- Author: Emily C. Dooley, UC Davis
Algorithm for AI enables low-cost tracking of invasive plant
To manage johnsongrass, a noxious weed that crowds out cotton and sickens horses, farmers have tried herbicides, burning and hand-pulling. Now, researchers at University of California, Davis, have developed a more high-tech weapon against the invasive weed: artificial intelligence and machine learning.
Using photos from Google's Street View database, UC Davis researchers have tracked down over 2,000 cases of johnsongrass in the Western United States for a fraction of the cost and time that it would take to do drive-by or other in-person surveys. They call their tool Google Weed View.
The advancement could help land managers easily and quickly survey for other problem plants.
“Once the model is trained, you can just go and run it on millions of images from Google Street View,” said Mohsen Mesgaran, an assistant professor in the Department of Plant Sciences at UC Davis. “We have huge flexibility, and its capability can be scaled up very quickly.”
The technique can easily be extended to other plant species. All that is needed is to label the new item in Street View photos and train the algorithm to identify that object in the images.
By providing location information, Google Weed View also offers an opportunity to examine how climate affects the growth and spread of weeds and invasive plants at very large scales.
“I think it can be both useful for management and for people with interests in more basic questions in ecology,” Mesgaran said.
A colleague's query
Mesgaran began looking at using Google's photo database of roadways, streets and highways after Kassim Al-Khatib, a professor of Cooperative Extension in the same department, asked if he could survey Western states for johnsongrass.
Al-Khatib studies where johnsongrass grows, ways to manage it and how this perennial has evolved to be so prevalent and resilient. He's also working with scientists at the University of Georgia to decode the genome of johnsongrass, which is one of the top 10 most invasive weeds worldwide.
Johnsongrass can crowd out native plants, harbor pathogens and affect agriculture. It grows up to 7 feet tall with flowers that are green, violet, dark red or purplish brown depending on maturity, according to a UC Statewide Integrated Pest Management Program briefing page.
“Johnsongrass is a major weed not just in California but worldwide,” Al-Khatib said. “It's very difficult to control. It's a problem on vineyards. It's a problem for cultivated crops. It's a problem on orchards.”
Google Weed View allows for rapid, convenient scanning. It is continuously updated via everyday users with compatible cameras and images collected by Google. “Instead of a day of in-person driving, we can use AI to determine if johnsongrass is in a county or not,” Al-Khatib said.
Setting the parameters
To find the weeds, Mesgaran went to Google Street View, which hosts billions of panoramic photos. It didn't take long to find johnsongrass.
“The pictures are really good quality,” he said. “You can see plants and flowers.”
Street View's photos offer a 360-degree view, so in his request Mesgaran set parameters, based on street direction (bearing), to only see the side view. He also specified latitude and longitude, and other factors. To train the deep, or machine learning model, he chose Texas, where johnsongrass is prevalent.
A student sorted through over 20,000 images from that request to find pictures with johnsongrass and drew rectangular shapes around the weeds. They located 1,000 images.
The labeled photos were fed into a computer to train a deep learning algorithm capable of identifying johnsongrass in Google's images. The model was run again to capture potentially more images containing johnsongrass. These additional images were then labeled and used to further refine the model. With each iteration, the algorithm learned and became more accurate.
“This deep learning model was trained by these images,” Mesgaran said. “Once we had a semi-working model, we ran it against about 300,000 images.”
For Al-Khatib's request, researchers focused on 84,000 miles of main roads in California, Nevada, Oregon and Washington states. The team discovered 2,000 locations with johnsongrass.
Google Weed View cost less than $2,000 to purchase the images and teach the model. A traditional car survey to cover the same area would cost an estimated $40,000 in gas, hotel, food and other costs.
“In a matter of months, we came up with 2,000 records and I can do it for the whole U.S.,” Mesgaran said.
Next up? The entire United States.
This story was originally published on the UC Davis College of Agricultural and Environmental Sciences news site.
- Author: Trina Kleist, UC Davis
One more reason to adopt sustainable cultivation
California wheat farmers could both maintain their yields and improve soil health by growing annual wheat without tilling the soil year after year.
This could be one more encouragement to farmers to adopt a sustainable practice commonly called conservation tillage, no-till or minimum-till cultivation, impacting how we grow a grain that supplies about 20 percent of the calories and protein for people around the world.
A new study, by a team led by Mark Lundy, University of California Cooperative Extension specialist in UC Davis' Department of Plant Sciences, offers new insight for decades-long discussions around soil conservation, sustainable agriculture and climate-warming emissions related to growing our food. The study has been published in the journal Soil and Tillage Research. For the first time, researchers have shown that annual wheat that is not tilled each year is better for stashing carbon in the soil than perennial wheatgrass, while still yielding more crop in Central California.
Previous studies have looked at annual wheat that is tilled each year, annual wheat that is not tilled, and a cousin species, perennial intermediate wheatgrass (trademarked Kernza), which also is not tilled. But until now, no one has looked at all of the benefits and trade-offs together. Most importantly, “no one has ever controlled for tillage,” Lundy said. “And, no one has compared annual wheat to perennial intermediate wheatgrass over multiple years in a Mediterranean climate, which is what we have in California.”
This study also is unique because it delves into the deeper question of what is going on in the soil that drives the different results for carbon there. Soil carbon reflects various processes linked to plant activity and soil health. Measuring the different forms of soil carbon may also signal whether a farming system is accumulating carbon in the soil over time – a plus for reducing climate-warming gases in the atmosphere.
“Measuring soil carbon is complex and nuanced,” said Kalyn Taylor, the lead author on the paper. “We started this experiment because we wanted to know whether and how plant activity and tilling or not tilling would affect the carbon story belowground in California's climate.”
“When we started this study, we thought the crop being perennial or annual would drive the differences in carbon storage in the soil,” Lundy added. Specifically, they had expected perennial wheatgrass would lead to more carbon in the soil because of its deeper, better-established root system. “But that's not what we found,” he went on. “What we found was, it was the lack of tillage, plus the level of productivity of common annual wheat, that made the difference in soil carbon here in California.”
Soil carbon in annual vs. perennial grain
In 2017, Lundy, then-graduate-student Taylor, UC Davis Professor Emeritus Kate Scow and others on the team started measuring different forms of soil carbon in test plots at Russell Ranch, west of campus. Plots were planted with annual wheat that was tilled each spring, annual wheat that was not tilled and perennial intermediate wheatgrass (Kernza) that also was not tilled.
Each year, the researchers measured the carbon present in the soil, the amount of soil organisms (which have carbon in their bodies) and the amount of material the plants created.
At the end of three growing seasons, they found that land planted with no-till, common, annual wheat had the highest amount of soil organisms, measured as biomass, of the three treatments.
The researchers also found soil carbon is more likely to remain stable in the no-till, annual plots, compared to both tilled wheat and wheatgrass.
In addition, the no-till, annual wheat produced plant material more consistently than the perennial wheatgrass across the three years, which saw variation in rainfall.
“Overall, annual wheat grown without soil disturbance or tillage had both higher productivity and higher potential for storing carbon in the topsoil than perennial wheatgrass in our Mediterranean climate,” Lundy said.
Related research
“No-till annual wheat increases plant productivity, soil microbial biomass, and soil carbon stabilization relative to intermediate wheatgrass in a Mediterranean climate,” is online now and will be published in the January 2024 edition of Soil and Tillage Research.
The team also found that tilled annual wheat vs. Kernza stores total carbon at different depths in the soil profile and hosts distinct soil fungal communities, primarily in the root zone and topsoil: Taylor, K., Samaddar, S., Schmidt, R., Lundy, M. and Scow, K., 2023. Soil carbon storage and compositional responses of soil microbial communities under perennial grain IWG vs. annual wheat. Soil Biology and Biochemistry, p.109111.
Previous work comparing the perennial grain known as intermediate wheatgrass (trademarked Kernza) to annual wheat had not distinguished the extent to which soil health benefits are a function of the perennial nature of the crop. Read the story here.
This story was originally published on the UC Davis News site.
/h3>/h3>/h3>- Author: Emily C. Dooley, UC Davis
The project will also train plant breeders for the future
Wheat products account for roughly 20% of what people eat every day around the globe. As climate changes, wheat crops must adapt to new weather patterns to keep up with demand.
The University of California, Davis, is leading a five-year, $15 million research project to accelerate wheat breeding to meet those new climate realities, as well as to train a new generation of plant breeders.
“Everything is less stable,” said Jorge Dubcovsky, a plant sciences distinguished professor who is leading the grant research. “Everything is changing so you need to be fast. You need to be able to adapt fast.”
The grant from the U.S. Department of Agriculture's National Institute of Food and Agriculture will create a coordinated consortium of 41 wheat breeders and researchers from 22 institutions in 20 states. Researchers from Mexico and the United Kingdom are also participating.
Breeding needs to speed up
“Breeding crops for the future will require new traits, breeding platforms built for quick transfer of traits to elite cultivars, coordination of breeding efforts in public and private domains, and training for current and future plant breeders and researchers,” NIFA said in an announcement about this grant and others related to breeding.
The program involves on-the-ground research, identifying molecular markers and data analysis from multiple institutions to determine genes that will help wheat crops mitigate the effects of climate change. Plant breeding will follow to prove out those findings.
Wheat is unlike other crops in that 60% of the plant varieties — generating about $4 billion in annual production — are developed by public breeding programs rather than private corporations. In many states, wheat growers tax themselves to support basic breeding efforts at public institutions like UC Davis.
Increased coordinated research
The NIFA grant money will lead to more coordinated, sophisticated research. “This grant allows us to do breeding at a level that a good, modern company would do,” Dubcovsky said. “This grant is essential to maintain modern and effective public breeding programs in the U.S.”
The consortium will bring together data and research from across institutions, allowing for more expansive analysis while reducing redundancies. “We can take advantage of the data from everybody,” he said. “By doing that we don't need to duplicate efforts.”
A team in Texas will analyze plant images taken from drones at each institution to extract information about plant growth, water use, nitrogen levels and other data. “Using technology, we can see beyond our human capabilities,” Dubcovsky said. “You can extract a huge amount of information from every plant variety.”
The data from those images will allow researchers to document the plants throughout the life cycle and determine which plants fare better under certain conditions. Genotyping will help researchers obtain information about the plant genome. The combination of these two types of data could speed up breeding cycles, helping wheat crops adapt to a changing environment.
“If we can breed fast, we can adapt to change,” Dubcovsky said. “We are trying to make sustainable improvements in time.”
Training the next generation
The project will also train a cohort of 20 plant Ph.D. students in active breeding programs where they will participate in fieldwork, collect data from drones and DNA samples, and learn to integrate that information to accelerate wheat breeding. The students will participate in online and face-to-face workshops, as well as educational events and national scientific conferences.
Colorado State University, Cornell University, Kansas State University, Michigan State University, Montana State University, Oklahoma State University, Purdue University, South Dakota State University, Texas A&M University, University of Idaho, University of Illinois, University of Minnesota, University of Nebraska, University of Wisconsin, Utah State University, Virginia Tech, Washington State University, and U.S. Department of Agriculture Agricultural Research Service branches in North Dakota, Washington, Kansas and North Carolina are also participating in the consortium.
/h3>/h3>/h3>/h2>- Author: Mike Hsu
Giving 1,200-pound cows access to one of California's most fragile and biologically rich ecosystems seems a strange way to protect its threatened and endangered species.
But a recently published study suggests that reintroducing low to moderate levels of cattle grazing around vernal pools – under certain conditions – leads to a greater number and greater variety of native plants.
Ecologists consider vernal pools – ephemeral ponds that form seasonally – “islands of native habitat” amid California's grasslands that are dominated by exotic grasses. These biodiversity hotspots harbor about 200 native species of animals and plants, such as the coyote thistle, which germinates under water and forms a snorkel-like straw to deliver oxygen to its roots – and then “fills in” its stem as the pool dries.
Specially adapted to survive in those stages of wet and dry, many of these species are found only in vernal pools scattered across California – making those pools an urgent priority for conservationists.
During the 1970s and 1980s, vernal pools were fenced off in parts of the state, in the hopes of protecting the flora and fauna from grazing cattle. In the early 2000s, however, UC Davis researcher Jaymee Marty found that grazing was actually crucial to vernal pool biodiversity: once livestock were removed from areas that had been grazed historically, the diversity of plants plummeted.
“Her research was critical to rethinking the best ways to protect the diversity in California's vernal pool ecosystems,” Eviner said.
The Michaels-led study, published in the Journal of Applied Biology, builds on Marty's work, by looking at scenarios where cattle had been blocked from vernal pools for decades, and then observes the rate at which biodiversity returns after reintroduction of the animals. Michaels said she wanted to provide some initial answers to the practical questions that ranchers and land managers have in potentially reintroducing cattle.
“A lot of them had these areas that had been fenced off from grazing for the last 20–30 years, and they were very concerned about what happens if we let cattle back onto these vernal pool grasslands – are there going to be negative impacts because that land had been at rest for a few decades?” Michaels explained.
They discovered that, after reintroducing cattle to areas that had been fenced off since the 1970s, there was a greater abundance of native flora (species like the vernal pool buttercup, bractless hedge-hyssop and bristled downingia), as well as increased diversity among the plants (both in number of species and in how evenly distributed they were).
“Encouragingly, diversity is rapidly restored,” Eviner said, “providing conservationists with strong data to show that rapid action can enhance plant diversity.”
And as for potential worries about cattle making a snack of vernal pool plants, Michaels and her colleagues observed that the cattle appear to be more interested in munching on grasses.
“Anecdotally, we saw very few signs of herbivory on the vernal pool species because the timing is such that [the plants] are underwater for a good part of the late winter and early spring, and then by the time they're blooming, there's plenty of good forage around for the cattle,” Michaels said.
In fact, the cattle seem to be performing a function filled for millennia by native grazers (namely, the once-abundant tule elk), helping to knock down vernal pool species' chief competitor in those transition zones: the grasses.
Instead, microdepressions created by the cattle appeared to encourage the proliferation of native plants. Each hoofprint became a miniature basin – “a vernal pool within a vernal pool.”
“Right in those transition zones, where they could be hosting either the vernal pool species or the upland grasses, just a couple centimeters of soil topography can make a big difference,” Michaels explained. “If a cow comes and steps in that transition zone, and that lowers the soil surface so it stays inundated a little longer, you end up seeing these pockets of vernal pool species that are able to persist.”
Michaels is currently conducting a follow-up study on the hoofprints to pinpoint their role in boosting native plant abundance and biodiversity. Because the prints can last for several years, they might be able to deliver some enduring benefits – and land managers might not have to bring cattle in to graze the pools as often.
“If it's really the hoofprints making the big difference, maybe we don't need to graze every year – only during certain times of year when we know the hoofprints will form well and harden, and then we're good for a few years,” Michaels said.
- Author: Brad Hooker
Over the last three millennia, the practice of growing rice has evolved and spread throughout much of the globe. From China, through India, to Greece and parts of the Mediterranean and from Europe to the Americas, rice has demonstrated its versatility in desert regions and wetland deltas alike. Abundant in carbohydrates, it is today one of the world’s most widely eaten foods.
While University of California researchers develop rice varieties more tolerant to the modern challenges of climate change — flooding, heat stress, drought — California rice farmers each year discover more new threats in the form of non-native and herbicide-resistant weeds. So well adapted are these weeds that if left unmanaged, they cause rice yields in some places to plummet to nearly nothing.
The introduction of rice to California in 1912 was fraught with weed challenges from the start. The traditional dry-seeding method allowed barnyard grass to quickly overrun fields. While a new water-seeding technique suppressed the weed, it led to a whole other set of problems. In continuously flooded fields — still the most widely used practice in California today — an imported weed, late watergrass, flourished. Aquatic weeds took advantage of the new environment while others gradually became more flood tolerant. For many years, advanced herbicides allowed farmers to gain ground over these weeds.
Then, beginning in the early 1990s, several weed species, including late watergrass, were found to be evolving resistance against the most powerful herbicides. A metabolic resistance to one herbicide, researchers discovered, could lead to resistance for another.
Weeds also found new ways to outcompete rice. One invasive weed, Ludwigia, grows fast and tall — as high as 10 feet. Shadowing the rice plants, it spawns tiny seeds that travel well in water. Other weeds, meanwhile, are small and run along the ground to avoid combines and some emerge earlier in the season than rice, dominating resources.
In the Department of Plant Sciences at UC Davis, professor Albert Fischer’s laboratory is battling rice weeds on a variety of fronts: by researching the evolution and mechanisms of herbicide resistance, finding traits that make rice varieties more competitive, developing resistance techniques through field testing at the industry-supported Rice Experiment Station in Biggs, Calif., and by encouraging farmers to diversify management methods.
One system Fischer encourages is the stale seedbed technique, which allows weeds to emerge first from a reserve of seeds in the soil. Once that flush is up, farmers use a general herbicide to kill the weeds. At least one local farmer with a bad weed problem has controlled late watergrass this way. By replacing herbicides with shallow tilling, organic farmers can use this method.
With each management system is a different combination of growing techniques and herbicides, depending on weather, soil moisture and soil temperature, among other factors. Fischer’s team at the experiment station spends much of its time testing these herbicides on new weeds.
UC Cooperative Extension farm advisors encourage growers to also sanitize equipment, rotate crops, scout for surviving weeds and apply herbicide only when necessary, easing selection pressure on weeds while reducing environmental impact. Along that line, Fischer’s team is discovering how switching growing techniques and irrigation systems may be helping farmers meet higher environmental standards, addressing a trend of steeper water prices in California. Other researchers see this as an opportunity to reduce greenhouse gases released from decaying rice stalks post-harvest.
For each strategy, researchers weigh costs over benefits to select the right weapons for arming farmers entangled in this ongoing war with weeds.