.	[20]
OFoot [40]	-	-	Uses farm equipment, infrastructure, consumables, land data, and farm operations for assessments. Provides options of carbon footprint computations for multiple case study scenario.	[20]
Farm-to-Market [41]	USA	-	Works with biodiversity, energy use, GHG emissions, irrigated water use, land use, soil carbon, soil conservation, and water quality. Provides sustainability performance and operational efficiency scores while showing national and state average scores for comparisons.	[27]

provides an overview of 21 such tools for GHG emission calculation and visualizations.

3.3. CFT Case Studies

Few studies in this literature review are particularly relevant to Atlantic Canada, as they focus on crops and livestock that are relevant to local agriculture or geographically nearby, sharing similar environment and farming practices. One such study highlighted corn soybean cover crop cultivation in Illinois [42]. Researchers collected agricultural practices and yields data from a local farmer service program known as PCM, used by over 400 farmers in the state to generate inputs for CFT. Their in-depth analysis showed that despite lower GHG emissions, cover crops had lower returns. They highlighted the importance of understanding the potential trade-off between environmental sustainability and financial burden and concluded that the use of programs providing economic assistance to farmers who use cover crops is necessary to alleviate the financial burden.

Another American study focused on dairy farming in New York using CFT-based emission estimations from six organic dairy farms and compared CFT performance to COMET-FARM [6]. Although CFT is closer to a holistic “cradle-to-farm-gate" assessment, it lacked flexibility in customizing feed inputs and local energy data in their evaluations. In a similar study based in Ecuador, CFT was applied in 85 rural dairy farms in the Northern Ecuadorian Andes [43]. Researchers collected data by conducting 55 min interviews with the farmers on pasture and livestock management. Additional information was collected from a workshop held by the National Autonomous Institute for Agricultural Research (INIAP), and soil and climate characteristics were determined using secondary sources. They used CFT on the complete dataset and determined that enteric digestion was the primary contributor (with 80.96% emissions) to GHG emissions. Vetter et al. [5], on the other hand, studied CFT for more than 3 years in 10 large egg producers across USA for GHG emissions calculation. CFT reported the highest emission in feed management, with transport and manure management being the second and third highest emission areas. Their long-term applications and adaptation of CFT showed about a 25% decrease in emissions over 3 years, proving the benefits of CFT adaptation in real-time.

In Southwestern Bangladesh, researchers used CFT to compare conventional cultivation (CT) with zero tillage with rice straw mulching (ZT) for potato cultivation [44]. Their experiments were conducted during two growing seasons on farmer’s fields, where potatoes grown using CT and ZT were harvested at the same time. The results showed that ZT produced lower emissions than CT, partially due to increased productivity when using ZT, and both methods created the most emissions from soil and fertilizer. In Northwestern India, 118 farmers from small, medium, and large potato farms were surveyed to collect quantitative data about seeds, diesel fuel, fertilizers, biocides, irrigation, and tuber yield [45]. The researchers commented on the necessity of further research to determine more user-friendly options for adopting local sustainability recommendations, as the farmers were reluctant to adopt CFT. Maize-wheat cropping systems under different nitrogen fertilizer regimes in Eastern India used CFT for emission calculation in [46]. The results with field-level management data from both direct sources (i.e., fertilizer-induced N₂O) and indirect ones (i.e., fertilizer production) suggested that emissions tend to rise with the increment of nitrogen inputs due to the carbon cost of fertilizer production and the nitrogen behavior in the soil, spotting inefficiencies in resource usage.

CFT was used to analyze the impact of COVID-19 on farming practices for French vines in three different countries in Europe in [47] by surveying actors from the value chains. All additional information was collected through press releases and government releases. While lower exports during the pandemic resulted in lower CO₂ emissions, no drastic changes were observed in the current systems, leading to the decision that the pandemic had no noticeably lasting effects on the long-term agricultural practices. Another survey-based CFT research was discussed in [48], using a questionnaire with queries on cultivation, soil characteristics, fertilization, energy use, and water use and collecting data from 40 tomato producers in Greece. The highest source of CO₂-eq was machinery and fertilization, guiding them to suggest adopting integrated management strategies while limiting the use of machinery and synthetic mineral fertilizer. In another Greek study by Kakabouki et al. [49], researchers worked on quinoa seed production with 40 Greek growers. They reported the main sources of GHG emissions by source: fertilizer production (40.53%), fertilizer application (30.18%), fuel/energy use (25.15%), and residue management (4.14%). This method of breakdown allows researchers to quantitatively assess the emissions and make targeted recommendations for emission reduction.

Other recent papers focused on crops such as cotton, coffee, sorghum, peanuts, sunflowers, sugarcane, rice, maize, and quinoa in Kenya, Republic of Cote Id’Ivoire, South Africa, India, Greece, South Africa, Pakistan, and the United States [11,50]. These studies used CFT to recommend more sustainable farming practices like improved wastewater systems, limited use of synthetic fertilizers, and improved resource efficiency. Additionally, some studies highlighted the importance of data integrity and understanding the social factors that may affect uptake of tools like CFT [51]. Some other recent studies from Ghana, India, South Africa, and Chile presented CFT applications on a variety of different crops and livestock [52,53,54,55,56]. Similar studies compared different farming practices but were not based in any specific country [4,57]. In one study, CFT was used to determine the effects of COVID-19 on farming in Burkina Faso, France, and Columbia [47], while other studies used CFT to examine the environmental impact of the agri-food industry [58,59]. These studies highlight the potential for using CFT carbon footprint estimates to compare the efficiency and sustainability of farming practices.

While the majority of studies focused solely on carbon footprint estimation, few studies used the other modules of CFT. Kayatz et al. [60] examined the efficiency of CFT in water usage while proving its insufficiency in absolute value guidance of field irrigation. On the other hand, CFT was used to evaluate biodiversity in [61] based on science-backed management practices. Researchers in New Zealand used CFT’s biodiversity metric as a conceptual starting point to develop a more locally grounded framework aimed at supporting biodiversity assessments tailored to the conditions of New Zealand farms [62]. Additionally, researchers in Cyprus adapted CFT’s carbon footprint tool to create a vineyard-specific carbon footprint calculator known as the Vineyard Carbon tool (VCT) [63].

3.4. Key Takeaways

3.4.1. Farm Type and Location

The vast majority of farms using CFT in the recent literature have involved different types of crops, dairy, poultry, as shown in

Table 2. CFT in crops emission assessments by country and continent.

Continent	Country	Main Crop(s)
Africa	Burkina Faso	Cotton
	Kenya	Sorghum
	Republic of Côte d’Ivoire	Cacao
	South Africa	Peanut, Soybean, Sunflower
Asia	Bangladesh	Potato
	Cyprus	Vineyards
	India	Sugarcane, Maize, Wheat, Rice. Potato
Europe	France	Vines
	Greece	Tomato, Quinoa
North America	United States	Soybean
South America	Columbia	Coffee

and

Table 3. CFT in livestock type emission assessments by continent.

Continent	Country	Animal(s)
North America	United States	Dairy cattle, Poultry
South America	Ecuador	Dairy cattle

.

Further quantitative analysis on the CFT case studies are represented in Figure 1, Figure 2, Figure 3 and Figure 4. As shown in Figure 1, the majority of the studies were based in Africa or Asia, while Figure 2 shows the number of case studies per country, with India containing most of them. The split between case studies with the primary focus being crop or livestock is shown in Figure 3, and studies that focused on both crops and livestock were not included in the graph. Finally, Figure 4 shows the frequencies of the distributions of CFT features in recent case studies, showing that the GHG estimation tool was the most popular by a wide margin compared to the others.

3.4.2. Data Collection Methods

Although most studies collected data through interviews [43,53,64] and surveys [45,48,54,56,63], data was also partially or entirely sourced from secondary sources [65] like press releases, peer reviewed articles [47], and pre-existing datasets [66] in some of them. Some researchers opted for digital tools such as GeoFarmer APP and DSSAT simulations [49,53] for data collection. Researchers from different parts of the world proved the need for local data collection for applying a new tool to the region. They also showed that the barriers of such adaptation included both tool limitations (technical or feature limitations) and farmer reluctance.

3.4.3. Effectiveness of CFT

CFT is often compared to other tools, with varying strengths and weaknesses, and it is identified as a comparable and trustworthy tool for stakeholders from various domains [15]. Although the literature review focused mostly on environment, crop type, livestock, and region similar to Atlantic Canada, the variation of crops and the applications of CFT present in those works showed the diversity of the tool as well as its adaptability. Additionally, CFT is shown to be user-friendly, with a simple interface with intuitive visualizations [16]. Key features of CFT identified by researchers and other users include detailed input structure [57], low-cost carbon estimates in the absence of advanced measuring equipment [67], and emission identification on a per hectare or per unit of yield basis [45]. Moreover, CFT has been identified as a suitable tool for livestock emissions calculations [26]. These studies on CFT helped to demonstrate the potential in Atlantic Canadian agriculture, in comparison to similar tools, highlighting the opportunities for adoption in local farms. Additionally, the existing research works from other regions provided the necessary background and ideas for possible adaptation approaches to explore.

3.4.4. Barriers Affecting Uptake

When assessing digital tools such as CFT, it is important to consider barriers that could prevent farmers from adapting such tools to their daily farming operations. These factors include inadequate resources, time, money [52], and social factors [51]. Additionally, as seen from the comparison between CFT and other similar tools, CFT lacks the same level of scenario analysis, static and data-driven recommendations, and cost–benefit analysis. There are also several constraints in CFT-based emission calculations, such as not incorporating CH₄ mitigation strategies like cover and flare systems [68]. Furthermore, the limitations include lack of consideration of livestock used in cropping systems [53], social and economic factors [20], and differences in organic production [4]. Notably, the lack of CFT applications in Canadian contexts requires investigations to reveal adaptation challenges.

3.5. Research Problem and Questions

Building on the insights from the literature review, the research problem can be defined as the absence of empirical applications of CFT in Canadian agriculture and the limited attention to livestock and poultry systems in existing studies. While prior research has demonstrated CFT’s potential across diverse global contexts, the evidence base remains heavily crop-focused, geographically concentrated in Europe, Asia, and parts of South America, and largely absent from North America. This gap poses a significant barrier for Canadian agriculture, which is under increasing pressure to adopt climate-smart tools and meet net-zero targets but lacks regionally validated digital decision-support systems. To address this gap, the present study is guided by the following research questions:

• What are the current farm practices, sustainability perceptions, and barriers to digital tool adoption among poultry and dairy farmers in Atlantic Canada?
• To what extent are existing CFT modules (GHG, water, biodiversity) aligned with the realities of Atlantic Canadian farms?
• What region-specific requirements and adaptations are necessary to improve the usability and adoption of CFT for Atlantic Canadian poultry and dairy producers?

Given the exploratory nature of this study and the limited sample size, we do not propose formal hypotheses. Instead, the research questions outlined above serve as the guiding framework for survey design, data collection, and analysis. Future research with larger and more representative datasets may build on this foundation to develop and statistically test hypotheses regarding adoption patterns and sectoral differences.

4. Survey Design and Data Collection

This section outlines the rationale behind the data collection, the structure and scope of the poultry and dairy farm surveys, and the procedures followed to reach participants across the region.

4.1. Study Context and Objectives

Building on the literature review, we translated themes of digital tool adoption and GHG assessment into a practical field data collection framework. Our objective was to gather detailed operational, environmental, and management data from local poultry and dairy farms to inform CFT customization. By documenting current practices and farmer attitudes toward emission tracking, we aimed to identify adaptations needed to improve CFT’s usefulness and uptake in the region. As highlighted in the literature review, region-specific practices (feed types, manure management, energy use) may not be fully addressed by CFT. Therefore, our surveys focused on establishing baseline farm data and gauging openness to digital emission tracking. Understanding farmer needs and constraints allows us to pinpoint what CFT modifications or support systems (e.g., decision support, predictive modeling) are required for successful local adoption. We consulted prior studies on farmer perceptions of new technologies to align our approach with established methodologies, as shown in

Table 4. Summary of existing surveys used globally to understand farmers’ perceptions of technology.

Technology Type	Region	Survey Size	Responder Type	Data Collection Methods	Data Analysis Methods
Precision livestock adoption for dairy [69]	Italy	52	Dairy farmers	Online distribution	Descriptive statistics, Pearson’s correlation
Precision livestock adoption for sheep [70]	England, Wales	439	Sheep farmers	Postal distribution	Descriptive analysis, Exploratory factor analysis, Multivariable logistic regression
Consumer perceptions of precision livestock [71]	Europe	56	Consumers	Group discussions in Finland, Netherlands, Spain	Descriptive analysis
GHG emissions of smallholder dairy farms [72]	Kenya	384	Dairy farmers	Interviews	Fractional response model
Cattle feedlots for climate-smart farming [73]	South Africa	161	Cattle farmers	Print-out distribution	Chi-squared test, Cronbach’s Alpha
GHG emissions from smallholder pig farms [74]	China	272	Pig farmers	Questionnaires in 3 towns	Emission calculation formulae, ANCOVA
Climate change mitigation [75]	Ethiopia	25	Farmers	On-site interview	Mitigation Options Tool (MOT)
Technology on poultry farms [76]	Germany	53	Poultry farmers	3rd-party distributions	Kano Model
Technology on livestock farms [77]	Germany	98	Livestock farmers	3rd-party distributions	Kano Model
Precision livestock adoption for dairy [78]	Ireland	311	Dairy farmers	2018 National Farm Survey	Multinomial logistic regression, Binomial logistic regression

. Notably, ease-of-use and clarity were prioritized in our questionnaires, since CFT’s eventual adoption may hinge on farmer-friendly interfaces. Surveys were anonymous, with no personally identifiable information collected unless voluntarily provided in the “Future Contact” section. Ethical considerations were addressed by ensuring voluntary participation, informed consent, and secure storage of all data. The dedicated “Future Contact” section invited respondents to provide contact information for follow-up research or pilot trials, enabling us to identify participants willing to engage in future phases while preserving the anonymity of all other responses. No pilot testing was conducted until now. The surveys were reviewed by the research team and subject-matter experts for content relevance, but they were not trialed with farmers before the full rollout. Consequently, we could not refine question wording, adjust response categories, or assess completion time using farmer feedback.

As illustrated in Table 4, data analysis methods in comparable studies range from descriptive summaries to advanced regression models and Kano analysis. These decisions typically reflect both study objectives and the scale of available data. Given the exploratory nature of our research and the small sample size, we adopted descriptive statistics, summarizing results with counts, proportions, means, and standard deviations. This method was chosen to provide a clear overview of farmer practices and attitudes while avoiding over-interpretation of patterns that cannot be generalized from such a limited dataset. The emphasis was on generating preliminary insights and identifying areas for further study rather than testing predictive models.

4.2. Survey Methodology Framework

To enhance clarity and transparency, a structured survey methodology framework was developed to guide the design, implementation, and analysis of this study (Figure 5). The first stage focused on the design of two tailored questionnaires, one for poultry producers and one for dairy producers, which were structured into six thematic sections: general farm information, farm practices and operations, sustainability and greenhouse gas management, waste and plastics management, willingness to adopt the Cool Farm Tool (CFT), and future communication needs. Both multiple-choice and Likert-type scales were combined with open-ended questions to capture both quantitative and qualitative insights. Following this, the surveys were distributed through two primary channels: in-person at the Atlantic Poultry Conference in Wolfville, Nova Scotia, and online via provincial commodity boards, industry networks, and direct outreach. The next stage involved response collection, yielding 23 completed surveys followed by data handling and preprocessing. Analysis on the preprocessed data then provided the final results and were translated to interpretations.

4.3. Data Collection Strategy

Data collection was conducted in early 2025 using a two-pronged strategy. First, we conducted in-person surveys at a regional industry event to gather initial responses. In February 2025, at the Atlantic Poultry Conference (Wolfville, NS), poultry farmers were interviewed, producing responses from 13 poultry farms across the Atlantic provinces. Where possible, interviews were audio-recorded (with consent) and later transcribed for accuracy. Following the conference, we expanded our reach with an online survey (Microsoft Forms) distributed to poultry and dairy producers in Atlantic Canada. To maximize participation, 21 poultry, 6 dairy, and 2 mixed (dairy and poultry) producers were initially contacted. Survey links were shared by email, with two follow-up reminders issued at two and four weeks after initial contact. By 28 August 2025, we received 20 poultry and 3 dairy survey submissions, totaling 23 participating farms. The poultry responses predominantly came from the in-person effort (conference) supplemented by 7 online submissions, whereas all dairy responses were obtained via the online survey. Response rates were approximately 87% for poultry and 37.5% for dairy. No incentives were offered for participation, and missing entries in questionnaires were coded as blank with no imputation.

4.4. Participant Demographics

Among the 20 poultry and 3 dairy farms in this study, poultry respondents represented all 4 Atlantic provinces: 12 from Nova Scotia, 3 from New Brunswick, 2 from Newfoundland and Labrador, and 1 from Prince Edward Island. One poultry farm reported operations in both Atlantic Canada and Ontario/Quebec, while one respondent did not specify a province. This geographic distribution indicates coverage of diverse regional contexts, though all dairy respondents were located in New Brunswick, creating a provincial skew. Figure 6 illustrates the distribution of poultry farm types among our respondents.

The cohort included broiler (meat chicken) farms (8), layer (egg) farms (7), and mixed-production farms (e.g., broiler–breeder, broiler–turkey, or layer–pullet). In terms of scale, poultry farms ranged from small to large: using categories based on flock size, we had 2 small farms (less than 10,000 birds), 10 medium farms (10,000–50,000 birds), and 8 large farms (greater than 50,000 birds). Most poultry respondents thus operated medium or larger commercial operations, with only a couple of small-scale producers in the sample. The dairy farms were generally moderate-to-large family operations, averaging 433 acres in farm area and managing sizeable herds. On average, the dairy farms reported about 62 milking cows (with additional dry cows, heifers, and calves bringing the total cattle inventory higher). All three dairy farms used robotic milking systems, milking more than twice daily, and reported daily milk production averaging about 39 L per cow. Notably, all dairy farmers indicated they use farm management software tools for record-keeping or herd management (i.e., DeLaval herd management systems, QuickBooks for farm finances). In contrast, digital record-keeping among poultry farmers was less uniform: only a subset of poultry respondents (mainly those from the online survey) reported using software tools such as Microsoft Excel, SmartBarn, or Google Sheets to manage farm data.

4.5. Data Handling and Preprocessing

All survey responses, whether collected on paper, via interview, or online, were digitized for analysis. We pooled the poultry data from the conference and online surveys, since the two instruments covered largely the same questions. Except for questions unique to one version, results from both poultry surveys were combined to increase sample size and statistical power. Prior to analysis, we performed basic data cleaning such as standardizing units and responses for certain quantitative questions. For example, farmers initially had the option to report electricity usage either in cost (CAD) or in energy (kWh), which complicated comparisons among farms. In the online survey, we refined this by requesting electricity use specifically in kWh. In-person data contained mixed formats, which we addressed by separating those entries from standardized responses. Unanswered or non-standard responses were treated as missing data. Given the descriptive nature of our study, no transformations beyond unit conversion were applied. We computed summary statistics (count, mean, and standard deviation) for the numeric responses in each survey section to characterize central tendencies and variability. The results are presented for key variables and in aggregate form in

Table 5. Number of responses, mean, standard deviation, and statistical significance for variables in the “Current Practices” section of poultry surveys. Mean values represent the average levels reported for each farm practice, while standard deviations indicate the variability among surveyed poultry farms. Statistical tests (Kruskal–Wallis or Fisher’s Exact) evaluate differences across farm size categories—small (< 10,000 birds), medium (10,000–50,000 birds), and large (>50,000 birds).

Category	Responses	Mean	Standard Deviation	Statistical Test	p-Value	Significance
Number of birds	19	90,018.94	178,127.15	Kruskal- Wallis	0.032	*
Feed per bird (grams)	5	832.9	1296.74	Fisher’s Exact	0.721	ns
Monthly fuel (litres)	11	1400.41	1406.74	Kruskal- Wallis	0.156	ns
Monthly electricity (Canadian dollars)	7	2400	1648.38	Fisher’s Exact	0.089	ns
Monthly electricity (kWh)	4	9715.50	8133.47	Fisher’s Exact	0.443	ns
Monthly water usage (liters)	11	178,610.18	241,021.77	Kruskal– Wallis	0.234	ns

Notes: Kruskal–Wallis test [79] for continuous variables; Fisher’s Exact test [80] for categorical variables. Significance codes: *** p < 0.001, ** p < 0.01, * p < 0.05, ns = not significant.

. The mean and standard deviations were computed from the response values, and two statistical tests (i.e., Kruskal–Wallis test and Fisher’s Exact test) were applied for calculating the p-values determining statistical significance.

Given the relatively small sample size, we did not employ dedicated qualitative coding software. Instead, two researchers manually reviewed open-ended responses, independently highlighting themes, categorizing practices, and comparing coding decisions for consistency. Discrepancies were resolved through discussion, resulting in a set of broad descriptive categories (e.g., “energy costs,” “data management challenges,” “interest in renewables”). This qualitative coding process was intended to capture key attitudes and practices rather than to build a formal codebook.

Alongside descriptive statistics, we conducted limited exploratory tests to assess whether interest in adopting CFT was associated with specific farm characteristics. For this, adoption responses were grouped into three categories: Interested, Unsure, and Not Interested. These were cross-tabulated with explanatory variables common across survey modes, including (i) farm size, (ii) farm type (broiler, layer, mixed), (iii) province of operation, and (iv) reported participation in an energy audit or sustainability program. Because of the small sample and sparse cell counts, we applied Fisher’s Exact Test [80], which is robust for contingency tables with low expected frequencies. Analyses were conducted in R (v4.3.3) [81] using standard statistical packages. These tests were strictly exploratory: they did not yield statistically significant associations, but they demonstrate the type of analysis that could be expanded with a larger dataset.

5. Survey Responses and Analysis

The key findings from the poultry and dairy farmer surveys conducted to inform the adaptation of the CFT for Atlantic Canada are presented and analyzed in this section. The results are organized to reflect both the general characteristics of respondents and their reported farm management practices, sustainability perceptions, and openness to adopting CFT. Data are drawn from both in-person and online responses, with quantitative measures summarized using counts, means, and standard deviations, and qualitative answers grouped by recurring themes.

5.1. Response Overview

In this paper, we analyzed survey data from 23 farms in Atlantic Canada: 20 poultry and 3 dairy. Among the poultry respondents, ~40% were broiler, ~35% layer, and the remainder mixed or dual-purpose flocks (Figure 6). The dairy farms were year-round indoor operations with herds of about 100, including several dozen lactating cows, and all used robotic milking systems, indicating a relatively high level of technology adoption. Poultry results combine on-site and online surveys (n = 20), with online-only items noted separately. Dairy responses (n = 3) are included for comparison but should be interpreted cautiously given the very small sample.

5.2. Farm Management Practices

5.2.1. Feed and Water Use

Feeding practices varied considerably, particularly among poultry farms. Respondents reported diverse feed types: mash (8 farms), pellets (2), crumble (2), and specialty formulations such as “lay mash.” A minority also sourced feed on-farm (grain or organic rations). This diversity, with no single dominant feed, suggests that diets are tailored to farm preferences or supplier availability. Dairy farms were more uniform: two used partial mixed rations and one a total mixed ration, typically of corn silage, dry hay, and concentrates. Reported intake averaged 50.7 kg per cow per day (fresh weight), though values varied by farm. None of the dairy farms practiced grazing or pasture feeding.

Groundwater wells were the primary water source in both sectors. Eighteen of twenty poultry farms relied on wells, one used lake water, and one pond water; none used municipal supply (Figure 7). All dairy farms also depended on on-site wells. Most respondents did not monitor water consumption: poultry farmers generally lacked usage records, and dairy farmers were uncertain about monthly totals. Wastewater management was limited: only one poultry farm stored and field-applied treated wastewater, while all others discharged or spread it without prior treatment.

5.2.2. Animal Housing, Cleaning, and Manure Management

Hygiene routines and manure handling differed markedly between sectors. Poultry farms reported variable cleaning schedules: every 6 weeks (2 farms), every 3 months (1), annually (4), or 75–80 weeks (1). Others gave conditional responses such as “as needed” or annual disinfection with minor interim clean-ups. In contrast, all three dairy farms cleaned and disinfected barns daily, facilitated by alley scrapers or automated removal—practices essential for cow health and characteristic of modern dairy housing.

Manure management in farms also diverged. Among poultry farms, the most common strategies were direct land application (8 farms) and composting (6). Other practices included drying and storing manure on belts (1), using an unspecified technology-based system (1), or selling/giving manure to neighboring farms (3). Only one poultry farm reported lagoon storage, consistent with the solid or semi-solid litter systems typical of poultry production (Figure 8). In contrast, all dairy farms managed manure as liquid or slurry, using lagoon or pit storage; one also handled solids from bedded pack areas. Because none of the dairy herds grazed, all manure was collected indoors. Bedding also differed: dairy farms mainly used sawdust, with two supplementing straw and one adding sand. Poultry bedding was not surveyed but is typically wood shavings or straw.

5.2.3. Farm Energy Use and Equipment

Poultry and dairy farms used diverse energy systems, with limited renewable adoption. Among poultry farms, only two had installed solar panels, though six expressed interest in renewables and others reported no interest. None of the dairy farms had renewable systems, but all three indicated future interest (primarily solar).

Conventional energy use in poultry barns was heterogeneous. Gas heating (propane or natural gas) was the most common, used by at least five farms, with additional mentions of propane “box” heaters (2), L.B. White radiant heaters (1), and general propane use (3). Alternatives included a biomass pellet heater (1), electric heaters (2), a heat exchanger (1), and an oil furnace (1). Cooling relied mainly on ventilation, with isolated reports of evaporative pads, large fans, or unspecified air-cooling systems. In dairy barns, all three relied on electric heating; one also used summer cooling pads and a diesel heater for winter extremes.

Dairy fuel and electricity use varied widely: two farms reported 100 L and 3000 L of diesel per month and electricity usage of 1650 kWh and 6300 kWh, reflecting differences in herd size and mechanization. Poultry energy data were harder to standardize, as many reported costs instead of kWh, and some did not track fuel use, reflecting a broader lack of monitoring.

Open-ended responses on equipment use showed that the poultry farmers employed a variety of mechanization tools, including automated augers (2), skid steer loaders (1), dump trucks (1), specialized nest systems (1), automated pan feeders (1), egg packing machines (1), manure conveyor chains (1), and tractors (1). This indicates targeted investment in labor-saving technologies despite reliance on conventional energy sources. Both sectors rely on traditional energy sources but show growing interest in renewables. Both also face challenges in sustainably managing waste streams (manure, wastewater, plastics), discussed further below.

5.3. Sustainability Perception and Practices

Few poultry farmers actively track GHG emissions: 3 of 20 (~15%), including 1 using the NESTT tool [8]. Most did not monitor emissions—six responded “not applicable,” five expressed future interest, one was uninterested, and others said “no” (Figure 9). None of the dairy farms tracked GHGs, though all three expressed interest, consistent with their enthusiasm for adopting emission-reducing technologies.

Mitigation practices were more common. Nearly all poultry farms had implemented at least one measure: energy-efficient lighting/ventilation (6), precision feeding (5), efficient heating (1), LED lights (2), high-efficiency fans (1), cooling pads (1), enriched feed carts (1), composting mortalities (1), and improved manure storage/treatment (2). Only one poultry farm reported no mitigation strategies. Two dairy farms had undertaken similar actions, including efficiency upgrades (2) and improved manure storage (1). None used feed additives for methane reduction, reflecting a knowledge gap.

Participation in formal programs was limited (Figure 10). In poultry, only 2 of 20 had completed an energy audit, 6 were unsure, and 12 said no. Interest in carbon credit or sustainability programs was moderate (6 interested, 2 participating, 4 unsure, others not interested). Dairy farmers had not yet engaged in audits or programs, though their expressed interest in tracking suggests potential uptake.

General attitudes toward sustainability were moderate. Poultry and dairy farmers rated sustainability at an average of 3.67/5 (Figure 11), suggesting it is valued but not a top priority. Poultry farmers cited “industry longevity” (2), lower emissions (1), and lower costs (1) as benefits, while dairy farmers emphasized efficiency (2) and cost savings (3). The latter aligns with their recognition of financial co-benefits from energy efficiency.

Interest in new technologies was also evident. Poultry farmers mentioned solar panels (2) and heat exchangers (1), while dairy farmers highlighted solar and the robotic systems already in use. Overall, most respondents are engaged in sustainability through low-cost measures and are curious about advanced options, but participation in structured programs remains low. Sustainability is viewed as pragmatically important, but it is often framed through cost savings and efficiency, suggesting that clear incentives and tools like CFT could help to convert interest into broader adoption.

5.4. CFT Adoption Willingness

Survey responses revealed clear differences between poultry and dairy farmers in their openness to adopting the CFT. For poultry farmers, the most common barrier was cost (6 of 20), including potential subscription fees, equipment, and time as economic burdens. Other barriers included uncertainty of benefits (4), complexity or data demands (3), lack of time/resources (1), labor constraints (1), farm size concerns (1), government-related issues (1), and lack of incentives (3). This pattern reflects both financial caution and skepticism about the tool’s utility.

In contrast, dairy farmers did not view cost a barrier. Instead, two of three cited uncertainty of benefits and one noted time requirements. None mentioned complexity or incentives as barriers, suggesting greater confidence in integrating digital tools.

Interest levels also diverged. In the poultry conference survey (n = 13), only 3 farmers (23%) expressed interest, while 6 (46%) were not interested, and 4 were unsure or required more information. In the online poultry survey (n = 7), average adoption interest was low (2/5). One NESTT user rated potential CFT adoption at 4/5 if additional features were offered, while others scored it 2/5. A cost-free scenario raised average interest modestly to 2.67/5, indicating that financial barriers alone do not limit adoption; time, complexity, and unclear benefits also matter. Poultry respondents also did not identify specific incentives that would change their stance.

Dairy respondents, by contrast, showed strong willingness: All three expressed interest, with an average adoption likelihood of 4.33/5 and perceived usefulness rated similarly high. Their interest remained unchanged when cost was removed, suggesting they view CFT as a worthwhile investment. Two cited government incentives (e.g., subsidies, tax credits), and one mentioned grants as supports to encourage adoption, not to offset direct costs but to reduce perceived effort or risk through program participation.

5.5. Engagement Willingness

Beyond immediate CFT adoption, we examined farmers’ willingness to participate in broader sustainability initiatives and future collaborations (Figure 12). At the survey’s end, producers were invited to opt into future contact for follow-up studies or on-farm trials; several did so, indicating a subset willing to remain engaged in research and development. Interest in formal sustainability programs also reflected latent engagement. Six poultry farmers expressed interest in carbon credit or government incentive programs, though only two were enrolled. This gap between interest and participation likely reflects external barriers such as program complexity or limited awareness. Similarly, five poultry farmers expressed interest in GHG tracking despite not implementing it, while three had adopted monitoring (e.g., NESTT), with one farmer noting willingness to switch if CFT offered advantages. All three dairy farmers expressed interest in both tracking and adopting emission-reducing technologies, underscoring readiness for climate-focused initiatives.

Engagement was also evident in openness to new technologies. Poultry respondents mentioned solar panels (2), heat exchangers (1), and real-time emission monitoring (4), while dairy respondents highlighted renewables and robotics. Two poultry farmers had undergone energy audits, demonstrating willingness to collaborate with advisors. More broadly, farmers’ detailed qualitative responses suggest a reflective, participatory mindset: even CFT skeptics discussed sustainability practices, signaling readiness to engage in ongoing conversations and trials.

6. Discussion

The survey results are interpreted in this section in the context of the study’s objectives and the wider literature on farm management, sustainability, and adoption of decision-support tools like the CFT. It explores what the findings reveal about the current level of digital readiness, the main barriers and motivations for adopting CFT, and the sustainability practices already in place among poultry and dairy farmers in Atlantic Canada. Differences and similarities between the two sectors are highlighted to better understand where customization of the tool may be most needed. The discussion also considers the practical implications of these results for improving CFT’s usability and relevance in this regional context, as well as the limitations of the current study and opportunities for future work.

6.1. Digital Readiness and Tool Usage

The responses highlight the digital readiness (Figure 13) of Atlantic Canadian farmers. Overall, findings suggest moderate technological adoption, with significant differences between dairy and poultry sectors. The use of digital tools among all dairy respondents signals a high baseline comfort with technology in that group. This environment supports adoption of additional tools like CFT, as dairy farmers have available data (milk yields, feed intake, etc.) and skills to manage it, lowering barriers to using a carbon calculator. Robotic milking systems on all three dairy farms show their interest in investing in precision technologies, motivated by labor savings and herd monitoring benefits. This can be a potential indicator of farmers who more readily see the value in data-driven decision support tools (like CFT), especially if they can be integrated into their existing systems. This aligns with the observations in other regions; for instance, many European dairy farmers adopted precision technologies partly due to subsidies and industry programs.

In contrast, digital tool usage among poultry farmers had more variations in responses. The survey did not quantify poultry farms’ digital record use, but among the three online respondents, two used basic tools such as Excel or Google Sheets for record-keeping. Some poultry farmers (from the conference) likely used computers or smartphones for tasks such as finances or flock tracking, but the sample also included less digitized operations. This disparity suggests that CFT adoption will face a steeper learning curve in the poultry sector, where some farmers may not routinely use data software. These farmers may require more training and support to comfortably use an online GHG calculator. Conversely, some poultry respondents reported advanced equipment (e.g., automated feeders, climate controllers), suggesting that segments of the industry are technologically advanced. The key will be to align CFT with the existing workflows; for example, if a poultry farmer logs feed purchases and flock data in a spreadsheet, CFT could leverage those records to compute emissions rather than expecting separate inputs. Encouragingly, some farmers showed a willingness to adopt innovations such as real-time emission tracking or renewable energy systems. This indicates an openness to digital solutions if they address specific needs or bring clear benefits.

Another aspect of digital readiness is internet connectivity and device usage in rural areas. Although not directly surveyed, the fact that 23 farmers responded on paper or online suggests that most of them had basic internet access. If some farmers did not participate due to poor connectivity or discomfort with online forms, it suggests that digital tool deployment must consider access and interface simplicity. Tools like CFT need to be user-friendly for farmers with varying levels of tech-savvy. Since many older or remote farmers may prefer pen-and-paper, adaptive strategies could include offline data entry options or extension support (e.g., advisors inputting data on their behalf). Our findings also show that on-farm data collection and storage are not always consistent. For example, several farmers did not know their exact fuel or electricity usage, indicating their irregularity in tracking such metrics. To use CFT effectively, users must gather information on inputs, outputs, and practices. Thus, part of improving digital readiness is encouraging better record-keeping. Since many dairy farmers already keep detailed herd and milk records (often required for management), adding emissions-related record-keeping might be a smaller leap for them. In that case, one of the major challenges is integrating CFT into existing digital ecosystems on the farm. Ideally, CFT could import data from tools farmers already use, minimizing duplicate data entry.

Placing these results in context helps to clarify their significance. In European dairy sectors, adoption of digital tools such as CFT has been accelerated by subsidies and cooperative programs. New York dairy studies similarly show that uptake improves when tools are linked to efficiency and profitability rather than only environmental goals. In contrast, the U.S. rice sector demonstrates how adoption lags without clear incentives. Our survey findings show that dairy farms are technologically ready but lack supporting programs, while poultry adoption remains tentative in the absence of tangible benefits.

6.2. Barriers and Perceived Challenges

This study revealed several barriers and challenges that farmers perceive with sustainability tools like CFT, as well as general obstacles to implementing sustainable practices. Understanding these barriers is vital in strategizing how to encourage adoption and effective use of the tool. Figure 14 shows that the most frequently reported barriers to adopting CFT were cost and lack of time or resources, followed closely by uncertainty about its benefits. Fewer respondents identified complexity of the tool, government-related issues, or farm size as significant obstacles. This suggests that practical and perceived value considerations may outweigh technical or regulatory factors in adoption decisions. The main barrier for poultry farmers was cost, reflecting concerns that CFT expenses may not be justifiable. This cost concern includes direct expenses (e.g., subscription fees) and indirect ones such as time, labor, and potential equipment or software upgrades. It is interesting that cost was not echoed by dairy farmers in our sample as a major issue. A likely explanation is that dairy respondents, all with robotic milking systems, are larger or more capital-intensive operations, where an additional software tool is only a marginal expense. Poultry respondents, with a broader mix of farm sizes, included some who likely run on tighter margins or have less discretionary budget for non-essential tools. Therefore, CFT implementation must consider economic feasibility and demonstrate clear return on investment. Demonstrating CFT’s potential to save money—through efficiency gains or eco-certifications that yield price premiums—could mitigate cost concerns.

Another prominent challenge is the uncertainty of benefits. Farmers hesitate to adopt CFT unless convinced it will bring tangible operational benefits. This skepticism is rational; farmers face many new products and programs and often take a “prove it” stance before changing routines. To address this, pilot projects or case studies in Atlantic Canada could be instrumental, for example, showing a real farm where CFT use produced specific improvements (GHG reduction and cost savings, or efficiency gains). In the absence of local proof, European or Canadian successes may help, but local validation is more persuasive. The farmers also mentioned a lack of time, lack of personnel, and the complexity of tools as reasons they might shy away. If CFT is perceived as an administrative burden (data entry, learning software), adoption will suffer. This emphasizes the need to streamline the user experience via automation and integration. For example, if CFT could pull data from a milking system or feed silo readings automatically, this would reduce manual data entry. If governments or supply chain players (processors, retailers) begin to require or reward GHG tracking (through programs, subsidies, or premium pricing), farmers will have a stronger push to adopt tools like CFT. In Atlantic Canada, such incentives are currently limited, which could explain the relatively low participation in carbon programs. This suggests a potential strategy: collaborate with policy-makers to create or publicize incentives for on-farm carbon accounting and reduction.

A further obstacle is infrastructure and logistical barriers that indirectly affect sustainable practice adoption. For instance, in waste management, many farmers (especially dairy) indicated they do not recycle plastics simply because they lack access to recycling facilities. This barrier lies outside CFT’s scope but reflects broader sustainability challenges. Similarly, none of our participants had performed a carbon audit or calculation before, apart from one using NESTT, which could mean that the culture of carbon management is still nascent here. The learning curve itself is a barrier: many farmers are unfamiliar with concepts like carbon footprint, making education part of the challenge. Surprisingly, dairy farmers did not cite “lack of time” or “complexity” as barriers, though one mentioned time anecdotally. Possibly, the dairy farmers in our sample, being already engaged with tech, felt confident they could handle it. Barriers to behavior change were also evident in sustainability practices: for example, although some farmers expressed interest in carbon credit programs, they had not joined them. Barriers may include bureaucratic complexity or skepticism about program benefits, mirroring barriers to CFT adoption with unclear outcomes.

It is worth highlighting the difference in mindset between the two groups we studied. Dairy farmers appeared more solution-oriented, while poultry farmers were more cost/benefit cautious. Engaging poultry farmers may require addressing specific concerns, such as demonstrating how CFT could optimize feed efficiency (a major poultry cost) or improve manure management. If they see a direct farm management benefit (beyond just carbon accounting), they may be more inclined. Addressing these barriers requires a multifaceted approach: economic arguments or subsidies to offset cost, demonstrable results and testimonials to clarify benefits, user-centric design and data entry support to reduce time/complexity, and collaboration with industry/government to build a supportive ecosystem.

Overall, the barriers identified here—cost, time, and uncertainty of benefits—mirror the structural and behavioral challenges reported in prior studies of technology adoption. As seen in the literature, European livestock producers often cite administrative burden and unclear economic returns as major obstacles, while North American studies point to skepticism and limited incentives as persistent barriers, suggesting that Atlantic Canadian farmers face not unique but familiar challenges. Thus, similar strategies such as financial support, clear demonstration of value, and reduced complexity will be critical for overcoming resistance to CFT adoption.

6.3. Sustainability Practice Trends

Our results provide insight into sustainability practices and trends among Atlantic Canadian poultry and dairy farmers. A clear trend is that many farmers are already reducing environmental impact, especially through efficiency improvements. For example, most poultry and dairy respondents have implemented energy-efficient lighting and ventilation. This is a positive trend: farmers adopt these measures mainly to save electricity or improve animal comfort while also reducing GHG emissions. Such practices are often the “low-hanging fruit” of sustainability: relatively simple, cost-effective changes that also cut emissions.

The frequent mention of precision feeding by poultry farmers is another notable trend. Precision feeding (e.g., diets tailored to nutritional needs, phase feeding) reduces nitrogen excretion and methane from manure, improving feed conversion ratios. Several poultry farmers practice this, likely driven by feed cost savings, which also serves as an emissions mitigation strategy. In dairy, none mentioned precision feeding by name, though the use of robotic feeders by one farm suggests similar efforts to optimize feeding.

Another trend is the interest in renewable energy, even if current adoption is low. The survey responses indicate a growing trend towards exploring on-farm renewable energy generation. Motivators are both economic (offsetting power costs in remote areas with high electricity rates) and environmental (aligning with sustainability goals). As technology costs fall and more success stories emerge, we can expect more farms to adopt solar, wind, or biomass energy solutions.

Waste management practices show partial sustainability uptake. For instance, plastic pesticide or chemical container recycling/take-back programs are utilized by many farmers. This reflects the success of existing programs (often run by industry or government) that facilitate proper disposal of farm plastics. However, when it comes to other plastic waste (i.e., feed bags, baler twine, wrap), the trend is still dominantly landfilling or burning. Many farmers recycle or reuse some plastics, but landfill remains the most common destination for miscellaneous farm plastics. This trend highlights a gap: farmers will participate in recycling if convenient programs exist (like for chemical jugs), but lacking those, they revert to traditional disposal. All three dairy farmers noted lack of access to recycling facilities, underscoring a structural limitation: “improvement where systems exist, stagnation where they do not.”

We also see a trend in attitudes: moderate importance given to sustainability, suggesting that while farmers are not dismissive of sustainability, it might not be their top priority unless it connects to farm profitability or compliance. Environmental outcomes like lower emissions were mentioned, but economic resilience seems to be the primary lens through which sustainability is viewed. Another noticeable trend is the low participation in formal sustainability programs (like carbon credit markets or government-funded initiatives). Even though some interest exists, actual participation was minimal, suggesting that current programs are either not well-suited or not well-publicized to these farmers. It could also indicate that the programs (if they exist regionally) have barriers to entry.

A final trend to note is a forward-looking one: interest in knowledge and improvement. Many farmers, through opting for follow-up, asking for more information, or expressing interest in tools they are not using yet, demonstrate a willingness to learn and improve. If we continue providing accessible information and showcasing real-world benefits, these interested farmers can become early adopters who then influence others.

Taken together, these trends show that Atlantic farmers are adopting efficiency-driven sustainability measures, but often only where programs, infrastructure, or direct cost savings exist. This aligns with the literature on European and US-based studies. The parallels suggest that Atlantic Canada is not unique in its trajectory—progress in sustainability practices will depend less on awareness than on creating enabling conditions that make such practices practical and worthwhile.

6.4. Factors Affecting Interest in Adopting CFT

We explored factors affecting farmers’ likelihood of adopting CFT by analyzing survey questions on adoption interest. Responses were grouped as ‘Interested,’ ‘Not Interested,’ and ‘Unsure.’ Four contingency tables were created, pairing interest in CFT adoption with the following variables: poultry farm size, energy audit participation, province, and poultry farm type. Analysis was performed using R (v4.3.3) [81], and Fisher’s Exact Test [82] was used to determine whether there was a statistically significant relationship (p < 0.05) between the two variables. Fisher’s Exact was used, as it can support cell values of 0 and less than 5. No significant relationships were found, though results may change with more survey responses.

Table 6. Interest in CFT with respect to province of operation in Atlantic Canadian poultry farming.

	Interested	Unsure	Not Interested
Newfoundland and Labrador	1	0	1
New Brunswick	3	2	1
Nova Scotia	3	4	5
Prince Edward Island	0	0	1
Ontario	1	0	0

shows the interest of poultry and dairy farmers in CFT by their province of operation. With a p-value of 0.8382, the results show no significant dependencies between province of operation and interest in CFT. Similarly,

Table 7. Interest in CFT with respect to participation in an energy audit for poultry farms.

	Interested	Unsure	Not Interested
Participated in an Energy Audit	2	0	1
Unsure about Participation	0	2	0
Did not Participate in an Energy Audit	7	5	6

examines whether participating in an energy audit impacts dairy and poultry farmers’ interest in adopting CFT and shows no significance, with a p-value of 0.3181.

Table 8. Interest in CFT with respect to the type of poultry farm (according to bird types).

	Interested	Unsure	Not Interested
Broiler	3	1	5
Layer	1	0	0
Combination of Broiler and Layer	4	1	3

and

Table 9. Interest in CFT or similar tools with respect to the number of birds on the poultry farms.

	Interested	Unsure	Not Interested
Small (less than 10,000 birds)	0	1	1
Medium (between 10,000 and 50,000 birds)	3	3	4
Large (more than 50,000 birds)	4	1	3

show the interest of poultry farmers in CFT by poultry farm type and number of birds on the farm, respectively. With p-values of 0.7231 and 0.763, they represent no statistically significant relationships between CFT adoption and farm sizes and/or bird types. No significant relationships were found between flock size, participation in energy audits, province, or poultry farm type. These results are preliminary; further testing will follow with more survey data, including crop farmers.

6.5. CFT Customization Implications

The data gathered in this study carry several implications regarding how CFT should be customized and implemented in Atlantic Canada to maximize its effectiveness and adoption. Firstly, our analysis highlights specific gaps between the tool’s current capabilities and local farming practices. One gap is in dairy management: Atlantic Canadian dairy farms universally use robotic milking and zero-grazing (no pasture), which may not be default assumptions in CFT’s dairy module. CFT’s algorithms (based on IPCC methods) handle confined dairy systems, but some nuances may require tuning. For example, dairy herd structures include many heifers and calves raised on-site, and without accounting for emissions from young stock or replacements, CFT may underestimate carbon footprints. The farmers reported replacement rates of 23 heifers per month (or ~20% annually); ensuring the inclusion of emissions associated with rearing replacements (i.e., feed for heifers, their manure) is important.

Another gap is in dairy feed practices, which rely on partial/total mixed rations (corn silage, hay) without grazing. CFT’s emission factors should be adjusted to reflect Atlantic dairy diets, allowing for input of region-specific feed types and feeding methods. Similarly, the poultry manure management practices we recorded suggest a need for customization. CFT includes modules for manure management with default emission factors (stored manure, composting), but these may need validation or adjustment for local conditions (climate, storage duration). The latest version of CFT (v2.45.0) does not account for the effects of manure management in poultry. Anaerobic digestion has been shown to decrease lifecycle GHG emissions by up to 23% in egg production [83]. Therefore, including more information on manure management practices could provide more accurate results for poultry farmers in Atlantic Canada and elsewhere.

The variety of heating systems in poultry (gas, propane, oil, biomass, etc.) indicates that CFT’s energy section should include those options [3]. Custom emission factors for propane or oil may need to reflect Canadian fuel mixes if they differ. Another implication is the need to incorporate the range of feed types in poultry. Farmers listed everything from commercial mash/pellets to home-grown grain and even organic feed. If CFT expects a single “feed” input (e.g., kilograms of standard broiler feed), it may not capture feed composition differences. The high standard deviation in feed amounts observed (Table 5) underlines that one-size-fits-all assumptions would not work; customization should allow multiple feed entries or an average diet composition input.

One critical implication is the need for CFT to demonstrate value to farmers in tangible terms. Since farmers are skeptical of its benefits, CFT outputs should include metrics and interpretations relevant to them, for example, beyond carbon footprint in tons of carbon dioxide, showing potential cost savings or efficiency gains if they reduce emissions and effectively linking sustainability to farm profitability. Kano’s concept of must-have features means the Atlantic Canada CFT should incorporate the features that our farmers deem essential. Our findings suggest key “must-haves”: minimal time requirement, low or no cost, offline or limited-connectivity functionality, and assurances of data security and ownership. The surveys also highlighted areas outside carbon that farmers care about, like plastic waste management and water, and customizing CFT for Atlantic Canada could consider adding or emphasizing modules that reflect local priorities. For example, water was not a stressed issue in the responses (most have wells, and water scarcity is not acute in Atlantic Canada), but waste management was a concern (especially lack of recycling).

As more data are collected (including crop farms), regional benchmarks can be developed (e.g., average carbon footprint per liter of milk or per dozen eggs in Atlantic Canada). Customizing CFT to show how a farm compares to these benchmarks could be motivating, either to see that they are doing well or to identify specific improvement areas. However, care must be taken in how that is communicated to avoid discouraging those who are far from the benchmark. Since no adoption differences were found by province or farm size, customization does not require separate versions; one flexible tool can serve all, provided it allows for farm-specific data entry and remains relevant for both small and large operations.

6.6. Policy Implications and Support Programs

To practically achieve CFT adoption and net-zero emissions in agriculture, farmers will need supportive programs and policies.

Table 10. Summary of the government programs supporting adoption of GHG-reducing practices in Atlantic Canadian agriculture.

Program and Jurisdiction	Mechanism	Funding Range	Relevance to CFT Adoption
Agricultural Clean Technology Program—Adoption Stream (Canada, AAFC)	Cost-share grants for clean tech equipment	CAD 25 k–2 M per project (non-repayable; AAFC 40–50% cost share) [84]	Can fund equipment or systems that reduce GHGs (e.g., energy-efficient barn systems, solar). Could support hardware/software needed for CFT data collection.
On-Farm Climate Action Fund (OFCAF) (Canada, AAFC via provincial partners)	Direct payments for BMP adoption	Up to CAD 75 k–100 k per farm over program [85]	Funds practices like improved nutrient management, cover cropping, and rotational grazing that CFT may recommend; makes GHG-reducing practices financially viable.
Advancing Clean Technologies Program (Nova Scotia)	Provincial grant (60% cost-share) for climate-smart farm improvements	60% of costs up to CAD 150 k (tiers by farm income) [86]	Supports GHG-reducing tech in NS (energy efficiency, waste management). Can subsidize tools identified via CFT (e.g., efficient heating, manure handling upgrades).
Efficiency NS Agriculture BMP Energy Program (Nova Scotia)	Incentives/rebates for on-farm energy efficiency projects	Varies by project (often 50–75% of project cost) [86]	Helps farms conduct energy audits and upgrade equipment (lighting, cooling, etc.). Aligns with CFT energy module; CFT can quantify GHG savings of efficiency measures.
Anaerobic Digester Feasibility Program (New Brunswick)	Funding for feasibility studies (streams for farms and communities)	CAD 1.5 M program budget (covers study costs; per-farm support not specified) [87]	Encourages anaerobic digesters to cut manure-related emissions; if pursued, CFT will capture major emission reductions; program lowers feasibility barrier.
Sustainable Canadian Agricultural Partnership (SCAP)—Environmental Programs (Atlantic Provinces, fed–prov)	Various cost-shared BMP programs (provincial delivery)	Typically 40–75% cost-share; caps CAD 50 k–100 k depending on province	Broad support for environmental improvements (e.g., PEI resiliency, NB stewardship). Can cover nutrient management plans, precision ag, cover crop seed—measures reflected in CFT. Using CFT as a benchmarking tool can encourage adoption and progress tracking.

presents relevant federal and Atlantic provincial programs that could support the adoption of digital GHG tools, on-farm audits, renewable energy, and data systems. We compiled a shortlist of programs funding or encouraging sustainable farm practices and technology adoption, emphasizing those available in Atlantic Canada. As shown in the table, there are multiple funding streams that farmers could tap into when implementing changes highlighted by a CFT assessment. For example, a poultry farmer using CFT who identifies high-efficiency heating as an emission reduction strategy could apply to the ACT Adoption Stream or Nova Scotia’s clean tech program for a grant. A dairy farmer finding that covered manure storage or a separator would reduce emissions could leverage SCAP-funded environmental BMP programs in NB or PEI to cost-share that investment. The On-Farm Climate Action Fund, though focused on cover crops and nitrogen management, provides direct payments that facilitate adoption. By integrating CFT usage with such programs, we create a synergy: the tool quantifies benefits that justify the funding, and the funding enables the adoption of the tool’s recommendations.

From a policy perspective, the most relevant programs for CFT adoption in the region target energy and efficiency (ACT, Efficiency NS) and beneficial practices (OFCAF, SCAP environmental streams). Notably, programs like ACT and SCAP are already active in Atlantic Canada, so a near-term strategy could be to formally incorporate GHG assessment tools into their delivery. For example, provincial SCAP programs could require or encourage a “carbon assessment” (using CFT) as part of funding an environmental project, familiarizing farmers with the tool in a supported setting. Educational outreach through Environmental Farm Plan (EFP) programs could introduce CFT as a voluntary add-on, leveraging existing extension networks. Although not a government program, if processors or retailers start demanding GHG info, they might offer premiums or preferential contracts for lower-carbon products as part of supply chain incentives (i.e., a major grocery chain eventually asking meat and dairy suppliers for carbon metrics), as seen in some European retailer-led initiatives. Our small sample provides no evidence of this trend yet, but being prepared with tools like CFT is prudent.

7. Study Limitations

As an initial exploratory study in our four-phase CFT adoption, this paper analyzes the current survey results to establish an informed foundation for our future large-scale pilot project. While this study provides valuable insights, its limitations must be acknowledged. The primary limitation is the small sample size, particularly for dairy. The data is not a large or fully representative sample of the entire Atlantic Canadian farming community. Dairy findings should be interpreted with caution, as averages or trends may be heavily influenced by a single outlier. Similarly, the lack of statistically significant patterns observed is likely due to the low statistical power of such a small sample. As preliminary findings of our surveys, this limitation was anticipated, and more responses (including from crop farmers) will be collected to strengthen the analysis.

Another limitation is potential sampling bias. The poultry respondents were largely obtained from an industry conference and through outreach to known contacts. We applied non-probability or non-random sampling based on convenience and purpose. The potential participants or distributors were chosen according to their farming connections, expertise, availability, and prior personal and/or academic connections. Hence, our sample might lean towards farmers that are more engaged or interested in the subject matter. For example, poultry conference attendees may be more progressive or more connected to industry networks than average poultry farmers. Similarly, those who responded to an online survey or on-site interview may differ from non-respondents, potentially influencing results.

Geographic bias is also a concern, as all current dairy samples are from New Brunswick, creating gaps for Nova Scotia, PEI, and Newfoundland farms. Dairy farmers in other provinces may face different circumstances (e.g., milk buyer sustainability programs or provincial policies) not captured here. Similarly, our poultry sample had strong representation from Nova Scotia and few other provinces, but the one from PEI, for example, indicated no interest in CFT, which might or might not reflect PEI farmers generally.

Another limitation factor is time frame, as the surveys capture a snapshot of early-to-mid 2025. Agriculture is a dynamic domain, where awareness and interest in climate-related farming tools could be changing rapidly. For instance, a new federal program announced in mid-2025 could quickly alter farmer attitudes, which our data may not reflect. Current data show moderate interest in AI-based tool adoption, but this may shift with external factors such as market demand for low-carbon products or new government incentives.

Additionally, our surveys focused on poultry and dairy, not accounting for other significant agricultural sectors in Atlantic Canada such as beef, pork, or horticulture. Although the upcoming crop survey will address crops, they are not represented here.

The absence of complex analysis methods is another limitation. Due to the small sample size, we mostly limited ourselves to descriptive analysis. We applied a few Fisher’s Exact tests and found no significance, but more complex analyses (i.e., regression modeling of factors influencing adoption interest) could have been applied with a larger dataset. We can highlight correlations qualitatively with the existing data, but we cannot prove causation or clear predictors.

Finally, the lack of in-depth qualitative data, such as detailed interviews beyond the survey questions, can be considered as another shortcoming. We have short answers but not extensive farmer narratives or reasoning, limiting our ability to deeply understand the “why” behind some responses. For example, knowing five farmers are uninterested in carbon credit programs is useful, but knowing why would be more valuable—a nuance not captured in this study.

8. Future Research Directions

Although this study has laid a foundation for adapting CFT to Atlantic Canada, several key avenues remain to be explored for practical and sustainable deployment: broadening the scope of data collection, applying advanced analytical methods, improving data handling pipelines, tailoring the tool to specific agricultural sectors, and engaging in long-term validation with stakeholders.

8.1. Expansion of Survey Scope and Design

As mentioned earlier, the generalization of local requirements is only partially captured by the small amounts of responses. But, they were able to provide us with some basic idea of the Atlantic Canadian farming reality and the farmers’ perspectives. These responses can be used to refine future surveys to be more focused, targeted, and precise. Building on the initial surveys of poultry and dairy producers, future work will expand the survey scope to include crop farmers. Incorporating a crop farmer questionnaire will capture insights from grain, horticulture, and other field-crop operations that are integral to Atlantic Canada’s agricultural emissions profile. This broader coverage is important for designing tools that are truly “fit-for-purpose” across the diverse farm types in the region. In conjunction with expanding the audience, the survey design will be refined for clarity and effectiveness according to the lessons learned from the initial surveys, such as which questions yielded unclear answers or lower response rates. Each sector’s survey (poultry, dairy, and crops) can be customized with terminology and examples relevant to that sector, making it easier for respondents to understand and provide accurate information. The selection of target audience and distribution of the surveys will also be enhanced with industry associations and provincial agriculture departments to reach a larger and more representative sample of farmers. These refinements in scope and design are expected to yield higher-quality data and more nuanced insights, strengthening the foundation for subsequent CFT adaptations.

8.2. Advanced Analytical Approaches

With a larger and more diverse dataset, advanced analytical techniques will be employed to uncover deeper insights into technology adoption and user satisfaction. One planned approach is to apply the Kano model to the survey results to identify which features or aspects of CFT drive farmer satisfaction. The Kano model [77] is a product development framework that categorizes features into must-haves, performance attributes, and exciters based on their impact on user satisfaction. It can distinguish features farmers consider essential from those that pleasantly surprise, based on feedback about tool features and support services. It also highlights dissatisfaction from absent features (e.g., poor data integration) and satisfaction from added features. This will help in ensuring that the pilot system includes all “must-have” features to meet baseline expectations and focuses development effort on features that maximize farmer satisfaction and buy-in.

Advanced analytics will shift the project from descriptive findings to predictive insights, crucial for widespread CFT implementation. Applying statistical and machine learning (ML) techniques can therefore be explored to find predictors of willingness for adaptations. Techniques such as regression analysis or classification trees can be used to identify farm characteristics and perceptions that correlate with high interest in using CFT. For example, data may show whether farm size, farm management software use, awareness of sustainability programs, or perceived barriers (cost, complexity) determine adoption readiness. By training predictive models on survey responses, we can uncover the combinations of factors that best distinguish enthusiastic adopters from reluctant ones. ML could also help to detect subtle patterns in the responses, such as grouping farmers by their attitudes toward technology or sustainability. This can aid in crafting targeted strategies to increase tool uptake. If lack of time or unclear benefits are found to be common barriers, those can be directly addressed in future training and outreach, whereas if certain incentives or supports are strong motivators, they can be discussed and emphasized by government organizations and/or NGOs.

8.3. Sector-Specific CFT Adaptations

A major outcome will be sector-specific CFT adaptations tailored for poultry, dairy, and crop farming in Atlantic Canada. A top priority is to incorporate local emission factors and activity data to replace generic assumptions in the tool to represent distinguishable local characteristics compared to European agriculture. For example, dairy manure management practices and poultry feed types or barn climates differ from default assumptions. Future work will assemble the best available regional data (from literature and/or field trials) on emission coefficients for local practices, ensuring that CFT calculations reflect Atlantic Canadian realities. Hence, the precision of farm-level emission estimates will improve, giving farmers more confidence in the tool’s output. For instance, default methane emission factors for liquid manure storage will be adjusted to cooler, humid regional conditions influencing methane production. Similarly, emission factors for crop residue management or fertilizer use can be tuned to local soil and weather conditions, making the recommendations and benchmarks far more relevant to local users.

Additionally, sector-specific operational parameters and modules will be introduced by developing new input categories or management options within CFT to mirror common Atlantic agriculture practices. For poultry, CFT could include typical local barn heating systems and bedding materials that affect energy use and nitrous oxide emissions. For dairy, it could incorporate features for pasture-based feeding periods (since many farms here allow seasonal grazing) or local feed supplement types aimed at reducing enteric methane. A future crop module would address prominent regional crops such as forage corn, hay, or potatoes, including their rotations and cover cropping practices that influence carbon sequestration and emissions. Importantly, these adaptations will be guided by the needs and gaps identified by the farmers themselves. If poultry producers report challenges tracking barn fuel use, the poultry module will emphasize energy accounting; if dairy farmers report difficulty quantifying manure improvements, the dairy module will emphasize user-friendly manure emission estimates. This sector-by-sector tailoring will ultimately contribute to a more effective context-based overall strategy for agricultural GHG reduction.

9. Conclusions

CFT is widely used to estimate GHG emissions, assess water use efficiency, and identify opportunities for enhancing biodiversity. However, existing literature underscores a critical limitation: the tool’s default parameters often fail to reflect local conditions, particularly with respect to energy sources, crop and livestock types, fertilizer applications, environmental characteristics, and management timing. When evaluated with regional data—such as climate conditions, soil characteristics, and relevant agricultural practices—CFT has the potential to yield actionable insights for local farming practices and their influences on GHG emissions. In this study, we proposed a four-phase framework for AI-enhanced customized CFT adoption in Atlantic Canada and provided our contributions to the first phase (i.e., data collection). We designed, distributed, and analyzed two surveys (online and offline) targeting local poultry and dairy farmers. These surveys aimed to gather detailed information on farm-level practices relevant to GHG emissions. A total of 23 responses (20 poultry, 3 dairy) were collected from the on-going surveys to prepare a dataset for comparative, qualitative, and statistical analysis, generating insights into current farm management practices and local sustainability perceptions.

Farmer engagement through these tailored surveys yielded valuable perspectives on the feasibility of digital tools like CFT for GHG assessments in Atlantic Canadian poultry and dairy farms. Although the small sample size limits the generalizability of our results and their interpretations, they successfully highlight several critical needs: requirement to specify the local emission factors; practical cost–benefit analysis; and accessible, sector-specific digital tools. Although sustainability and digital innovation are not yet primary concerns for many farmers, there is a demonstrated openness to adopting new tools—provided that adequate support and incentive structures are in place. This foundational study establishes an initial reference point for understanding the GHG-related practices of farming operations in Atlantic Canada while also identifying key barriers such as implementation costs, time/resource demands, and unclear benefit/incentive structures (government and NGO supports). Additionally, our findings emphasize the importance of a collaborative approach—one that incorporates farmer feedback, industry perspectives, and academic expertise—in the development of context-sensitive and sustainable solutions. Moving forward, further data collection, including a survey for crop farmers, will be essential to extend the relevance and applicability of this work. Future works will include advanced data analytics, system integration, and pilot testing to evaluate the adapted CFT in real-world farm settings. Ultimately, a regionally tailored CFT could assist Atlantic Canadian farmers with a practical tool to reduce their environmental impact and actively support the transition toward net-zero agriculture.