Moisture Management in Peaches
For years I’ve been researching how to optimize irrigation. Every year I am given more reason to believe soil moisture is the most important variable for plant health and yield.
While water savings may seem like the apparent goal to anyone paying attention to the water crisis in the Colorado River Basin, improvements in tree health are where the true opportunity lies. Moisture is simply the best toggle growers have at their fingertips.
The NRCS advocates an evapotranspiration-based approach to inform irrigation decisions. This relies on using the measured evapotranspiration (ET) in the local area to determine irrigation quantity. ET is the amount of moisture evaporating off the soil and transpiring out of the plant leaves (and then evaporating). The idea is simply to replace the water that is lost. There are a few crude assumptions on rooting depth, soil texture, and “max allowable depletion” to help determine the irrigation frequency. While this planning method is 10x better than no planning method, and probably quite good for a mature pasture, it’s flawed for my orchard for a few reasons:
My microjet sprinklers don’t have uniform coverage across the surface of the orchard. This massively changes the math and essentially makes the entire approach useless because my field doesn’t receive uniform water.
The assumption on “root depth” is provided for mature trees. But assuming a young tree has a 2-3’ deep rooting depth is obviously problematic. So what is the correct “rooting depth” to use for a young tree? Who knows. Similar to a vegetable crop, I believe managing moisture in the top 6” in year one is best.
Soil texture is based on Soil Web Survey data with a large scale (I believe 20,000:1). Texture often changes with topographic changes and with soil depth. In my case, four separate lab analyses of texture have shown three different results, all different than what the Soil Web Survey indicates. Changing this variable makes the math vary widely.
Acclima volumetric soil sensors.
Therefore, I believe a better approach is to use actual soil sensors, which the NRCS advocates as well. However, there is little guidance on how to use them. In previous years I’ve used volumetric soil moisture meters buried at various depths in my soil profile to help track my soil moisture.
The sensors are connected to a datalogger that stores the data and also displays it real-time on my office computer along with my weather station data.
In-field datalogger from Campbell Scientific.
I’ve programmed the datalogger to send me a text message every 12 hours with important metrics so I can track what’s happening on my farm.
If you are interested in a similar set-up for your farm, please contact me as I’m happy to help.
2024 Project
This year I decided to double down on the sensors. Instead of getting more volumetric moisture sensors, I decided to explore two alternative measurements that I thought may have more promise than volumetric water content. The first is a measurement of the soil’s water status, and the second is a measurement of the tree’s water status.
Teros 21 tensiometers from MeterGroup before being buried at three depths.
Soil Water Potential
Instead of volumetric sensors, I purchased three Teros 21 tensiometers.
Volumetric sensors measure what percent of the soil is water.
Tensiometers measure water potential: how much energy it takes to extract the water from soil pores.
While the difference may seem minor, tensiometers are technically a more direct measurement of soil moisture because they show what the plant root is actually experiencing and are independent of soil texture.
FloraPulse installed in trees.
Tree Water Potential
Measuring the water status of trees directly is a very novel concept. Previously this was only possible with a very expensive and time-consuming device called a “pressure bomb”. Luckily, a few researchers at UC Davis invented a microchip tensiometer that is installed in the trunk of the tree. It measures the water status in the xylem of the tree.
I believe that measuring the tree water status is the most direct—and therefore the best—measurement for optimizing irrigation. The crop is what matters.
With grant funding from the LOR Foundation, I purchased an additional six sensors to measure the moisture level of the soil and the actual trees simultaneously. The goal was to do a simple comparative analysis of the different kinds of sensors to determine which is best.
10 Lessons from the Data
I collected a season of data and will display it graphically. For easier viewing, I zoomed in on a two-week period of the season and will show the absolute value of the FloraPulse data.
This is what the three FloraPulse in-tree sensors look like. The green lines are sensors in one tree, while the purple line is a single sensor in a second tree.
Learning 1: The diurnal wave shows how the tree demands more water at the hottest time of day to cool itself. At night, it recovers by continuing to transpire.
This is an average of the three sensors. As the soil dries out over the course of a week, the wave trends upward as the trees get “thirstier”. The sharp reduction in the trend is due to irrigation.
I wanted to explore what factor was primarily driving the diurnal trend. This is the in-tree water status alongside temperature through the day, showing the strong correlation between water demand and temperature.
This is the in-tree water status alongside solar flux through the day.
This is the in-tree water status alongside VPD through the day, a combined measurement of temperature and relative humidity.
Getting back to tree and soil water status, this is what the volumetric soil sensors look like alongside the in-tree sensors at three different depths in the soil profile (6”, 12”, 22”).
Learning #2: After visualizing the data over multiple weeks, the volumetric moisture of the soil at 12” deep seems to be best correlated to tree water stress (dark gray line in the graph above). Moisture at 22” deep didn’t easily explain tree water stress. Moisture at 6” was correlated to water stress, but the value at which the tree experienced water stress was highly variable. The optimal depth to measure soil moisture will obviously change based on the tree age and root structure. I believe soil moisture measured at 6” is likely the best indicator of water stress for young trees.
For a bit more context, I’ll zoom out on a longer period of the season and display the raw FloraPulse data (not the absolute value).
This shows what all the FloraPulse sensors look like through the majority of the season:
This is what all the sensors look like averaged together.
Learning #3: FloraPulse in-tree sensors seem to be the best sensor for determining when to irrigate. While no targets for what defines “water stress” have been developed and validated for peaches, I chose to start irrigating when the sensors showed a daily maximum of about -8 bars, which was roughly weekly in the heat of the summer. In the graph above, water stress surpassed -10 bars in a few instances in early August and again in early September (oops!)
Learning #4: It was brought to my attention that the “recovery” of the tree’s water status at night gives invaluable information about general plant stress. While I was focused on the daily maximum reading of the FloraPulse, the nighttime minimum may actually be where the magic lies.
Learning #5: Moisture through the soil profile isn’t always linear and continuous, making moisture sensors less predictive of tree water status. This primarily happens when precipitation doesn’t infiltrate to the first moisture sensor. Here is one example to illustrate the point: on July 18th it rained almost ¾”. The rain didn’t infiltrate to the 6” sensor, but it absolutely reduced water stress in the tree (the tree roots were taking up water in the top few inches of soil!) The next day it was cool and overcast for much of the day. While the soil sensors indicated a dry soil at 6”, the overall water stress was low. The soil was wet in the top few inches, but dry at the 6” sensor. The FloraPulse told the true story.
This is what the volumetric soil sensors at two depths look like alongside the FloraPulse sensors.
Learning #6: While volumetric soil moisture at 12” seemed to be the best indicator of water stress, it wasn’t consistent. The FloraPulse would measure -8 bars when 12” soil moisture was between 28-31%.
This is what the tensiometer soil sensors look like alongside the in-tree sensors at three different depths in the soil profile.
Learning #7: The Teros 21 tensiometers didn’t seem to measure as well as the Acclima volumetric sensors. Two of the three Teros 21 sensors had to be replaced within six weeks under warranty. Due to this issue, good soil tensiometer data wasn’t obtained until late July. Additionally, tensiometers show 0 when the soil is above about 33% volumetric water content. The volumetric sensors always give an actual percentage, indicating “how wet” the soil really is. The tensiometer data also jumped around more, likely due to poor soil contact.
This shows what both types of soil sensors look like alongside the in-tree sensors at the 12” depth, which I believe is the best depth for predicting tree water status in my peach trees in their “third leaf” (~2 inch caliber trunks). This is a comparative graph showing the correlation between the three types of moisture meters available for growers.
Learning #8: FloraPulse in-tree sensors aren’t a great tool for determining how much to irrigate. While the data shows a drop in tree water status during an irrigation set, the best sensors for determining how much to irrigate are volumetric soil sensors at multiple depths.
Learning #9: FloraPulse in-tree sensors must be installed carefully in order to function correctly. Of the three sensors, one required replacement. In one installation, I didn’t hammer the sleeve deep enough into the tree to reach the xylem and a second installation attempt was necessary.
Learning #10: FloraPulse in-tree sensors only last one season. This limits their attractiveness for growers who want to re-use them as their primary sensor for determining irrigation each year.
Grower Recommendations
Use volumetric soil sensors at multiple depths in the root zone. Determine depth based on size and age of trees.
For one season, install FloraPulse in-tree sensors to observe when water stress occurs. This will help determine the optimal depth of soil sensor placement and the ideal moisture target for starting irrigation.
If the FloraPulse sensors prove to be overwhelmingly useful, install a new one each year. This is what I plan to do in my own orchard. If they don’t, continue to rely on your soil moisture sensors.
Rather than predetermining your irrigation duration, turn off your irrigation when all soil sensors at various depths show they are wet.