First experience with university farm


A few weeks ago I had a great opportunity to meet with one of the farms at McGill (it was a rainy day so there was a slot when they didn’t need to be out in the field). Its really easy to interview and talk to someone who understands farming and their operations inside-out as well as the task I have at hand as a researcher. For example, right off the bat, when he didn’t know now much compost was being applied my survey respondent said “Well if you have a tape measurer feel free to go in back and measure the heaps of compost and the machine we use to spread the wood chips and then I can tell you how many loads we put on a year”.  He was also able to tell me what other initiatives at the university I should be contacting (most of which I had sent an initial email to) but more importantly the person I should be talking to. One really can’t underestimate the value of “local” knowledge. Doing the surveys in person is really letting me get more information that just the numbers I need to the P budget, its allowing me to get information to be sure a get surveys to all the right people.


I have found that although respondents don’t always have the quantitative data, they are more than willing to validate your calculations and give you access to the bags, sites, or information you need to make the calculations. With the eco-quartiers’ I surveyed a few weeks ago, I had to go find containers to measure and a little research on the size of fertilizer bags. I took measurements and pictures and was able to email the respondents and ask them if that was correct. In other words, I think it is very important to ask for details that will allow you to get the data you need. If they don’t know the size of a plot then offer to go measure it. If they don’t know the weight of the compost bags then ask if they have one lying around.

Taking pictures of inputs (who is that slightly creepy girl at the hardware store taking pictures of everything?)

When I was creating my survey I thought (and it was also suggested by colleagues) that I look at the type fertilizers and soils available on the island and the available sizes these inputs came in. However, after doing an initial compilation of what was advertized online, and seeing the variability I decided it wasn’t really the best option (as I couldn’t really make generalize what “a bag” size was or what the N:P:K of pellet fertilizer would be). In addition, I thought I was going to be doing the gardening survey online and the limesurvey question types were not going to be conducive to including pictures of lots of different bags and fertilizer options.

However, after doing a few surveys I found myself wanting to know about the range of inputs easily available to gardeners so that I could ask them the right “probing” questions and also be able to complete information about inputs if they only knew some of the information. For example, if they know it was a 30 l bag of soil of a certain company, I can look-up what the weight and maybe even some N:P:K ratios.

I bit the bullet and spent a weekend going to a Canadian Tire, Home Depot, Rona, and a Walmart (the big ones with lots of selection) and taking pictures of every type of garden (so excluding lawn or flower) soil, fertilizer, manure, compost, and mulch (and other amendments). I then made an excel spreadsheet with all the available info about each input, including brand, volume, weight, blend components, and N:P:K ratio. I then validated them online and looked for other products and sizes available for the brands sold.  I also looked at the websites for the smaller gardening stores on the island to include their products in my “master” list.

Unfortunately the Canadian government does not require companies to test for the nutrient concentration of the soils they sell. Thus, even though I contacted the largest companies that sell top-soil, potting-soil, and soil-compost mixes they were not able to tell me the origin, density, or the nutrient content of the soils. When soil was enriched with other amendments sometimes the N:P:K ratio was on the bag, but never for non-enriched soils. I contacted smaller garden centers on the island that sell their own soil mixes or do bulk orders in the hope that they might give me more information I can use to convert the mass of inputs I get from my survey questions into P flow values.


Problems with technology


The convertible tablet we purchased to be able to fill-in the surveys directly online has been a little bit of an adventure. It was acting really slow when we got it and did practice runs of the survey on it.  I knew we couldn’t have something slow in the field so I dropped it off at IT services. When it was “in the shop” so to speak, I was still spending about an hour per survey to do transcription and any of the calculations that I couldn’t be done one site (I got faster over time though).

Once we got the tablet back, it was responding much faster to commands but the sim card didn’t seem to be registering so we still could not use it in the field. It took a week of phone tag to get the phone company to explain that the sim card was registered and working but that because it was not a “real” tablet and was a computer I need to change setting on the computer to make it read the card. I am not that tech savvy and I have to admit it took my 4 days and a lot of blog searching to get things to work. But I did! Just before our first big week of garden surveys.

Unfortunately, once I got to an interview and tried to write using the screen as a pad, I realized the computer didn’t understand French and would not let me write anything I needed as it was making strange autocorrects that I could not fix (I would of have had to use the screen-touch keypad which somewhat defeated the purpose of having the tablet). After some more online searching I was able to download a French recognition package though. Still, I found it frustrating when the pad would auto correct my words to something I didn’t want in the next interview-survey I conducted.  During the meeting I ended up just flipping the tablet into computer mode and typing (which worked well and made it worth white having the sim card, but still a bit frustrating).

My assistant then took it out for surveys in community gardens. He had a little trouble seeing the writing on the screen with the sun, thought it was to heavy for biking from site to site, and thought that it seemed slower than simply writing on paper and transcribing.

This week, my other assistant and I conducted 2 surveys and again ended up flipping the tablet to computer mode as we thought it was taking us to long (and some of our attention away from the respondent) in a meeting with a busy actor. For the time being, when the surveys take place outside we will use paper copies and do the transcription after but continue to use the computer inside to minimize transcription time when ever possible.

First presentation at a community garden general assembly

We got lucky in the Ville-Marie “arrondissement” because the horticultural counselor called me back quickly and gave me the opportunity to come recruit volunteers at the annual garden meetings in his area. The first meeting I presented at was at Low Income Housing community garden (followed by 2 other garden meetings in the following weeks). Initially I asked him if I could attend the assemblies for the 2 largest gardens in his area (as I thought I had more chances of getting recruits) but he actually suggested also coming to this particular low income housing garden. He explained that the diverse immigrant community in the garden were likely to be using more inputs than some of the other gardens because they produce multiple crops a year and that getting volunteers at this garden would give me a little variety.

At the beginning of the meeting I was allowed to give a little 2 minute (bilingual because the community was very diverse) explanation of the project and then my field assistant went around the room asking volunteers to write down their name and contact information so that we can contact them once they were gardening, He also handed out a little card with my contact information in case they had any questions. We got 15 volunteers. I know they probably won’t all end up answering the survey but its still great to have a lot of initial sign-ups.

Before showing up to the meeting I had printed 40 copies of the survey with consent forms (half English and half French), sign-up sheets, contact information cards, and 20 copies of harvest tracking documents. I had also prepared a little speech explaining my research, what I was asking of them, and the benefit of the research and the fact that it would all stay anonymous and I would give them a little summary report at the end of the year. However, the horticultural counselor told me that with this audience it would actually be better to not show the survey and just say my project was generally about fertilizer application in community gardens. Once people have signed-up I can explain to the individuals the details, but he said that that much information up-front would be scary to them. I am really thankful that I have had access to such expert knowledge up-front and I will simply use all that material later on in the process.


Data collection: recap of the first two months

Its here, field season has started!

Although refining the collection methods is really never done (one needs to be iterative and reflexive about such things), I have now embarked on the journey of data collection. I have my online and paper version surveys, my base introductory email and phone speech, my Ethics Review approval, and my field assistants.

I have to admit I actually started my first meetings with stakeholders in March but they were mostly to explain the project or do a pilot survey. I really started contacting people at the beginning of April, so about a month and a half ago.

In the first few weeks, I made initial calls to “arrondissement” departments that are responsible for managing the community gardens in each of their areas. I contacted them to see if it would be possible to give a short 3 minute presentation at gardens’ annual “kick-off” administration meeting to recruit a few volunteers. I thought this would both be a more efficient way to recruit volunteers  (as opposed to standing in community gardens around the city trying to get people to fill-out the survey), as well as more respectful of everyone involved in community gardens, as I am asking for permission.

As anticipated however, its hard to get in contact with the right person at the right time. And we only got to present at meetings in one arrondissement (more on that experience in the the next post).

I also emailed and called all of the collective and institutional gardens and relevant city departments I knew about and used my  “standard” email and speech adapted to the situation of each garden. I always joined a copy of the consent form and the survey to emails. I realized after getting a few answers that the survey was a bit “scary” so I added a few sentences to the email about the fact that the survey looked long but could be done in 15 minutes in person with no great difficulty.

In the first 2 weeks I had a few meetings, which went really well, but then responses slowed down. I have had to call most people that I initially emailed, and then send them another email with all the information. I am now getting meetings though, which is great. A number of the actors want to be re contacted later in the season to schedule a meeting as they are to busy in april and may, which is fine. At least I have their contact information so I fell confident the surveys will eventually get filled out.

Last weekend we were able to visit 3 community gardens and get 3 or 4 gardeners in each garden to respond. The summer is starting to pick up!

Over the next few posts I will go into a little more detail on some of the encounters that have occurred since the beginning of the season to show how things have been progressing and how some element of my planning need to be adapted as the whole team (me and my assistants) actually use the data collection tools and protocols.


things are planted in the gardens and growing!

In the field: Montreal and UA at a glance

This summer I will quantify P cycling in UA on the island of Montreal.  Montreal is a good first study site because I speak both languages and I am familiar with the city and the UA initiatives and research teams working on UA.

The island has an (rural) agrarian history, as well as a UA history with community gardens since the early 1900’s (there will be a great master’s thesis published on this soon).  Over the past few years there has been a lot of attention on how to better support, coordinate, and expand UA initiatives in Montreal from the community, from research institutes, and from policy makers.

In 2011, the Groupe de Travail en Agriculture Urbaine (GTAU) collected 29,000 signatures to get a public consultation about UA in the city through the Office de la Consultation Public de Montreal (OCPM).  This was the first time citizens (and not a company) used this right to ask for a public consultation). The OCPM held the consultations in May and June 2012 and on the 3rd of October 2012 they published a 147 page report. The report summarized the oral and written presentations by community groups, municipality departments and citizens about UA in Montreal and also presented many recommendations for the City of Montreal. In response, the city announced they were going to create a permanent committee on urban agriculture where both selected community partners and city officials would co-jointly create and implement a plan to support UA on the island. Although this is far from the steps taken in other Canadian cities, like the Toronto food policy council or efforts in Vancouver, the committee is still a step in the right direction. I hope that the committee will permit a systematic organization of efforts in UA and the integration of good science in both practices and in planning.

Concurrent to political developments mentioned above, on-going research and community efforts have gained visibility and momentum over the past few years. The University of Quebec in Montreal (UQAM) has both a research lab (AU lab) and a student organization (the CRAPAUD) leading many research and outreach efforts on UA in the city. The Quebec Center for Biodiversity Science (QCBS) has also contributed to better documentation of trees in Montreal, which I think is relevant to UA because we can map fruit trees.  There is also a central calendar to share events relevant to UA in Montreal, including workshops, conferences, movies and much more . A radio show every week dedicated to UA. And tools (for example le guide du potager urbain and other websites) to help urban residents start and manage their own garden.

Its worth mentioning that the CRAPAUD have a cool website where people can register their garden (so we can find-out how much area is under cultivation in Montreal). Eventually it would be great to better understand yields and production of existing gardens like the open community-based research project Farming Concrete and better understand the available space for UA like the Urban Design Lab at Columbia University did for NYC  and the “croud-sourcing” data website can gather is a step in the right direction (and has an added advantage to allow for public “buy-in” to research). I think my research, and those of my collaborators will also contribute to UA quantification efforts and thus better recommendations.

There are approximately 75 collective gardens, 97 community gardens, 7 universities and colleges, 6 businesses, 10 farm producers and half of  Montreal households that say they practice some form of UA. I am excited to start quantifying flows in this system that is in the midst of putting in place a political structure that could greatly benefit from such research.


you can grow kiwis in Montreal!!

Training field assistants and building a field work team


Urban beehive downtown Montreal. Bees make me think of collaboration (as well as pollination ecosystem services and wonderful honey)

In order to get “first-hand” knowledge of P management (intentional or not) there are a lot of actors to contact. I am lucky enough to have a wonderful advisor who is excited and supportive of my project and thus gave me the opportunity to hire a part-time field assistant over the summer to help with data collection. She was also advising a really great undergraduate student for her honors thesis and saw some linkages in the students interests and my project and thus also teamed up with her for the summer.

It will me my first time leading a team for data collection so I am a novice and everything I say in this post might be wrong. Still, I thought it was useful to document the process so that I can’t get better over time, and that others might learn from my successes and my mistakes. Through this process I am trying to learn from my advisor and also integrate what I learnt “from the other side” when I was an undergrad working for grad students and integrate what I loved and learn from what I did not.


I put out a notice in the school newsletter and in several department listsevs that we were looking for a field assistant for the summer. I noted the skills we were looking for and a little bit about the project.  After getting emails from candidates and answering some of their questions, I selected 4 to interview in person. I came up with a list of questions, each one trying to determine if the candidate possessed the skill sets necessary for the position.

Skills I wanted:

  1. motivation (and thus punctuality and hard work all the way to the end)
  2. team work (taking direction but also capacity for leadership and collaboration (so independent but consulting for questions))
  3. data collection skills for surveys and field measurements (logic and attention to detail)
  4. adaptable (dealing with a new country and security and lots of different stakeholders)


I really want my field assistants to feel part of a team because I think that that is a better work environment and I hope keeps everyone motivated to do a good job. Ultimately I want the data collection to be done well and efficiently and I think that is more likely to happen if my team-members:

  • understand the project they are involved in
  • have “buy-in” the project
  • feel like they can ask questions
  • take initiative
  • have the resources (physical and human) to do the job

In order to provide an environment where all these things could happen I made everyone sign a contract with regards to what I expect of them and what they can expect from me. I also have a weekly meeting where we can discuss what we have been doing and what we will do. The meeting is an opportunity for training at the beginning, but more about learning from each-others experiences and coordinating efforts as we go along.  I write summaries of each meeting and send it to everyone to make sure key points are not lost (I am hoping to transfer this responsibility to my assistants though). I created a shared google calendar with color codes so that everyone can keep track of when people have meetings and are filling-out the survey with particular actors. I created a shared google document where we input data on all the actors we contact, who contacted them, a summary of the conversation(s) (or non-conversations) that have occurred with each (this document is also shared with our collaborators that are working with us for the case study city). This document allows us to make sure we don’t contact the same person twice and also “take-up” where someone left off if there is a change of plans. I also created a “field guide” document where I detail what I think are key points in protocol for contacting people, administering the survey, collecting and saving the data, and other important pieces of information that come along as we all gain more experience in the field.

Here are some of the important points I tried to highlight in the initial guide:

  • always have a tape measurer
  • take a lot of photos (we always need them for presentations and documentation but its something I often forget to do)
  • ALWAYS get the respondents consent (and give them a copy of the consent form)
  • Read all the questions and answers and don’t “lead” people to particular answers

In order to minimize bias originating from different survey givers I have:

  1. Made each field assistant read-through the survey and ask me questions and give comments.
  2. Made each field assistant administer the survey to a labmate of mine as a “test gardener” where I take notes and then tell them what worked and what didn’t.
  3. Made both field assistants come with me to observe how I administer the survey with actors “in the field”.
  4. We plan to do a minimum of 2 surveys a month as a team (a different person actual administering the survey each time) to ensure we are all still administering it the same way, and if not readjust so that we are in the future.

Training (for me and my assistants) is really a continuous process and thus I am trying to stay adaptive and reflexive throughout the field season.

Choosing how to contact people, how to administer a survey and collect the data [part 3]

Keeping records of the data and compiling survey responses:

I really wanted to find a digital tool to directly collect and store all the survey responses I was getting. I felt strongly about wanting to find this (these) tool(s) because over the years I have seen my friends struggle with transcriptions of interviews and survey results. Such a process is really time consuming, which means even though the survey only takes between 15 minutes and an hour to fill-out, you still have another hour of work on that one survey to digitize it to be able to use the results one day. In addition, there can be errors in the transcription process like putting information on the wrong line or not being able to read your (or someone else’s hand-writing).  Because I was collaborating, we also needed a central space to share data where we could all read it and update it, thus again wanting a digital option. Such concerns are actually also what let me to want people (or me) to fill the survey’s online.

I thus embarked on a journey to find a tablet that would allow me to digitally fill-in the survey and an application that would allow me to centrally and securely store the survey responses.

Here were the criteria I was using:


  • portable
  • sturdy (can resist being moved around on a bike and a plane, and withstand wind, dust, and a little rain)
  • long battery life (especially important in the field in Africa I think)
  • could take a sim card to do gps work and have internet access to look-up key information on the spot


  • could support at least 14 questions
  • able to export data for analysis
  •  protects privacy of the respondents
  • can make multiple-choice, tables, hierarchical questions, and drop-down menus
  • integrates explanation and pictures for questions
  • more than 1 language
  • free (or free to me through the university)

Here were my options:


  • Ipad or the a Lenovo thinkpad convertible computer

Applications (online or for a tablet directly):

  • online
  1. survey monkey
  2. qualics
  3. limesurvey
  4. gizmo
  5. make your own (like the household nutrient flux calculator)
  6. google survey (payed or just from the google drive options with your email account)
  • applications
  1. odk
  2. some mac specific survey apps

all these options had different costs and different functionality.

Here is what I chose in the end:

Lenovo thinkpad laptop that converts to a tablet with a sim card to be able to access lime survey.  Neither of these choices was perfect, but given financial restrictions and the functionality of the different options this made the most sense. Although bulkyer, the convertible laptop was a subsidized option at my university, and based on reviews is really good for field work as its sturdy and have an 8hour battery life.

The tablet is great to collect data digitally but it would not have been optimal to use a tablet application as we only have one and have at least 5 people collecting survey data. As such having the tablet with a sim card  (which was not the default option but we could get it installed) allows me to fill the survey out directly online when I have the tablet and allows others who are using paper surveys to transcribe everything into one central and safe location.

I chose limesurvey first and foremost because my University has a license and hosts the surveys you create on a secure server (thus I feel safe about the protection of the data, its free for me, and when I send out the link for those who want to fill-it out it looks very official).  Limesurvey didn’t have all the options for the response I wanted (like mixing fill-in the blank and drop-down menus in the same table) but I was able to find acceptable work-around. If satisfied all my other criteria.

IMG_3262 IMG_3264

Above are a couple of picture of the lenovo thinkpad we are using to administer the survey.

Note: It also seems really important to follow-up with your respondents after you do the survey. The follow-up builds reciprocity in case you need something later in the research project but also to not “poison” the field for other research projects using the same actors are you did. I am thus trying to remember to always send a thank you note and be sure that I send a 2 page report on my study to all actors that helped me (this means keeping a good and accurate list of everyone you contacted during the field work).

Choosing how to contact people, how to administer a survey and collect the data [part 2]

Administering the survey:

For some of the “larger” actors like markets and waste managers, I chose to try and meet them in-person to fill-out the survey so that I could be sure they were interpreting my questions the same way I was and to be able to clarify anything and ask follow-up questions. Ultimately however it seems that one needs to be flexible with how the actors feel more comfortable taking the survey (in person, on the phone,  on paper, or online).  Thus in my initial contact I favored the “meeting with them” option but also said they could fill the survey out in the other ways.

The real decisions were really about how to administer the survey to all the UA actors. There were a lot of actors and there were no good alternative ways to get the data other than the survey. I wanted to have as many actors as possible taking it (given that the data they provide are accurate however).  I really hesitated between in-person interviews so that I could make sure people were answering the questions “correctly” and I would have an idea of certainty of answers, and an online survey to get to the most people in the least amount of time.

However, when doing the pilot with friends and the first in-person survey, it became clear that it was generally necessary for me (or a collaborator) to be there to be sure people fully answered all the questions. Quantitative data seems straight-forward to me as scientist but its not that easy to do in a quick survey.

Still I was faced with the fact that I had a lot of people to go interview and thus needed assistants in addition to my great collaborators. I needed to train assistants to have good ethics, take good notes, and explain the project correctly as we want to minimize bias between data collectors (more on this in another post). In my contact with people I still offer the option to do the survey online but emphasize that its easier to fill-out in person because you have a resource to ask for clarification and help.