Soil Moisture Sensors in the Greenhouse For Better Yield and Quality

Dr. John Lea-Cox discusses how soil sensors lead to water savings, increased yields, improved quality, and a more efficient and profitable operation.

Professor John Lea-Cox, University of Maryland Professor and State Research and Extension Specialist for the Nursery and Greenhouse industry in Maryland, presents techniques for using soil moisture sensor networks to monitor and control irrigation events in nursery and greenhouse applications. He discusses how set-point control leads not only to water savings, but also:

  • Increased yields
  • Improved quality of the product
  • A more efficient and profitable operation

Next steps


Our scientists have decades of experience helping researchers and growers measure the soil-plant-atmosphere continuum.


Professor John Lea-Cox is a University of Maryland Professor and State Research and Extension Specialist for the Nursery and Greenhouse industry in Maryland.


See all webinars

Part 1: How to Interpret Soil Moisture Data

To get the most out of your data, you need to know how to interpret unexpected occurrences in your data set.


Soil hydraulic properties—8 ways you can unknowingly compromise your data

If your data are skewed in the wrong direction, your predictions will be off, and erroneous recommendations or decisions could end up costing you. Leo Rivera discusses common mistakes and best practices.


Methods of Sampling and Analyzing Soil Pore Solution

Leo Rivera and Chris Chambers teach different methods for sampling and analyzing soil pore solution.


Case studies, webinars, and articles you’ll love

Receive the latest content on a regular basis.


Good afternoon. My name is John Lea-Cox, and I’m a nursery and extension specialist at the University of Maryland. I’m a professor in the Department of Plant Science and Landscape Architecture. Today, I’m going to talk about using sensor networks to monitor and control irrigation events in nursery and greenhouse operations. And typically, I think most people are familiar with the monitoring side of using sensor networks. But we’re engaged in a very nice project, long-term five-year project, with Decagon Devices as one of our partners, that is really looking at pushing the envelope on developing control strategies for nursery and greenhouse systems.

So just give you a brief outline of the webinar today. I wanted to just talk to you about, assuming that you don’t, not all of you know about nursery and greenhouse industry in the United States, talk about the industry, what it looks like, and some of the operational efficiencies in the industry. I’m going to focus in on the Chesapeake Bay in particular and some of the current challenges we face in irrigation, water, and nutrient management for the Chesapeake Bay. I’m going to then kind of switch gears a little and talk about our SCRI-MINDS project. It’s called Managing Irrigation and Nutrition via Distributed Sensing, that’s what MINDS means. Then really get into a little bit about sensor networks, what we’re doing, and particularly illustrate the need for control, and really focus in on a couple of case studies, where we’re actually looking at using various tools to actually achieve what we call setpoint control. And then just to finish up a little bit about water savings, our current water savings, and the economic benefits that we’re beginning to see from this project.

So just to give you a very, very brief overview of what our industry looks like. It’s really segmented into three types of operations. The first type is the field nursery operation. This picture here is a typical tree farm in soil, grass middles in the rows and quite intensive. As you can see, these maples are quite closely planted. Typically they could be thinned out in years three and four, but this is long term production in soil. Kind of an intermediary, this is what we call a pot and pot operation. It’s still growing large trees, but typically, these are grown in 10, 15, 25, 20, 25 and 45 gallon pots and that’s about 160 liters, perhaps 180 liters for a 45 gallon pot. So that’s quite big. These trees are also very large. Typically the production cycles are shorter though. And it’s a more intensive production scenario because what they do is they use soilless substrates in these containers, a soilless substrate being a pine bark or perhaps say some kind of artificial media where they are not using soil because soil is extremely heavy and expensive to ship. The next type of container nursery is probably one that most are familiar with. This is where a lot of smaller type shrub material, annuals, perennials, herbaceous perennials, are grown in smaller containers, typically anything up to about a 10 gallon or 40 liter size. But typically a lot of these annuals are in 4 liter, maybe 8 liter, one and two gallon size containers. And as you can see, they’re very, very diverse. A lot of these operations will grow perhaps 300 to 400 different species of plants, and not typically in contiguous blocks. This really does make irrigation management very difficult. A lot of these operations are on overhead irrigation. Some are quite tightly spaced, like you see in this picture. But then when these canopies get large, you have to space them out. And of course if you’re on overhead irrigation, your efficiencies really drop down. And we will talk a little bit more about that. And then lastly, greenhouse operations. This is perhaps an atypical greenhouse operation because this is just a picture of poinsettias. This is what we call a flood floor. This is a very advanced greenhouse. It’s actually courtesy of Rutgers University, who loaned me this picture. And so but there are a lot of greenhouse operations who are very advanced, recycle all of their nutrient and water in big, large underground systems. Some greenhouse operations are on micro sprinkler, or boom irrigation overhead, but typically, they’re very, very dense, small material, this is probably the largest type material you’d see in the greenhouse. This is perhaps a one gallon pot, or perhaps a little bit larger, what we call a 12 inch pot. And so typically, where your annuals and bedding plants are grown, what you typically see in a Lowe’s or Home Depot, those bedding plants are typically grown in a greenhouse operation.

So given all of these different diversity of operations in the nursery and greenhouse industry, you can imagine how diverse the irrigation systems are. And what me and many others in the United States are trying to do is actually figure out how we can make these systems much more efficient, in terms of not only their water use, but also their nutrient use. And I’ll show a little graphic in a few slides where I tried to illustrate how water management and irrigation and nutrient management are all interrelated. And so you really can’t just think about one particular part of an operation in isolation. You have to think of it together. Current deficiencies, as you can see, there’s a very broad range. And that broad range is because typically, with irrigation, overhead irrigation systems, depending on the plant size, you could be as low as 20%, what we call interception efficiency. And interception efficiency is the amount of water that you apply would actually be intercepted by any one given plant. Up at 60%, 70%, 80%, that’s when you get drip irrigation systems, single pot micro sprinkler irrigations. And so those typically are much more precise, they’re much more easily controlled, and of course, all of that water goes in the pot. So those efficiencies are much, much higher. Contrast that with our nutrient, nitrogen and phosphorus, efficiencies, those typically 10% to 60%. That’s for various reasons, because of different forms of fertilizer that are being applied. On the right hand slide, you can see the slow release fertilizer in that picture has been applied to the top of the plant, you do get volatilization losses, you get leaching losses. And so the plant really has to kind of fight to retain or actually compete in terms of nutrient uptake for those nutrients. So that’s why we’re really focused on irrigation management, because typically, this is the reality of many nurseries.

This is a container nursery, actually a very good container nursery, they have lined furrows, they recapture their water, that water is not just running off, it’s actually running off actually into a pond. And so what we do in many container nurseries is they do have these ponds on site, they do recycle a lot of that water, and so it goes back to three, four, or perhaps even perhaps five times to the crop. But of course, if it’s accumulating nitrogen and phosphorus in particular, you tend to get algae blooms in summer when the temperatures are warm. And so it becomes a runoff issue. Of course, we got pesticide and herbicide runoff issues. And so really the key to thinking about these systems is to understand that everything’s synergistic. If you look at the practice, in terms of nutrient rates and the type of irrigation system, you really need to understand how to manage that particular system. Because we’re really, and most growers are acutely aware of this, in trying to optimize efficiency. And lastly, if you can’t get it one way by optimizing efficiency, you really have to do something about it in terms of mitigation. But mitigation is expensive. And so what we’re really focused on is really the upstream side of things in terms of really reducing the amount of water being applied to containers, to plants so we’re really focused on being as as efficient as we possibly can with water, not because not only because it’s a precious resource, particularly in many states that are prone to drought, and water is very expensive, like in California, but because of the environmental issues as well.

So water management really is the key to nutrient management. And this is the graphic I was referring to, where I’ve just tried to put it all together on one slide. Obviously, if you’re an outdoor nursery, you do have rainfall, rainfall supplements irrigation, which is a good thing. But it’s very difficult to control. But we can control irrigation amounts. And so in terms of water application, we’re really focused on trying to reduce the leaching component, reduce the surface water component, and absolutely maximize the amount of water in that root zone at any one time. So we can minimize everything else in terms of leaching and runoff.

So just to switch gears, and just to focus a little bit on Maryland, because that’s where I’m working and working with many other colleagues on this project. This project I’ll tell you about in a little bit later, though, is a national project. And so we do face all of these issues all over the country. And the reason we’re focusing on the Chesapeake Bay, of course, is that the Chesapeake Bay is a very large watershed. It’s a very shallow tributary. And so when you have a watershed that 64,000 square miles in size, and it drains to probably one of the largest estuaries in the northern hemisphere, what happens is that it ends up collecting an awful lot of pollutants. And given that there’s about 17 or so million people in this watershed, there’s an awful lot of human and anthropogenic activity in this watershed, including farming. So this is just a graphic to show some of the water quality concerns in the Chesapeake Bay watershed. It’s across all states. Maryland has a lot because they’re the end of the, it’s the end of the streams. We deal with a lot of water quality concerns that literally just come downstream from other states. And so as of 2010, there was a presidential order signed by President Obama that actually imposed what they call a total maximum daily load process on the Chesapeake Bay because we’ve been trying to clean it up for about 20 or 25 years, and quite frankly, we hadn’t been doing a very good job of it, and we still aren’t. But now we’re under a EPA mandate to actually do something about it.

So just a little explanation of what a total maximum daily load is. That’s the amount of a pollutant, and it’s not just nutrients, it could be pathogens, fecal coliforms from animal operations, could be mercury from industrial applications, could be all sorts of things. What we’re concerned with as serving the farming community, of course, is nutrients and sediment. And so what that total maximum daily load actually means is it’s the amount of pollutant that a watershed can mitigate on a daily basis. That’s the total maximum daily load. And those criteria are set by the states and the EPA. And they are set to basically set standards for water quality that everybody has to meet, whether it be urbanites, whether it be farmers, whether it be industrial applications or even sewage treatment plants. Those mandated reductions, and there are two year milestones in this process, by 2025, we’re going to have to reduce almost 25% of the nitrogen, 24% of the phosphorus, and 20% of the sediment that goes into the Chesapeake Bay. And that is not just a Maryland responsibility. That’s all six states in the Chesapeake Bay watershed, Pennsylvania, Maryland and Virginia, the three biggest states, but it does include West Virginia, and parts of New York State, parts of Delaware, and a little bit of West Virginia as well. So that’s an awful lot of nitrogen, almost 200 million pounds of nitrogen, 15 million pounds of phosphorus and yes, that’s right 7.3 billion pounds of sediment. And in fact, sediment is one of our biggest pollutants because that is what submerges our aquatic vegetation. That destroys our oxygen capacity generating capacity in the bay, and it has all sorts of consequences downstream. So agriculture. You probably can’t read this on your screen, but the purple is pretty much what agriculture is contributing to this process. And it’s about 35 odd percent of the nitrogen, it’s almost, just starting up the numbers, 45% of the phosphorus, and it’s nearly 60% of the sediment load. And so we’ve got a major challenge on our hands to reduce these numbers, not to say that everybody else doesn’t; it’s a challenge for all of us.

So switching gears, and this is the team that I’m very happy to be leading and working with in our SCRI project. It’s a multi disciplinary, multi institutional group. Yhey are really, as you can see, a really fun group to work with. Colin Campbell and Lauren Bissey, right up front, they’re from Decagon Devices, but all the people, graduate students and faculty members and advisory panel members and all people. This is a list of people, some of the people, involved in the project. These are what we call the project leaders. But there are many, many other graduate students, undergraduate students, and staff working on the project and all contributing to the success of this. I won’t go into all of the details of our objectives, but it’s safe to say that what we are is trying to develop next generation tools for the nursery and greenhouse industry that are specific for our needs. And so the the engineers at Carnegie Mellon and Decagon Devices have been working very hard in the last couple of years to develop a new node that actually will not only monitor sensor activity, but it’ll also have a control capability and are very happy to that we’ve just started implementing those in the last six months or so.

I’m going to share a few preliminary details with you and some exciting results that we’ve had. But we are doing many, many other activities. We’re really trying to understand the physiology, the economics, all of the benefits associated with this. And I have a graphic, here we go, of what we’re really trying to do with these sensor networks as a team. So if you look on the left hand side there, what we have is nodes out in a production area. We’re very focused not just on a very small scale, like in the greenhouse, but we’re focused on maybe deploying these sensor systems into large nursery operations, those could be as large as we’ve got one nursery grower who we’re working with, we have eight nursery growers who are firmly embedded in this project in Georgia, Maryland, Tennessee, Ohio, and in Colorado. And so what those growers are doing is we’re actually implementing these networks on their farms. And so this is just a graphical representation of some of what we’re doing. Those nodes are out in different fields. We use indicator species to understand what crops we’re monitoring. So we have a few key species that we’ve chosen, that we’re looking really intensively at to monitor. But we are gradually expanding into different crops and different ornamental crops. We’re all focused on ornamentals. But as you could see, it didn’t matter whether it were a greenhouse or a tree nursery, we’re trying to cover the range of those species in ornamental species.

So what we’re really focused on, and what I’m going to illustrate today at the top, is local irrigation control. Now, that’s a control that is literally a node based control. That takes information from the sensors, and it is interpolated by the node, and the node actually makes the decision to turn a solenoid on and off. And that node is being developed, as I mentioned. It’s initially called an NR5, I’m not quite sure why it’s called an NR5, but that’s what it’s called at this point, but what it needs obviously behind that is a software backbone to be able to, for us to easily program it and to interface with that node. So that comes through a data station, either to a local computer on the farm, or it goes to a remote server, and most of our local control nodes are actually going directly over the internet from the computer so it can be accessed from any place at any time. We can also mirror that data on a remote server if we wanted to ensure that we had adequate speed on that on that network. So that I’ll get into a little bit, and I’m going to show you some graphics, so some actual data that we’ve recently collected from from one of our sites. Obviously, from a utility point of view, we really want to be able to access this data from any place at any time. And so we’re really focused on, at this point, all of our applications are remote, and we can pick them up on an iPad from or an iPad or a droid from anywhere in the world, if you’ve got an internet connection. We also have a 3G capability, which Decagon currently has already. So there are various ways that this information can be transmitted to a smartphone or a handheld device, anywhere in the world. And so depending on the situation, depending on how rural the situation is, there’s a number of options that we can use to be able to do this.

So this is questions that quite honestly get asked, we get asked quite often, so I thought I’d address them here. So why why sensor networks? A lot of people say, Well, why do you need sensor networks? Why is irrigation so hard? Well, actually, irrigation is not hard, is not easy. Probably the hardest question for a grower to answer accurately is do I need to irrigate today? Because an experienced person will be able to judge that and to say, well, Yes, it hasn’t rained for two days, and it’s been very hot the last couple of days. Or it’s been very cloudy and cool, so I really don’t need to irrigate, I could skip some blocks. But then there are some species that use a lot of water, that probably wouldt need to be irrigated today. And then there’s some other species which are very slow growing that quite frankly, you could probably skip. But somebody who really is not familiar with the plant species or their water use or their growth rates, it’s really hard to make that judgment. And it’s all about how much root they have, what’s their growth stage. And so it’s just become very difficult to make that decision correctly on a consistent basis. So obviously, most growers in our business do irrigate. But surprisingly, very few of them actually monitor, actively monitor, their practices. They’ll go out, they’ll lift pots, or they’ll look at trees, they’ll look for signs of wilt. But honestly, if it gets to wilt, it’s oftentimes too late. You’ve impacted growth rate already by the time you see crops wilting in the field.

So we need to really move from precision irrigation, which is what we’ve illustrated, you can do that with drip, with micro sprinkler, with boom sprinklers in a greenhouse. But what we need to move to is precision plus decision agriculture. It’s the decision part of agriculture or decision part of irrigation management, which really is the key to providing much better, much higher quality information to a grower. And growers, quite rightly, typically won’t change practice until you convince them that it’s either going to help their bottom line, it’s going to increase their profitability, or quite frankly, it’ll improve their productivity, for example, their growth rate. So that’s obviously tied to profitability. But quality is sometimes a very important consideration for many growers. So when we implement these sensor networks, we need to be very careful because that technology is a few key features that really, that technology needs to, it needs to be very sound, it needs to be cost effective, that means that it needs to be robust. The sensors need to work well, they need to be accurate, it means that you can set them up quickly and easily. You could probably do an initial calibration but in a short period of time, you could actually be collecting data. And of course, once those systems are deployed, you want them to literally have low maintenance as much as possible. Obviously, everything needs maintenance, even your car, and you wouldn’t not maintain your car. So networks do need maintenance. But more importantly on the back end of things, when you’re collecting sensor net sensor data, you can collect an awful lot of information in a very short amount of time. The true key to being able to use information, of course, is how you can effectively visualize it, how you can graphically show it to somebody who really doesn’t have the time to get into all the nitty gritty details of it. So you have to use software to be able to make these decisions in a very short window of time. And our decision window is a five minute decision window. We set a standard, if a grower can’t see what he needs to see in five minutes, we’re not we’re not meeting the challenge, we’re not really helping him improve his management because most often he’s time limited.

So onto a few examples, and I’m going to focus in on one particular example. This is a pot and pot nursery operation, that’s part of our project. We’re doing some very interesting work there. We’re focusing on two indicator species. These in the foreground, the dogwoods, that was actually a shot in winter, they’ve leafed out by now. But as you can see, he’s growing these particular trees in a 15 gallon container, that’s a 15 gallon container, and it’s what we call a pot and pot operation, as I mentioned earlier, so that pot actually sits in a what we call a socket pot. So it’s a pot in a pot. And they do that for two major reasons. Number one, overwintering to protect the root ball from extreme temperatures. And quite frankly, to stop them blowing over. It’s as simple as that. When you get a lot of these containers in a field and they blow over, it becomes a real mess. And so this system is typically irrigated with a micro sprinkler, you can’t quite see it there, I think I have a shot of it. But just by the handle of the pot, you can see a little bit of a loop. That’s the micro sprinkler there. And that goes to an underground lateral irrigation pipe, which is all then plumbed in, across the farm.

So this is just a graphic of, actually, it’s the webpage for our sensor network. This is not Datatrack, if some of you are familiar with Datatrack, this isn’t Datatrack. We are developing a next generation tool. We needed to develop a software tool that would be almost as good as Datatrack, but obviously have the control capabilities that we could actually talk to these nodes and we could actually send them commands. And I’m gonna get into that in a little bit. This is an overview of this one particular nursery. As you can see, it’s quite large, it’s about 200 acres in extent, just the pot and pot side of it. And so what this piece of software allows us to do, we call it Sensorweb, for the want of a better word, but it allows us to not only monitor the data in real time, but it also allows us to make adjustments to the irrigation schedules over the internet. And you may notice that those blue bars at the bottom of the graphic there, that’s what we call our macro scheduler. That’s where we can take our daily window of time and the light blue corresponds to the time during the day that we we can actually say okay, well we can irrigate these blocks. Not that it’s going to, but that’s the window where it’s allowed to actually schedule because obviously there are many other blocks on this farm and so we need to obviously overlap these windows so that all blocks can have a reasonable amount of water. We have two irrigation and monitoring control blocks on this farm. I’m just going to show you the one. We have a maple block which is two inch maple trees. They’re about 14 foot high at this point, 2 year old maple trees and then the dogwood block are three year old trees. They’re about an inch and a half in diameter, fairly sizable tree. And those are the lower part of the farm.

So I’ve described the macro irrigation scheduling control, but this is really the heart of our control capability in the software, and this is what we call our micropulse routine. So within each one of those windows, I’ll just go back, within each one of those blue windows, what happens is that means that the water is available to actually irrigate that block from the pump, and that basically configures the pump to be active. The micropulse routine then sets the time at which the pulse duration is active. So, in this particular instance, we’ve selected micropulse two. You can configure your own micropulse routines, you can see there’s a number of them there. But for this particular block, we’ve chosen that a micropulse it would be 120 seconds on, and then 180 seconds off. Now what that does is if the sensors get to a set point, and I’ll just go forward two slides, and then oh, let me go back. I’ll describe it when I actually get to some graphs in a little bit. But what happens is when the setpoint of the sensor’s average gets to a certain point, that micropulse routine will come on, and it will actually deliver 120 seconds of water to that pot, and then it will switch off for 180 seconds. And all the time, it’s actually still the sensors are still sensing. And if the setpoint actually gets above the minimum setpoint, it won’t irrigate again. But if it needs further irrigation, then it’ll go through another micropulse routine, and it’ll give it another 120 seconds. And so by doing this, these micropulses, what it basically does is it puts a very little amount of water, small amount of water on the top of the pot, and it lets it slowly move through the pot. And if it needs a second irrigation, then it’ll give another irrigation. By doing this, what we can do is minimize the leaching of any particular one irrigation event. Because typically, what we find is if you give a long irrigation duration to these containers, after a couple of minutes, it’s basically just leaching through the bottom of the pot, and your losing a lot of water and a lot of nutrients. And I’ll show you some data to actually prove that.

So this is the dogwood block at this particular operation. And just to show the solenoid farm for this particular block. And what we have done is selected two rows of trees. This is just looking towards the bottom of that rows. But looking up the hill, basically what we’ve got is a monitoring row, on the left hand side, there’s 133 trees in that row, and a control row at the right, and that’s another 133 trees. We’re actively monitoring five trees in each row. And what we’re doing is we have two 10HS Decagon sensors. One at six inches and one at 12 inches. We actually have some new of the GS-3 sensors in those trees as well. That’s the we’re monitoring EC in those. We won’t talk about that today. But that’s also part of our project where we’re monitoring the electrical conductivity of salts in those pots. The monitored row is actually scheduled by the grower. He’s the person who’s basically irrigating that row of trees. The trees on the right, the control row, are actually being irrigated by our team, and we decide at what set points we are doing those. Now, given that this, I’ll tell you that this operation is a long way from Maryland, it’s actually in another state, it’s in the state of Tennessee. So it’s not exactly something that we can just jump in the truck and go and visit very often. So it’s a true test of our long distance capabilities to be able to do that. Because of that, we have an EM50 R which is the typical wireless monitoring node at the end of this row, and that acts as a safety check for us and ensures that we’re getting adequate water at the end of the row and that we’re not irrigating the bottom end of this row because it’s uphill and starving the trees over the hill. So we have some safety checks in place.

So this is the inside of the NR5 node. This one’s a little special actually. This one’s, what we’ve done is we’ve configured this one to actually hook up to a latching solenoid. And that’s pretty interesting because that means that we can go into a farm situation, and we don’t need power to actually operate solenoid valves. The node actually operates that solenoid. So it turns it on and off. And it’s done with those five batteries that are in that node. And so thanks to our engineers and our partnership with Decagon, we’ve been able to achieve actually some very, very impressive engineering goals in terms of power management. As you can see, we’ve also got a flow meter on this particular row. In fact, it’s almost a standard practice for us now to put flow meters on these, particularly from a research perspective because that’s a very good tool for a grower to be able to look at, as well as ourselves, to be able to understand what are the impacts of these irrigation events on actual water use. The other four ports are taken up by the 10HS sensors in this particular row. And so the actual irrigation decision is made, each one of those senses is in a different tree, and so what we do is we make the irrigation decision based upon an average of four sensors.

Now just to show you some data. So as you see in this chart, this is the monitoring rows. So this is the row that the grower is actually irrigating. And as you can see, he typically irrigates three to four times a day. You might think that’s a lot, particularly if you’re from a agronomic background, but remember, these are quite large trees in relatively small containers, and they have very porous soilless substrate in them. So even the grower, prior to us implementing the sensor based control, he is and still is doing a very good job at what we call cyclic irrigation. Cyclic irrigation is where you put on a small amount of water three or four times a day, and that really helps to optimize water use. So just before I move off the slide, just take a note at how sharp those peaks are. Because that’s an interesting, that’s what I thought was a typical irrigation signature peak. And basically what that means is the volumetric water content, which is indicated on the left, is going up with an irrigation event. And the red line that is gradually creeping up across the slide, that is flow meter data, and that’s on the right hand y axis. And at the end of the couple of days ago, it was just about 2000 gallons he had used since the beginning of the month. So you can actually see the small increments in gallonage each day.

Now, this is the dogwood control block. And I’ve actually zoomed out for almost three months since we started this study. I did this for a very specific reason because I wanted to show you how this node was actually increasing the periods or the durations of irrigation over time. And as you can see, back in April, when it was relatively cool and moist, we were only irrigating once or perhaps twice a day, and those irrigation durations were very short, one minute or two minutes, at most. And as we got into May, those advanced to majority being two and sometimes four, and then when we get up to this in June, when it’s much much warmer, trees are growing very actively, they’ve really got a lot of leaf on them, they’re using a lot of water, that’s when we’re up to sometimes four and eight minute pulse times, so an eight minute would be four pulses in an irrigation event. But this shows you that in fact, the irrigation signatures on the control block are very different. With that micropulse routine, it means that we don’t get those very sharp water peaks because what that actually means is the water is going on much more gradually, and as it passes the sensor, it’s just very slowly increasing in water content. And then it’ll drop down again and then maybe rise up again. So you can see the three signatures. One tree is reading fairly high because it’s probably a little smaller or doesn’t have as much leaf. The other two trees are quite nicely in sync. And so what we do is we take an average of those trees. And so the on setpoint in this particular block, I just overlaid that, is currently at 46%. That’s where we are comfortable with that particular setpoint. That’s the setpoint that we also came to when we looked at the end of the row. We actually changed that a few weeks ago because we thought that we were underirrigating those trees a little bit.

So just to show you what the grower is doing, typically, he’s putting on four minute irrigation events, four times, sometimes five times a day. And that is a long duration time, which produces a very sharp peak. And what happens is, that not only uses a lot of water, but it actually results in quite a lot of leaching from that pot from any one of those irrigation events. In contrast, the control events are much shorter, and each pulse tends to be 21 gallons per two minute pulse. And so those irrigation events are much shorter, and so we typically are using a lot less water. And I’m going to show you some data, running total data at the moment to do that.

Before I do that, though, I wanted to just show you some of the leaching data that we’re collecting from these pots. And so before we changed the micropulse routine last month, we were getting a lot lower leaching from control trees. We did bump that up. And so you can see we’re getting somewhat higher leaching. But when you consider the volume, that’s in liters, so between a half and a liter, less than a liter, of leachate per tree, compared to about 80 liters of water use per day, that’s a relatively small amount of leaching compared to a normal irrigation event. That event right at the end of, well middle of June, the large bar there is a rainfall event. And obviously, we can’t do much about rainfall events, but we certainly appreciate having rainfall when we get it.

So just to show you some of the cumulative water use that we’ve had to date, the dog M is the dogwood M water use in liters per tree per day. But first figure is the accumulated water on a daily basis from the end of April, when we initiated the study, compared to the same water use for dogwood, the control trees in the same time period, so you can see 2.8 liters of water per tree per day being used by the trees that are on monitored irrigation, compared to just over a liter of water on average that we’re applying with the control trees using this micropulse routine. We also just thought I’d show you the maple data as well, so you can see maples are using quite a lot more water, they’re much larger trees, more faster growing. And so the monitor water use is about 4.7 liters a day, compared to our control application of about just over two liters of water a day. Now just threw in a rainy period there as well. The first part of June was actually quite rainy. So you can see not only our control volumes applications dropped, but our grower actually did a very good job as well of adjusting his irrigation times to match rainfall. So you can see his application rates dropped during that time too. The bottom part of that table gives our overall efficiencies. And so you can see, when we compare dogwood control versus monitoring, we’re running at about a 0.37. That’s should be 37.7% versus, so bottom line, we’re using about a third of the water in our control strategy than we’re using in our monitoring strategy. A little bit higher in our maples, but in that rainy period, that dropped down to about 38% as well. So there’s no doubt that we are, it’s early days yet, but we’re very excited about the potential for this kind of more advanced decision irrigation, and quite frankly, to the point where we might be able to take the human out of the loop on some of these irrigation decisions, at least once you get confident in the technology and the ability, it can certainly be used as an educational tool, even if it’s not fully used in total control mode.

Now, just to finish up, I just wanted to give you a little bit of an economic, some economic data that we’re gathering from the project. And this is from another nursery, in Georgia. And last year, some members of our team in Georgia implemented a similar irrigation strategy. This was not control from using an NR5, but what they did is they used control based on sensors and time. And so what they did is they instituted an irrigation regime, which was much more reduced basically on setpoint control, but it was initiated when they were in a kind of a manual mode. By doing that, and this gardenia is a very sensitive crop to root rot, what they did is in that particular crop, they actually managed to reduce the loss of gardenia from about 30%, which is typical for this particular cultivar, down to zero. We’re definitely hoping we can repeat that again this year, we have a repeat study. But what was interesting is that when you looked at the actual production time, it wasn’t just about disease management. What happened was, that crop grew much more rapidly, and, in fact, it was reduced from a 16 month production cycle to a 10 month production cycle. And so what happened is when we ran the numbers on it in terms of including disease management, the reduction in loss, but also the fact that they were able to produce that crop in a shorter production time, and they actually sold that crop in a shorter production time, that resulted in a net, over $1 per square foot increase in profitability for that crop. It was a small area, it was only a half an acre, but that was a significant benefit in terms of just that very small area. And what it basically meant was that we could pay for that sensor network in less than two months. So it’s certainly, just that one single crop paid for that sensor network, it would have paid for a much larger network actually, for that particular farmer.

So if you’re interested in our project and any of our partners and further information, you can go to our project website. It’s called smart farms I wanted to thank the USDA Specialty Crops Research Initiative Program for supporting us with this project. I wanted to really thank the growers who are involved in this project or late growers. They inspire us every day, and they really are probably providing us with the real information that we need to be successful on implementing these sensor networks. And so with that, I’d just like to say thank you. Thanks for your attention. And if you have any questions, I’d be very willing to take them at this time.

icon-angle icon-bars icon-times