A Coherent Measurement-Modeling Approach to Assess Soil Organic Matter Accrual in Agricultural Land
Soil organic matter (SOM) provides many ecosystem services. I will present our latest framework to conceptualize SOM structure, formation, and persistence, and a coherent measurement-modeling approach we implemented and use. I will provide examples of applications of our approach to quantify and forecast SOM changes under regenerative agriculture.
icon search Searchable Transcript
Toggle between list and paragraph view.
[00:00:00.870]The following presentation
[00:00:02.250]is part of the Agronomy and Horticulture Seminar series,
[00:00:05.840]at the University of Nebraska, Lincoln.
[00:00:08.670]So, hello everybody, welcome to the seminar.
[00:00:12.410]Today we have Dr. Francesca Cotrufo,
[00:00:16.420]she's professor in the Department of Soil and Crop Sciences,
[00:00:21.410]at Colorado State University.
[00:00:24.230]She earned his bachelor's degree
[00:00:26.830]from the University of Naples in Italy
[00:00:30.230]and her PhD from Lancaster University in UK.
[00:00:35.010]Dr. Cotrufo is a soil ecologist and biogeochemist,
[00:00:40.020]internationally recognized for her work
[00:00:42.350]in the field of litter decomposition
[00:00:45.150]and soil organic matter, dynamics,
[00:00:49.590]and also in the use of isotope methodologies
[00:00:52.760]in these studies.
[00:00:54.360]Today, she will present a seminar entitled,
[00:00:58.450]Coherent Measurement-Modeling Approach
[00:01:00.960]to Assess Soil Organic Matter Accrual in Agricultural Land.
[00:01:07.590]Dr. Cotrofu, whenever you are ready you can start.
[00:01:13.330]So again, thank you very much for inviting me
[00:01:16.000]and thank you for the nice introduction.
[00:01:20.320]I hope you can see my slide well.
[00:01:26.330]So I'm here today to present some of the work
[00:01:30.650]we are doing in order to provide a framework
[00:01:35.650]that sees the interconnected use of measurement
[00:01:39.960]and model to really inform how to best manage the land,
[00:01:45.655]to address some of the big challenges that we have today.
[00:01:50.900]And in particularly, when I talk of big challenges,
[00:01:54.740]I refer to the fact that, our growing population
[00:01:59.560]and also given the uneven spread of the resources,
[00:02:03.600]is really putting a lot of pressure on our planet
[00:02:07.890]to extract from it more and more energy,
[00:02:11.040]water, food, land, at the same time
[00:02:13.840]as we have modified the climate.
[00:02:16.500]And so for actually our planet to produce the food
[00:02:25.190]under climate change has become even more of a challenge.
[00:02:29.560]And on top of that, we also have to reduce
[00:02:34.910]that concentration of CO2 from the atmosphere.
[00:02:38.630]Right now, we are basically emitting into the atmosphere
[00:02:43.510]something like, 40 billion tons of CO2 per year.
[00:02:48.530]And while we need to move away from a fossil fuel industry
[00:02:55.780]and therefore reduce those emission,
[00:02:58.500]just a reduction of the emission will not be enough
[00:03:02.660]and we actually have to draw some carbon
[00:03:05.860]out of the atmosphere.
[00:03:08.000]And guess what, the point I want to make is that the soil
[00:03:12.110]is the best place to look for absorbing the carbon
[00:03:17.990]from the atmosphere, but also for addressing
[00:03:20.960]all of the challenges we just said.
[00:03:23.490]And that is because actually our soil is at the nexus
[00:03:27.750]of a lot of the ecosystems
[00:03:29.860]and provides a lot of the ecosystem services
[00:03:33.100]that are relevant for this food and energy
[00:03:38.750]and climate challenge I just told you about.
[00:03:42.607]And particularly when we think about carbon,
[00:03:47.290]basically in every ecosystem of the world,
[00:03:51.730]there is a lot of carbon in the soils.
[00:03:55.997]And even in ecosystem like forests
[00:03:58.700]where we think that the majority
[00:04:00.280]of the carbon is above ground,
[00:04:03.440]actually there is a significant amount of carbon
[00:04:07.360]below ground, both in top and deep soils.
[00:04:12.180]And of course that amount of carbon becomes even more
[00:04:16.260]in the organic deep soils of the boreal systems.
[00:04:25.720]And so, the things that we need to understand,
[00:04:29.790]is that on one hand we need to protect that carbon
[00:04:33.950]and maintain it in the soil,
[00:04:36.870]but then we also need to manage agricultural systems,
[00:04:42.380]so that they can increase carbon.
[00:04:46.858]And the lesson that we cannot go ahead
[00:04:50.430]with the same management we have used in the past
[00:04:54.870]came, for example, from many different studies,
[00:04:57.740]but in particular, this was very striking.
[00:05:00.130]And with Jonathan Sanderman,
[00:05:02.150]reconstructed the loss of carbon,
[00:05:05.260]that has happened throughout the human use of land,
[00:05:10.260]in particularly in grassland and cropland.
[00:05:13.500]And he estimated that we have lost
[00:05:15.880]about 130 petagram of carbon.
[00:05:19.240]And so, this study really stressed how business as usual
[00:05:25.920]is not possible anymore,
[00:05:27.630]otherwise we will keep losing carbon from our soil.
[00:05:31.880]But on the other hand, the positive side of it,
[00:05:35.280]is that actually those soils have now generated
[00:05:39.150]a lot of capacity to store that carbon,
[00:05:42.390]to put the carbon back in, if we modify our management.
[00:05:47.770]And so it opened what has been defined an opportunity
[00:05:52.450]that has been called
[00:05:53.970]the Climate Smart Soil Management Opportunity.
[00:05:59.630]And there have been a number of studies and reviews
[00:06:04.720]that are summarized here in this figure,
[00:06:08.730]which suggested how by doing regenerative practices
[00:06:14.670]in cropland, but also in grassland and for example,
[00:06:18.330]applying agroforestry in our other management,
[00:06:21.820]we can sequester on average somewhere between two
[00:06:27.050]and maybe five, if you see,
[00:06:29.990]across all these different studies,
[00:06:32.566]gigaton of the CO2 in the soil.
[00:06:36.990]And the good news is that if we do that,
[00:06:40.500]we really go into a win-win situation
[00:06:44.220]because restoring carbon and creating a quality soil
[00:06:53.313]and an healthy soil,
[00:06:54.987]also end up not only mitigating climate change
[00:06:59.560]or contributing to mitigating climate change,
[00:07:01.767]all of what I'm saying always have to come on top
[00:07:05.650]of, of course, getting off of fossil fuel emission use.
[00:07:11.550]But those both things will end up in improving
[00:07:15.090]the overall human wellbeing
[00:07:17.260]and also the socioeconomics of farmers.
[00:07:20.100]And so really there is a win, win, win situation
[00:07:23.430]that we need to use now.
[00:07:29.110]However, one and particularly, when we think about
[00:07:33.290]soil health soil management, soils are not done
[00:07:39.190]just to store carbon, we do not want to increase
[00:07:43.150]carbon storage at the expenses of the other services
[00:07:48.047]the soil provide, and in particularly fertility.
[00:07:51.630]And as Harry Janzen said now many years ago,
[00:07:56.260]we are posed in front of a dilemma,
[00:07:59.050]which can be summarized as have our cake and eat it too.
[00:08:05.020]In the sense that on one hand
[00:08:07.050]we want to increase soil organic matter storage
[00:08:11.170]and we have to increase soil input
[00:08:14.610]and reducing emission to achieve that.
[00:08:17.430]But at the same time, we also want to maintain
[00:08:20.650]an healthy soil so that it can recycle fertility
[00:08:26.640]to support plant growth.
[00:08:28.020]So we do want the soil to function and cycle,
[00:08:31.450]we don't just want to stock organic matter in there,
[00:08:35.280]but at the same time we want the soil
[00:08:37.550]to function efficiently, maintain a couple carbon
[00:08:42.130]and nitrogen cycle
[00:08:47.000]and at the same time be more efficient
[00:08:49.780]in the microbial transformation,
[00:08:51.900]so that per unit carbon input, more carbon stays in the soil
[00:08:56.840]and less goes into the atmosphere.
[00:09:00.290]And so in a nutshell, this is really the question
[00:09:03.300]that has driven my research now for several year,
[00:09:07.720]that can be summarized like
[00:09:09.810]how can we form organic matter
[00:09:14.580]has been turned over and persist?
[00:09:18.020]And one thing that I found that has helped me in forming
[00:09:22.190]how to do this in agriculture, is the fact that actually
[00:09:26.430]I haven't only studied agricultural systems.
[00:09:29.400]We have addressed this question all the way
[00:09:32.160]from the Arctic to the tropical forest.
[00:09:37.500]We have done it in cropland, grasslands of forest systems.
[00:09:43.870]We use some highly managed approach
[00:09:47.480]in particularly with isotopes
[00:09:48.970]that I'll be telling you about in a while,
[00:09:51.880]and of course, we do a lot of processing in the lab
[00:09:58.800]but my recipe, I guess, to do all this
[00:10:02.170]and maintain my people happy in doing this,
[00:10:05.400]is to really have fun while we do this work.
[00:10:08.480]It is very rewarding to work on soils.
[00:10:13.450]But let's start from the beginning,
[00:10:15.620]soils really form from the decomposition of plant input.
[00:10:20.940]And that's where I started in my career.
[00:10:23.300]I started my PhD studying mass loss from litter bags
[00:10:28.780]and then at a certain point, and mostly what we were doing
[00:10:32.180]was to ask the question,
[00:10:34.320]how quickly does a litter mass get lost from a litter bag
[00:10:44.200]where we had then trapped that mass in?
[00:10:47.630]And the idea was that basically,
[00:10:49.940]only the residues that wasn't trapped in the bag,
[00:10:53.440]would contribute to soil organic matter formation.
[00:10:56.770]And we did really ask what happened
[00:10:59.340]to all these other part of the carbon?
[00:11:01.480]Is it entirely mineralized to CO2, where does it go?
[00:11:05.710]A lot of the research also involved
[00:11:08.100]studying transformation in the soil.
[00:11:11.810]But a part of my research, in particularly
[00:11:16.250]I started this year with the journey,
[00:11:18.440]so my first student here at CSU,
[00:11:21.510]was to say, "Okay, has the litter decomposed?"
[00:11:25.200]Actually, a significant portion of the carbon,
[00:11:28.170]is not decomposed to CO2, mineralized to CO2,
[00:11:32.150]but is lost as dissolved organic matter
[00:11:35.980]and enter the soil.
[00:11:37.980]And so it also might have an impact
[00:11:41.260]on soil organic matter formation,
[00:11:43.780]besides what is formed from the structural residues
[00:11:47.140]that now I will call from now on
[00:11:50.870]particulate organic matter.
[00:11:52.990]And so this really started modifying our thinking.
[00:11:57.050]You can think of this like the T that forms from plant
[00:12:02.390]and again is not mineralized by (indistinct)
[00:12:05.750]And so following this line of thoughts, I started saying,
[00:12:09.490]okay, if a significant amount of organic matter
[00:12:12.903]entered the soil as dissolved organic matter,
[00:12:16.420]then actually can be very efficiently picked up
[00:12:20.630]by the microbes and make more of the microbial macromass,
[00:12:25.600]that we know is an important precursor
[00:12:28.310]together with the dissolved organic matter
[00:12:31.220]to what is considered the most stable form
[00:12:34.140]of carbon in soil which is the soil associated
[00:12:38.190]and bonded to mineral.
[00:12:40.310]And so I propose what has been called the MEMS framework,
[00:12:45.480]or the Microbial Efficiency Mineral Stabilization framework,
[00:12:50.910]suggesting that the plant material that actually
[00:12:54.630]were the one that decomposed faster,
[00:12:58.867]but were the more labile, the one that had more
[00:13:03.790]water soluble component or cellulose
[00:13:06.280]that could quickly be transformed into soluble material,
[00:13:14.340]formed more DOM and more microbial products,
[00:13:18.680]that then in soils that had high capacities
[00:13:22.370]for stabilization, so with a lot of silt and clay,
[00:13:27.490]would actually preferentially form
[00:13:30.260]mineral associated organic matter.
[00:13:32.890]Whereas the organic matter that were more lignified
[00:13:36.640]as less solubles, would actually stay
[00:13:41.080]in the floor soil organic layer for longer,
[00:13:47.180]but when decomposed would mostly result in CO2.
[00:13:52.150]And so to address this question, we mostly use isotope
[00:13:58.770]because we need to trace the carbon and nitrogen
[00:14:01.760]from the plant litter into the soil.
[00:14:04.360]And in my laboratory, in my group, we do this
[00:14:08.750]by producing isotopically and reach plant material
[00:14:13.790]in our label chamber.
[00:14:15.590]And then we use that both in laboratory or field studies.
[00:14:19.490]Once we use it in the field, we can study decomposition
[00:14:23.440]as if it was a natural environment,
[00:14:25.730]without the use of litter bags.
[00:14:28.740]We can also put lysimeter and collect exudates
[00:14:32.590]and do work with soluble and exudates
[00:14:37.080]and also burn the organic matter
[00:14:39.360]and do study with biogenic organic matter,
[00:14:42.290]which I'm not gonna talk to you today,
[00:14:44.850]but is actually an exciting also part of my research.
[00:14:51.000]Finally, more recently,
[00:14:52.600]we have started working with mycorrihizal
[00:14:55.430]and created mycorrihizal back
[00:14:57.540]to also have enriched mycorrihizal tissues.
[00:15:01.770]What we typically do,
[00:15:03.500]we trace the carbon and nitrogen all the way down
[00:15:10.360]from the litter into the soil.
[00:15:12.890]We use fractionation, we measure gas effluxes.
[00:15:16.790]We do some compound specific work.
[00:15:19.160]We have started some collaboration also to do NanoSIMS,
[00:15:22.970]but most typically we use the mixing model,
[00:15:25.680]the mass balance approach to determine
[00:15:28.520]the irrelative contribution of plant inputs into the soil.
[00:15:34.440]In terms of fractionation,
[00:15:36.210]I use physical fractionation approach.
[00:15:39.770]And with we typically,
[00:15:42.900]we use this understanding of soil organic matter
[00:15:46.080]as made primarily of mineral associated organic matter
[00:15:51.400]and particulate organic matter that can be found
[00:15:55.300]either free or included in different
[00:15:58.020]aggregate size fraction.
[00:15:59.860]And then there is the heavy course organic matter
[00:16:03.340]which is sense size fractions.
[00:16:06.360]And we do different type of fractionation
[00:16:09.750]depending on the questions,
[00:16:11.310]but also on the size of the work.
[00:16:14.150]For sure, just as simple by size POM
[00:16:16.820]and MOAM fractionation can already
[00:16:18.830]be enough to address a lot of questions.
[00:16:22.450]Anyway, doing these approaches
[00:16:25.430]in the one of the first study was in Kansas,
[00:16:28.600]in the using big bluestem,
[00:16:32.420]what we observed adding a labeled big bluestem in the field
[00:16:36.840]was that first of all, without litter bags in three years,
[00:16:40.330]all the litter had disappeared.
[00:16:42.660]And at the same time, the carbon from the litter
[00:16:46.310]was recovered in soils,
[00:16:48.380]but with different efficiency over time.
[00:16:52.660]After analyzing also the carbon and nitrogen
[00:16:57.300]in different fraction and in the microbial biomass,
[00:17:01.760]and linking this to the different form of carbon
[00:17:06.110]that were lost progressively over time,
[00:17:09.070]we propose what we call the two pathways
[00:17:11.940]of soil organic matter formation
[00:17:14.370]by which the dissolved organic matter
[00:17:17.460]that is lost from the litter soluble
[00:17:21.010]and from the cellulose that are not en classed in lignin.
[00:17:25.350]This can either directly go to form
[00:17:28.510]mineral associated organic matter,
[00:17:31.070]or after being transformed by microbes
[00:17:35.275]and form mineral associated organic matter
[00:17:37.410]as microbial necromass.
[00:17:39.630]Or the litter, the residues that are formed of the fiber
[00:17:45.160]from plant and also in the meantime, microbes
[00:17:48.360]and fungal necromass building the litter,
[00:17:51.170]those structures fragment and enter the soil
[00:17:54.890]as light fragments through physical transfer
[00:17:59.750]also mediated by soil fauna and those form,
[00:18:03.330]what we call the particulate organic matter in the soil.
[00:18:08.050]And so, we have worked a lot on trying to understand
[00:18:13.010]more and more whether this two pathway of formation
[00:18:16.880]is really true, and we have now a lot of studies
[00:18:20.930]and more and more confirmation that is in fact true.
[00:18:24.370]And so if you add, for example, in this study,
[00:18:27.330]we added root residues from two different lineage
[00:18:32.670]of corn that had different lignin content.
[00:18:37.530]And what we found was that once you add root tissues
[00:18:42.030]in particularly if you add them adapt,
[00:18:44.590]you mostly form particulate organic matter and more
[00:18:49.010]in the Z13 was the corn variety
[00:18:54.840]that had the more lignified roots.
[00:18:59.230]On the other end, when you add the root exudates,
[00:19:02.880]those form mostly mineral associated organic matter.
[00:19:07.440]And again, the more is from the root
[00:19:10.360]that have the highest water soluble.
[00:19:13.470]And so this is just one example.
[00:19:16.230]We have done many other studies now,
[00:19:18.917]all confirm these two pathways where the POM
[00:19:22.390]is formed from the structural
[00:19:24.270]and the MAOM is formed from the soluble.
[00:19:27.600]The other things that we are started looking at
[00:19:30.920]is what controls the efficiency
[00:19:35.640]of formation of POM and MAOM.
[00:19:39.190]And we, this was the results of a laboratory incubation
[00:19:43.520]with the variety.
[00:19:44.630]These were litter residues
[00:19:46.240]at different stage of decomposition.
[00:19:48.250]And the blue is the leachate.
[00:19:51.030]And what we noticed was that the POM is actually promoted.
[00:19:56.030]The efficiency of POM formation is promoted
[00:19:58.660]by both the microbial diversity
[00:20:03.300]and the chemical diversity in soils.
[00:20:05.650]Those were top soils and those were sub soils.
[00:20:10.930]Whereas for the MAOM, the microbial diversity
[00:20:14.900]actually reduce the efficiency of formation.
[00:20:19.010]And we are still trying to do subsequent work
[00:20:21.790]to better understand what's going on here.
[00:20:25.390]But what I want to leave you is that the two pathway model
[00:20:31.430]is really contrasting the original idea
[00:20:37.180]that organic matter slowly formed
[00:20:40.950]from the progressive decomposition
[00:20:43.570]of particulate organic matter
[00:20:45.580]and showing actually that those two pathway are separate.
[00:20:50.080]And so, if in the soil we want to,
[00:20:53.430]if we want to sequester more carbon in soil,
[00:20:56.700]we need to think about different strategies
[00:20:59.900]that form either the particulate organic matter
[00:21:03.570]or the mineral associated organic matter
[00:21:06.120]because they're not the same.
[00:21:08.010]They are formed from different pathways I showed
[00:21:11.090]and also probably have different properties
[00:21:14.360]and function in soil.
[00:21:15.690]And actually this is the one of the research objectives
[00:21:19.720]that we are working mostly now,
[00:21:21.830]is to linking POM and MOAM to properties and function.
[00:21:25.810]But what we do know from the experiment
[00:21:28.290]we have already conducted,
[00:21:30.110]is that they do actually have different vulnerability,
[00:21:33.900]different potential to sequester,
[00:21:36.400]and they respond differently to management.
[00:21:40.190]In particularly in a study that I did in Europe
[00:21:43.540]across a big dataset, these were grasslands in blue
[00:21:48.950]and forest in red, we showed that you need more nitrogen
[00:21:53.940]per unit carbon to form POM
[00:21:57.270]and instead to more nitrogen per unit carbon to form MOAM
[00:22:01.440]that has a much more restricted carbon to lower
[00:22:06.510]and smaller variation in the carbon generation ratio
[00:22:11.330]than the POM in particularly for system.
[00:22:15.320]But also the mount saturate,
[00:22:18.800]there is an upper limit.
[00:22:20.200]We're doing a lot of work now to try and understand
[00:22:22.900]what the term in this upper limit's not just the silt
[00:22:26.170]and clay in the soil.
[00:22:27.450]There's a ton of other stuff
[00:22:28.680]that control the upper limit,
[00:22:30.470]but there is an upper limit for the amount
[00:22:34.090]per unit gram of soil that can suppress MOAM,
[00:22:39.300]whereas there is not upper limit for POM
[00:22:41.760]because being fragmented, not require mineral to stabilize,
[00:22:45.940]you can imagine to have as many of them as you want
[00:22:51.170]and just increase the depth of the soil,
[00:22:54.860]or for example, reduce the bath density.
[00:22:58.660]The other thing that is coming by measuring 14 C
[00:23:01.980]is that the POM and the MOAM
[00:23:03.720]while there is some significant overlaps and again,
[00:23:06.500]we're working out to better understand how to separate
[00:23:09.520]those pulling and whether there is less overlap.
[00:23:11.910]But this I think is also dependent on the fact that
[00:23:14.273]this data across all type of systems,
[00:23:17.590]but typically on average,
[00:23:19.240]the mineral associated organic matter is older
[00:23:23.073]than the mineral associated organic matter,
[00:23:25.420]so persist for more.
[00:23:27.310]Then my student Kate Rocci did a meta-analysis
[00:23:30.569]which she also demonstrated that they respond differently
[00:23:34.980]to a number of different global change prediction.
[00:23:42.340]And so putting on and in generally POM is more responsive.
[00:23:46.780]And so putting these all together,
[00:23:49.830]we are now working towards a framework
[00:23:53.960]that really help us understanding how a soil is structured
[00:23:59.080]and therefore how much carbon we may expect in there
[00:24:03.800]to be vulnerable to changes
[00:24:06.110]and how much the nitrogen like demand it is.
[00:24:09.980]So typically what you would have is that in soils where
[00:24:17.360]the carbon accrue as mostly plant input,
[00:24:21.540]because the micro that physiologically inhibited
[00:24:24.860]by lack of oxygen or a very cold temperature.
[00:24:29.410]You can imagine like the well wet peat soils,
[00:24:34.620]in this case, the soils are very reach in carbon.
[00:24:39.010]There's mostly in POM but it's very vulnerable to changes
[00:24:43.030]if for example, that wetland is drained,
[00:24:46.000]or if the permafrost warm,
[00:24:54.450]then the micro become active and can go at that carbon
[00:24:58.020]and can start breaking it down.
[00:25:00.370]On the total other end,
[00:25:01.800]you can imagine soils where the majority of the plant input
[00:25:06.880]have gone microbial transformation.
[00:25:09.840]And therefore the carbon is actually much less.
[00:25:13.640]There is much more nitrogen per unit carbon
[00:25:17.350]because it has gone through microbial transformation,
[00:25:21.050]but then that microbial necromass
[00:25:23.430]is mostly stabilized through mineral association.
[00:25:27.080]And so it's not any more accessible
[00:25:30.980]for farther decomposition
[00:25:33.280]because it's protected by a bunch of different mechanisms.
[00:25:37.420]And so in this case, it can also persist for century
[00:25:41.230]to millennia, but it's not more vulnerable
[00:25:44.410]that is not any more very vulnerable to disturbance.
[00:25:49.350]Of course, all the carbon in soil
[00:25:51.720]is vulnerable to some extent.
[00:25:54.670]And then in between, you have all those soils
[00:25:57.610]that have an intermediate share of POM and MOAM
[00:26:00.430]and intermediate vulnerability and carbon stocks,
[00:26:03.470]where some of the soil is still from plant material
[00:26:08.430]that still has to undergone decomposition,
[00:26:11.150]some is stabilized.
[00:26:12.680]And where the microbe are not fully inhibited,
[00:26:15.017]but really the decomposition is controlled by the factors
[00:26:19.890]that limit microbial activity like energy availability,
[00:26:23.830]nutrient availability, moisture, and so forth.
[00:26:26.900]And so applying this key really can help us
[00:26:29.820]understanding better also where our soils are,
[00:26:33.290]what potential, what do we need to do to accrue more carbon,
[00:26:37.070]what are the mechanism maintaining there?
[00:26:41.555]And along this line of thoughts,
[00:26:44.100]I found it very useful again, to think about
[00:26:47.870]soil organic matter has this share
[00:26:50.470]between particulate and mineral associated organic matter.
[00:26:54.100]And what allowed me to do that is the fact as I said,
[00:26:57.510]as I show you before in those other fractionation scheme,
[00:27:00.800]is that those two contrasting form of carbon
[00:27:05.500]for so many different properties
[00:27:08.050]actually have different size and density
[00:27:12.620]and therefore we can't separate them.
[00:27:15.380]It's not as straightforward
[00:27:17.360]because there are a lot of other complexity
[00:27:20.240]and other forms and they can be formed in aggregate,
[00:27:23.520]but overall we can separate them
[00:27:26.330]and do further analysis on them.
[00:27:28.200]And so being measurable really help us studying
[00:27:32.720]and proposing them as proxy for understanding
[00:27:36.390]the state of your soil and its health.
[00:27:40.830]But one problem is that measuring POM and MAOM
[00:27:45.240]is somewhat tedious and can be time consuming.
[00:27:49.290]And so one of the other things my lab, we are exploring,
[00:27:54.350]in particularly knowing that we are going to this
[00:27:58.130]large scale need for quantifying soil organic matter,
[00:28:03.210]is to increase to throughput.
[00:28:04.870]And the way which we plan to increase throughput,
[00:28:10.080]On one hand, I have an NSF project
[00:28:17.670]where we collaborate with some engineering
[00:28:22.320]to actually build an instrument that can automatize
[00:28:26.330]the density flotation
[00:28:27.730]which is one of the most tedious approaches.
[00:28:31.370]The other thing we are doing,
[00:28:33.470]is to work with mid infrared spectroscopy
[00:28:36.980]to create some reliable calibration,
[00:28:40.420]so that let's say we take a thousand of soil sample.
[00:28:44.660]If we, and if we fractionate maybe 400 of them
[00:28:49.280]and scan the rest, we can create calibration
[00:28:52.810]that predict POM and MOAM
[00:28:54.850]and that seems to be a pretty good approach.
[00:28:58.420]We're doing a lot of work on that,
[00:28:59.847]and I feel very strong about getting in that direction.
[00:29:06.000]And another thing we have actually
[00:29:07.720]have one NSF project proposal to go be submitted soon
[00:29:14.840]is to use artificial intelligence,
[00:29:18.580]leveraging the large data sets that are there
[00:29:22.180]to upscale those measurements.
[00:29:26.670]And that's the approach that, for example,
[00:29:29.830]I used in Europe where I went on sabbatical 2015
[00:29:34.390]at the JRC, the Joint Research Center,
[00:29:37.880]and they had this amazing data set that they called LUCAS.
[00:29:42.960]There was a combined with physical soils
[00:29:46.390]and they had like 20,000 soils.
[00:29:49.240]So what we did was to subset and obtain 400 samples
[00:29:54.530]that were well representative of the 20,000 that they had.
[00:29:59.100]And then we fractionated them by POM and MOAM
[00:30:04.010]and use a random forest approach
[00:30:06.650]to actually predict POM and MOAM
[00:30:09.360]at the continental European scale.
[00:30:12.270]And you can see, as I was showing you before
[00:30:16.300]that the hypothesis is confirmed,
[00:30:18.473]that when there is more carbon and soil,
[00:30:21.430]the POM is where the majority of the carbon is present.
[00:30:25.270]And instead MOAM really accumulate in the soils
[00:30:28.770]that have the least amount of carbon.
[00:30:31.440]And when you start doing this separation
[00:30:35.940]by the different land users,
[00:30:37.900]you start seeing that arable soils
[00:30:41.570]really are very different from natural soils
[00:30:44.590]in the sense that they have lost the majority of the POM
[00:30:49.570]because it is the most vulnerable fraction of soil
[00:30:53.450]and also with tillage, we really have removed
[00:30:57.130]any microbial limitation and also reduce the input.
[00:31:01.070]And so the majority of the carbon in arable is in MAOM.
[00:31:06.210]But because arable land in Europe at least
[00:31:10.000]is the depth of this column are the area, the extensions,
[00:31:17.190]once you multiply the stocks by the cover in Europe,
[00:31:21.960]becomes that arable land actually make the majority
[00:31:25.930]of the carbon stock in European soils.
[00:31:29.200]And therefore, there is a high potential there
[00:31:32.710]for further accrual.
[00:31:34.330]On the other side, corn, for example,
[00:31:36.820]but even grassland, much more balanced POM and MOAM
[00:31:39.350]that can really inspire us
[00:31:41.930]when we think about soil health, what does soil health mean?
[00:31:45.190]In my mind, it really mean to bring arable soil,
[00:31:49.080]to have both a carbon to nitrogen and a fraction partition
[00:31:53.820]that is more similar to grassland.
[00:31:57.540]And so, putting all this concept together,
[00:32:00.720]I'm now working a lot in thinking how we can use
[00:32:06.420]the POM and MOAM concept to understand
[00:32:11.051]what controls carbon and nitrogen cycling in soils
[00:32:15.640]and how we can maintain those balance systems,
[00:32:19.750]where you have an equilibrium POM and MOAM,
[00:32:22.610]you have an efficient nitrogen recycling
[00:32:25.500]and therefore maintain productivity.
[00:32:30.095]And I believe that to do that,
[00:32:31.890]we really need to think about what control input limitation
[00:32:37.130]and what control output limitation.
[00:32:40.070]When input limitation I consider, plan input.
[00:32:44.880]And so, for example,
[00:32:46.720]you can think of cold wet natural system
[00:32:50.410]that have very little input.
[00:32:52.350]And as I was saying in permafrost or in wetlands,
[00:32:55.560]you accumulate POM.
[00:32:57.740]Our agricultural soils now are mostly here,
[00:33:01.890]where they're input limited
[00:33:03.970]because we don't put that much residues in our soil.
[00:33:08.440]We only have cover for a short period of the year
[00:33:11.930]and therefore the microbe basically go quickly
[00:33:16.050]through the little input,
[00:33:17.980]transform them all in MOAM and the nitrogen,
[00:33:21.920]there isn't enough plans to pick up the nitrogen.
[00:33:24.490]And often you might have a lot of nitrogen losses
[00:33:27.790]to the environment, of course,
[00:33:29.160]also because we over fertilize them.
[00:33:32.140]Anyway, this is to give you or this is the most,
[00:33:37.859]the framework that currently my lab is working on
[00:33:41.390]and trying to see whether it's a good framework
[00:33:44.290]to ask question both in natural system
[00:33:47.460]and in agricultural systems.
[00:33:49.830]But as I presented you, what we do in measurements,
[00:33:53.430]then what we do and what I think is very helpful
[00:33:57.080]for informing regenerative agriculture is to transfer
[00:34:01.840]this information into models and really create
[00:34:06.040]what I mentioned is coherent measure modeling framework.
[00:34:12.630]And so a little bit for branding.
[00:34:14.629]We've also called our model, the MEMS model.
[00:34:19.010]And we got two or the MEMS model
[00:34:22.220]starting from first modeling
[00:34:24.760]the litter layer when Neil Campbell
[00:34:26.760]was a PhD student here at CSU,
[00:34:29.890]we recreated those experiments I showed you.
[00:34:33.740]And to separate the input of soluble
[00:34:36.970]and structurally into the soil.
[00:34:39.403]Then I had Andy Robertson as postdoc
[00:34:42.500]now is working at Shell and he build the,
[00:34:46.490]we build together the MEMS 0.1
[00:34:49.782]that expanded from the litter layer,
[00:34:54.200]created the soil structure, again, in particular,
[00:34:57.920]the mineral associated organic matter,
[00:35:01.300]but the basically only represented carbon.
[00:35:04.720]We didn't have water or nitrogen in this model.
[00:35:08.590]It actually worked well.
[00:35:09.820]We used it to represent the data sets I showed you
[00:35:13.150]about grasslands and forest in Europe.
[00:35:17.200]And overall, we had a pretty good correspondence
[00:35:21.070]between measure MEM model,
[00:35:23.380]as you can see from the map of the residues here.
[00:35:27.810]But we then wanted to move farther
[00:35:32.670]and create a true ecosystem model
[00:35:35.540]that represents also nitrogen water multiple soil layers,
[00:35:41.060]separate rhizosphere processing from the box soil processing
[00:35:45.540]and the litter layer from the soil.
[00:35:47.830]And that is the work that really Yao Zhang is leading here
[00:35:52.420]at CSU together with me.
[00:35:54.350]And we are now doing lot of work
[00:36:00.398]now to further develop MEMS particularly to represent
[00:36:04.680]regenerative croplands and grasslands.
[00:36:07.780]Our first publication demonstrated the ability
[00:36:11.591]of the MEMS tool to represent grassland sites.
[00:36:14.970]And we use data that Rebecca, even in my lab,
[00:36:17.910]fractionated from the neon grassland sites,
[00:36:22.350]and overall, we use four of the sites for calibration
[00:36:25.583]and two for validation.
[00:36:27.650]And of course it wasn't perfect, but overall,
[00:36:30.230]we had some good agreement
[00:36:32.280]both for the share of POM and MOAM
[00:36:34.130]as well as for the depth dynamic.
[00:36:37.270]Of course, MEMS 2 also represent flexes
[00:36:41.284]and so in, it can be very useful to be calibrated
[00:36:48.630]and validated across class ecosystem flexes
[00:36:53.265]taken (indistinct) for example
[00:36:56.070]So this is, this model was really built
[00:37:01.553]to connect the measurements with the model so that we can do
[00:37:05.860]a better job calibration and validation,
[00:37:09.650]but also at using the model has an inquiry tool
[00:37:15.540]for research as well as pair it with field work
[00:37:20.750]and field study to better understand
[00:37:23.040]what's going on and possibly project into the future.
[00:37:27.330]And I'll show you now a few applications,
[00:37:34.650]because again, what I think what I said so far
[00:37:40.100]can be really helpful to inform management solution.
[00:37:45.180]And I like to apply the framework that I just presented you
[00:37:50.830]to regenerative agriculture.
[00:37:52.870]I think that framework really help us trying to understand
[00:37:58.610]in under which soil condition and what kind of management
[00:38:03.260]would best promote one pool versus the other.
[00:38:09.580]And so, for example, increasing root structures
[00:38:13.940]with perennial, so with cover crops,
[00:38:16.880]we might have more particulate organic matter
[00:38:19.710]possibly more nitrogen input, organic nitrogen input
[00:38:24.640]through manure for example,
[00:38:27.450]legumes might simulate more microbial
[00:38:29.667]to and over a mineral associated organic matter.
[00:38:32.900]But overall, to address the carbon dilemma,
[00:38:36.380]we really want to increase the inputs
[00:38:39.890]so that they can maintain the tool
[00:38:42.290]and over microbial activity and health
[00:38:44.460]so that they can provide fertility
[00:38:46.460]to support plant growth.
[00:38:48.220]But at the same time, have some saving
[00:38:51.090]that can be stored in mineral associated organic matter
[00:38:55.110]and that's possible if the micro turnover efficiently.
[00:38:59.840]And so now a lot of my students
[00:39:03.090]are working on those questions,
[00:39:05.360]in particularly Iron Prairie
[00:39:08.160]is almost completing a meta-analysis
[00:39:10.620]where is looking at how different management practices
[00:39:15.860]differently affect POM and MOAM.
[00:39:17.500]These are only preliminary data,
[00:39:20.820]but it seems like as we would've expected that POM and MOAM
[00:39:25.890]respond differently to the management in particularly
[00:39:29.650]with the POM being by far more responsive.
[00:39:33.180]One aspect that we think are important is in particularly
[00:39:36.860]with perennial and cover crop,
[00:39:38.950]we see the highest POC in accrual,
[00:39:42.050]and that can really also be effective at that
[00:39:45.740]when we think about perennialization.
[00:39:48.630]Have one student of my lawyer Van der Pol
[00:39:51.490]that is working on how to incentivize, for example,
[00:39:56.100]perennial use this study was done
[00:39:59.879]with Kernza perennial wheat grass
[00:40:03.960]on several farms in Kansas.
[00:40:06.700]And we saw a significant increase
[00:40:09.290]of particulate carbon with the perennial wheat grass.
[00:40:15.270]Laura also worked at what I started from that,
[00:40:19.590]that soil carbon dilemma.
[00:40:21.590]And again, the idea is that here in wheat follow,
[00:40:26.540]you really have those limited input system.
[00:40:29.890]Of course, if you intensify the rotation,
[00:40:33.350]you can put more carbon in the soil,
[00:40:36.170]but you are gonna put the same type of carbon
[00:40:39.763]and you might just lock it in
[00:40:42.750]and not stimulate that turnover and fertility of the system.
[00:40:47.810]And so the idea is that to really address
[00:40:50.160]the solar carbon dilemma and sequester carbon
[00:40:52.960]and the same time as maintained productivity,
[00:40:55.940]is by diversifying the crop rotation in particular,
[00:41:00.560]including a legume.
[00:41:02.433]And Laura addressed that question in farms in Colorado
[00:41:07.220]and Nebraska, that had done either continuous
[00:41:15.212]with fallow included, a continuous rotation
[00:41:19.110]without grain rotation, without legume
[00:41:22.220]or a continuous grade rotation with legume.
[00:41:25.320]And she observed that across the four year,
[00:41:28.330]the continuous rotation had increased productivity,
[00:41:32.630]but only where the legume was added,
[00:41:35.590]she saw an increase in organic matter,
[00:41:38.440]and in particularly an increase
[00:41:40.640]in the mineral associated organic matter
[00:41:44.800]that we explained with the legume stimulating
[00:41:47.700]microbial transformation and providing that nitrogen
[00:41:50.710]that is needed to form the mass.
[00:41:53.110]But this was of course a shot in time.
[00:41:55.560]The farm had done that for about 10 years
[00:41:58.650]and it, we don't really know what was the trajectory
[00:42:02.010]of the carbon changes and how it would go into the future.
[00:42:05.750]And so that's where that real combination
[00:42:08.410]of measurement modeling is helpful
[00:42:10.800]because what Laura did, was to use a data set
[00:42:14.810]with the MEMS and first verify that
[00:42:18.920]she had a good prediction across all the fractions.
[00:42:22.290]And then she used the MEMS to forecast into the future.
[00:42:26.150]And what she found out was that first of all,
[00:42:28.920]the farm, the width fallow had lost a lot of carbon
[00:42:32.400]from the initial Prairie system
[00:42:35.690]and it would continue losing carbon if nothing happened.
[00:42:39.350]But if you, with the continuous grain,
[00:42:43.820]you would stop the loss,
[00:42:45.400]but it would really be the continuous grain with legume
[00:42:49.510]which would allow that addition of carbon,
[00:42:53.480]stabilized carbon in the map.
[00:42:56.640]The bad news is that these results
[00:42:59.460]cannot really be extended everywhere
[00:43:02.270]and very much depend on the system where you are.
[00:43:05.650]This work was done on again,
[00:43:08.670]dry system in Colorado and Nebraska,
[00:43:12.590]but this other study that Rui et al had done,
[00:43:17.280]instead in Mollisols system,
[00:43:19.820]they observed it after very long time,
[00:43:22.670]even the system where they include the legume
[00:43:26.320]didn't really change the needle much,
[00:43:28.880]but it was when they converted it back to grazed pasture
[00:43:35.250]that the MAOM significantly increase.
[00:43:38.070]And so, again, regenerative, we have the principle,
[00:43:41.360]I think we can understand them,
[00:43:42.940]but we also need to think that the responses most likely
[00:43:48.090]are gonna be site specific.
[00:43:50.060]And so we need to work with farmers to understand what works
[00:43:53.850]in their sites.
[00:43:55.980]But I want to finish by with this grazed pasture,
[00:44:00.620]because that's actually one of the plays with the majority
[00:44:03.400]of my work is currently happening
[00:44:07.120]where we are really looking at how grasslands
[00:44:10.910]can be improved to increase carbon stocks.
[00:44:14.330]There has been a lot of work and previous synthesis
[00:44:18.390]suggesting that improved management
[00:44:20.700]can move the needle in the (indistinct) soils,
[00:44:24.130]and one of first studies that we did
[00:44:27.130]in the southeast of the United States
[00:44:29.280]with Sam Mosier was my PhD student at the time,
[00:44:32.610]she saw that by implementing adaptive management
[00:44:38.690]with rotational grazing, you again,
[00:44:41.440]could move the needle increased carbon,
[00:44:43.650]and in particular increased carbon
[00:44:45.411]in the mineral associated organic matter.
[00:44:48.290]So I would say that regenerative management
[00:44:51.710]have a lot of potential.
[00:44:54.010]However, we do need to be careful the moment in which
[00:44:57.590]this carbon sequestration is immediately transferred
[00:45:01.090]to payment for carbon markets.
[00:45:04.060]And while I think it's good to stimulate to provide
[00:45:09.060]a revenue for farmer, to transform their system
[00:45:13.190]and be more regenerative,
[00:45:14.960]I think that is important that we have some specific rules.
[00:45:18.840]And that's what we try to do, for example,
[00:45:22.590]in this opinion piece,
[00:45:24.610]but that's mostly what we are working on at Colorado states.
[00:45:29.460]We have created this Carbon Solution Center
[00:45:32.710]where we want to bridge our scientific understanding
[00:45:35.920]and bring it to producer and make sure that
[00:45:39.540]that all of this is done in a way that help our planet,
[00:45:43.193]help our farmer and all of us together.
[00:45:46.220]And to achieve those objective,
[00:45:49.240]I believe that education is really at the basis.
[00:45:52.750]And so I also do a lot of school and public outreach
[00:45:56.960]where now we are interested to hear from farmer
[00:46:01.070]knowing their issues and work with them,
[00:46:03.920]but also try to stimulate earlier on in the citizen,
[00:46:09.510]the appreciation for soils and what they can do.
[00:46:12.487]And to be honest, it's also very fun and rewarding.
[00:46:15.930]And so with that, I saw that I use quite a bit of my time.
[00:46:19.350]I will now, you can read the conclusions,
[00:46:25.010]but I want to say that all the work I presented,
[00:46:28.770]it's really the work of a lot of people.
[00:46:32.890]I have the fortune to have an amazing group of people
[00:46:35.810]that work with me and collaborators
[00:46:38.160]as well has a significant source of funding
[00:46:41.010]from all the federal agency and also private companies.
[00:46:45.220]And with that, I like to thank you.
[00:46:47.240]And I'm happy to take your question
[00:46:50.330]and I'll stop sharing so I can see you.
[00:46:57.500]Thank you, Dr. Francesca.
[00:46:59.810]Appreciate your seminar.
[00:47:01.870]So I have one first question.
[00:47:04.630]So in, I think in the 10 beginning minutes of your seminar,
[00:47:09.050]you mentioned the contribution of root structures.
[00:47:13.310]And I would like to say if you saw some contribution
[00:47:19.340]in terms of comparison of the root and above ground part
[00:47:23.970]of the cover crops.
[00:47:25.920]Yeah, so actually I didn't have the time
[00:47:28.980]to show this results.
[00:47:30.620]We have done some, a couple of studies
[00:47:34.270]both in the lab and the field where we convert
[00:47:36.900]the aboveground and the root contribution.
[00:47:41.140]In particular, this work was done with the sorghum
[00:47:43.827]and there is a publication food transmitted
[00:47:47.901](indistinct) biology 2019
[00:47:52.800]I don't remember the (indistinct) the year
[00:47:55.520]but what we saw typically is that the roots
[00:47:59.940]produce more carbon,
[00:48:01.920]mostly because the composition inside the soil is lower.
[00:48:05.970]You have more particulate organic matter formation
[00:48:09.110]and also possibly more efficient mineral associated
[00:48:13.130]organic matter formation.
[00:48:15.870]The aboveground, you might have more,
[00:48:20.950]first of all the decomposition above ground,
[00:48:23.760]it's very much dependent on the climate.
[00:48:26.200]And so if you have irrigated system and stuff like that,
[00:48:29.550]it moves very quickly.
[00:48:31.580]And a lot of it, more of it is lost to CO2
[00:48:35.270]than moves down into the soil.
[00:48:37.440]So there are two things is the chemical make up
[00:48:40.400]but is also actually the place
[00:48:43.160]where the transformation happen.
[00:48:45.600]And the fact that you have the root
[00:48:50.200]entering the soil already, being already in the soil,
[00:48:55.010]that makes it more efficient.
[00:48:57.050]The fact that typically they have suberin
[00:49:00.293]and are more rich in recant structure,
[00:49:04.600]make them for more particulate organic matter
[00:49:09.350]and roots that if they are deep roots,
[00:49:13.270]then when you have the most efficient formation.
[00:49:16.490]But typically of particulate organic matter,
[00:49:19.820]where the roots inputs are very effective reforming MOAM
[00:49:23.700]is from the exudates.
[00:49:27.017]So the roots form those different things.
[00:49:30.320]The more I work on this, the more I think that
[00:49:34.100]below two, five centimeter, it's almost entirely roots.
[00:49:40.670]So if we want to breed for carbon,
[00:49:44.253]we need to breed for roots.
[00:49:46.420]And that's one thing we are doing quite a bit.
[00:49:51.620]We have a question from Dr. Lidonc.
[00:49:56.780]So he said, hi, Dr. Ctrufo,
[00:49:59.040]what do you think could be the potential and the limitation
[00:50:03.390]of modeling soil organic carbon
[00:50:06.260]using artificial intelligence techniques
[00:50:09.190]such as random forest?
[00:50:15.090]So for sure, with the data intensity that we are gonna have
[00:50:21.950]in the very near future,
[00:50:23.880]we are gonna be overloaded by data
[00:50:26.540]and artificial intelligence will,
[00:50:29.960]can really help us further scaling the data
[00:50:34.970]that we have.
[00:50:35.803]So let's say that we have,
[00:50:40.040]so I'm gonna say that artificial intelligence
[00:50:44.010]is gonna be very helpful and very effective at predicting
[00:50:49.620]what we measure at even lighter scale.
[00:50:53.554]However, if we don't understand the mechanism,
[00:51:00.740]it will be hard to predict the responses
[00:51:03.460]both to management and disturbance and climate change.
[00:51:06.930]And so that's where artificial intelligence
[00:51:11.290]really fails a little bit.
[00:51:13.690]So I think artificial intelligence can be very helpful,
[00:51:17.300]especially to improve our data,
[00:51:21.350]especially it will be hard to use it for forecasting
[00:51:28.150]because we will have to assume that the condition stays
[00:51:31.510]as they are today.
[00:51:32.720]There are no mechanism in there to predict changes,
[00:51:36.660]if that makes sense.
[00:51:38.010]And so I think that the future is a combination
[00:51:42.820]of empirical modeling and artificial intelligence.
[00:51:47.870]Yeah, I think we don't have any more questions.
[00:51:51.780]I wanna say, thank you so much for your presentation.
[00:51:55.060]We really appreciate that, yeah.
[00:51:57.770]Thank you, was really a pleasure.
[00:52:00.214]Yeah, thank you.
Log in to post comments