Albedo of crops as a nature-based climate solution to global warming

Surface albedo can affect the energy budget and subsequently cause localized warming or cooling of the climate. When we convert a substantial portion of lands to agriculture, land surface properties are consequently altered, including albedo. Through crop selection and management, one can increase crop albedo to obtain higher levels of localized cooling effects to mitigate global warming. Still, there is little understanding about how distinctive features of a cropping system may be responsible for elevated albedo and consequently for the cooling potential of cultivated lands. To address this pressing issue, we conducted seasonal measurements of surface reflectivity during five growing seasons on annual crops of corn-soybean–winter wheat ( Zea mays L.- Glycine max L. Merrill —Triticum aestivum L. ; CSW) rotations at three agronomic intensities, a monoculture of perennial switchgrass, and perennial polycultures of early successional and restored prairie grasslands. We found that crop-species, agronomic intensity, seasonality, and plant phenology had significant effects on albedo. The mean ± SD of albedo was highest in perennial crops of switchgrass ( Panicum virgatum ; 0.179 ± 0.04), intermediate in early successional crops (0.170 ± 0.04), and lowest in a reduced input corn systems with cover crops (0.154 ± 0.02). The s trongest cooling potentials were found in soybean ( − 0.450 kg CO 2 e m − 2 yr − 1 ) and switchgrass ( − 0.367 kg CO 2 e m − 2 yr − 1 ), with up to − 0.265 kg CO 2 e m − 2 yr − 1 of localized climate cooling annually provided by different agroecosystems. We also demonstrated how diverse ecosystems, leaf canopy, and agronomic practices can affect surface reflectivity and provide another potential nature-based solution for reducing global warming at localized scales.


Introduction
Land surface albedo is the portion of incoming solar radiation from the sun that is reflected by the land surface.The monitoring of albedo is critical in managed landscapes (e.g.croplands, grasslands) as it constitutes an opportunity to increase surface reflectivity, reduce warming via crop selection, agricultural management, and the potential for radiative cooling (Muñoz et al 2010, Bright et al 2012, Carrer et al 2018, Zhu et al 2024).Land use and land cover is dependent of land management through intentional land cover design (e.g.conversions from bare ground to biofuel, grassland, and cropland) and agronomic practices (e.g.tillage, planting density, crop species choice, conventional vs conservation management), which in turn can significantly alter surface albedo (Pielke et al 2011, Cai et al 2016, Lei et al 2021, 2023, Sieber et al 2022a) that redefines surface energy budget (Sieber et al 2020, Abraha et al 2021).These imbalances due to surface albedo changes are referred to as albedoinduced radiative forcing (RF α ).The RF from albedo difference is the amount of energy reflected for assessing the warming/cooling potentials of different surfaces (Lei 2022, Lei et al 2023).However, to gauge the cooling potential of agricultural surfaces, the analysis of the variability and dynamics of albedo across unique landscapes needs to be performed at multiple temporal scales and at different crop types.
Changes in albedo due to land use can have quantifiable impacts on the amount of heat an ecosystem can reflect (Georgescu et al 2011, Caiazzo et al 2014, Carrer et al 2018).As a result, albedo, through solar radiation management, has been proposed by the Intergovernmental Panel on Climate Change (IPCC) for maximizing climate cooling on the landscape (Lenton andVaughan 2009, Chen et al 2021).Still, current scientific understanding of forcing effects of albedo is relatively low (IPCC 2022, Sciusco et al 2022).Previous studies have attempted to bridge this knowledge gap using satellite data (Fang et al 2007, Zhang et al 2010, Sciusco et al 2020, 2022), where spatial measurements are effective but discontinuous temporal measurements may miss key crop phenological stages; or by using biophysical models (Smith et al 2020), where albedo is often estimated by generic values not representative of individual crops on the land surface.In contrast, field-based observations can be employed to more accurately observe changes in phenology, document unique agronomic managements over the course of a growing season, and identify crop-specific surface reflectivity benefits not easily discerned from satellite pixels.Here, we use field measurements of surface albedo to more intensively investigate albedo differences among unique cropping and grassland ecosystems and associated agronomic practices.
We focus on three aspects of agroecosystems that influence albedo: (1) crop composition, (2) seasonality, and (3) management practices and agronomic inputs.First, plant characteristics, including crop type (e.g. annual vs perennial), species (e.g. corn vs wheat), varieties (e.g.different season lengths), and phenology can vary substantially, and can potentially account for large differences in surface reflectivity (Hollinger et al 2010).The relationship between plant canopy, chlorophyll content, and surface reflectance has also been acknowledged in past research (Bray 1962).For example, Daughtry et al (2000) observed that remote sensing observations of canopy reflectance of corn was influenced by leaf area and leaf chlorophyll.The influence of these factors is also connected to phenological events, such as flowering, leafing, and leaf fall, which are also correlated with measures of surface reflectance (Stöckli and Vidale 2004).However, this canopy-chlorophyll-albedo relationship has only been explored in forests (Ollinger et al 2008, Bartlett et al 2011) but not empirically examined in grasslands and croplands that cover a vast majority of the landscape in the United States Midwest, and other global regions.
Second, changes in planting dates, weed management, harvest dates, residue retention, and other agronomic factors can lead to large differences in albedo over the course of a growing season.In a previous study, summer months were found to be particularly important for producing albedo cooling (Lei et al 2023), suggesting that a deeper understanding of seasonal agricultural patterns would be essential to improving Nature-based climate solutions (NbCS).
Third, agronomic management can be a crucial determinant of albedo level and dynamics.Albedo is influenced by farmer's choices of crop, timing of planting and harvesting, and management practices including tillage, weed control, fertility management, and cover crops, each of which can lead to significant variations in albedo among different agricultural systems (Davin et al 2014, Miller et al 2016).For example, Stock et al (2019) showed that winter application of manure can affect radiation absorption, but its absorption depends on the timing of application and seasonal snowfall.Clearly, there is a need to further understand the cumulative effects of farmer practices over the growing season.
Here, we investigate how cropland characteristics can change over time and affect surface albedo for quantifying the magnitude of the cooling effects in ten agroecosystems typical of the upper Midwest USA.We hypothesize that perennial crops will have a higher time-integrated albedo compared to annual crops, primarily because they maintain a well-developed vegetative cover for a much longer growing period annually.We also examine the relationship between plant nitrogen and surface reflectivity, as plant nitrogen content can be directly affected by nitrogen fertilization and agronomic practices.We predict that higher plant nitrogen concentrations will increase surface reflectance.Understanding these dynamics will further inform the potential of employing agronomical systems as a NbCS to global warming (Robertson et al 2022).

Study design
Our study was conducted at the W.K. Kellogg Biological Station (KBS) Long-term Ecological Research (LTER) Main Cropping System Experiment (MCSE, https://lter.kbs.msu.edu/)(figure 1).The average air temperature at KBS is 9.9 • C, with an average of −7 • C-28 • C over the course of the year.Precipitation averages around 1005 mm annually, with around 17% of precipitation occurring in winter, and the remaining 83% evenly distributed across spring, summer, and fall (NCDC 2013, Robertson andHamilton 2015).The landscape is a mature glacial outwash plain and moraine complex, consisting of fine-loamy to coarse-loamy mesic soils (Thoen 1990, Crum andCollins 1995).The MCSE consists of an area 1.4 km long and 0.40 km wide that were divided into a randomized complete block experiment with 1-hectare plots in each of six replicate blocks.We used six treatments in three blocks for this study: three annual cropping systems and three perennial systems.The annual cropping systems are corn-soybean-winter wheat rotations (Zea mays L.-Glycine max L. Merrill-Triticum aestivum L.; CSW) managed differently as described below.The perennial systems include switchgrass (Panicum virgatum L.), an early successional community burned annually, and restored prairie installed as a prairie strip (Kemmerling et al 2022) in one of the annual cropping systems.Perennial systems green up once the snow and frost thaws on the landscape in early spring.The CSW systems were wheat in 2019, followed by corn in 2020, then soybean in 2021, wheat again in 2022, and corn again in 2023.Winter wheat is planted in the fall and harvested in mid-July.Corn and soybean are planted in May and harvested around mid-October.

Assignment of treatment and control conditions
2.2.1.Three annual cropping systems were used in this study, as follows CSW-Conventional (CSW-CO): A conventional corn-soybean-winter wheat rotation, where crops received fertilizer and pesticide inputs typical of those in the U.S. Corn Belt and conventional tillage (chisel plowing) to control weeds and incorporate crop residue into the soil surface.The system is not planted to a winter cover crop other than wheat during the winter wheat portion of the rotation.
CSW-No Till (CSW-NT): This system is managed identically to the CSW-CO system but for tillage.This system has been continuously no-tilled since 1989.Due to the COVID-19 pandemic, this system was sampled only in 2023 and only in one replicate plot.
CSW-Reduced Input (CSW-RI): This system is managed similarly to the CSW-CO system but with reduced chemical inputs, mechanical weed control, and cover crops.In this system herbicides are banded only within rows, resulting in 30% spatial coverage, and mechanical cultivation controls weeds between rows.The system also receives 30% of the nitrogen fertilizer applied to CSW-CO, with other nitrogen inputs from cover crops.After wheat harvest, a winter cover crop of red clover (Trifolium pratense L.) is planted, while fall-planted wheat provides cover after soybean.Additional details on long-term management practices for each system within this study can be found in Robertson and Hamilton (2015).

Three perennial systems were also used in this study, including
Switchgrass: A switchgrass monoculture grassland was established in 2019, where the previous system was alfalfa from 1989 to 2019.Switchgrass was fertilized annually with 28% urea ammonium nitrate (UAN) at 50 lb.N/ha.
Early Successional: A polyculture early successional community derived from the seed bank and unmanaged but for annual spring burning.Dominant species included Canada goldenrod (Solidago canadensis L.), arrow-leaved aster (Aster sagittifolius.), red clover (Trifolium.pretense L.), and smooth bromegrass (Bromus inermis Leyss).This system was converted from row-crop agriculture with tillage in 1989 and is burned every spring to control the establishment of woody species.
Restored Prairie: In 2019, 30 m wide prairie strips were added to CSW-Reduced plots; these were grown annually and burned the following spring after harvest of the main treatment.This perennial polyculture grassland strip was planted with dominant species of big bluestem (Andropogon gerardii; C4), indiangrass (Sorghastrum nutans; C4) and prairie junegrass (Koeleria cristata; C3) native to Michigan and commonly found throughout the Midwest (table 1).

Data collection and instrumentation
From 2019 to 2023, field data were collected at the thirteen MCSE sites every two weeks during the growing season (May 1st through October 31st) in order to capture the changes in vegetation and soil properties.Measurements were taken across thirteen sites at 39 semi-permanent sampling stations of No. 3, 4 and 5 to minimize trampling of crops (figure 1), yielding a total of 317 measurements each year (see S table 1 for a detailed summary of measurements per crop).
We used a portable four-component net radiometer (Kipp & Zonen, Netherlands) mounted on a survey pole, where the heights of measurements were changed according to the current plant height in order to maintain optimal sensor field of view to the vegetation (1.34 * average plant height) (Zeri et al 2011, Lei et al 2023).Shortwave incoming (SW ↓ ) and outgoing (SW ↑ ) radiation were recorded at one-second intervals and converted to per-minute averages (equation ( 1)).Each plot was sampled for 1-2 min to ensure stable conditions.All observations were logged using a Campbell 1000X datalogger (Campbell Scientific Inc., Logan, UT, USA).All measurements within each site were taken approximately between 10:00 and 13:30 h local time to ensure that data measurements were consistent around solar noon.
To investigate how plant 'greenness' affects surface reflectivity we measured nitrogen content (chlorophyll) within crops at all blocks and treatments using a Soil Plant Analysis Development chlorophyll meter (SPAD 502, Spectrum Technologies, Inc. IL, USA).During measurement, a crop representing species majority within the treatment plot was selected (in the case of polyculture perennials sites), to ensure the chlorophyll measurement retrieved was representative of dominant crops.Three measurements were taken per leaf, where the leaf was kept under shade to avoid color variance caused by solar angle and sunlight intensity.A linear regression was then completed between albedo and chlorophyll content, to analyze the relationship between foliage nitrogen and potential reflected radiation.

Variation between adjacent plots
Each plot was measured at three replicated sites to assess the robustness of data measurements observed and any variation between crop species and land use.Each replicate block had the exact crop species and agronomic management, and all were located within close proximity to ensure similar climate and growing conditions.There were rare occasions when measurements were missed because personnel were not able to access sampling sites, due to removal of sampling flags within plots or conflicts related to agronomic inputs (e.g.fertilization, social distancing).

Statistical analysis
Data were analyzed in the statistical software R (R Development Team 2013) and Esri ArcGIS v.10.2.2 (ESRI 2011).ArcGIS was used to map albedo ( α s ), which was calculated as: where SW ↓ is the incident shortwave radiation and SW ↑ is the reflected shortwave radiation.Albedo was averaged over the year to assess surface reflectivity across treatments and years.Seasonal changes and agronomic management were then examined against albedo to detect recurrent patterns between crop species and albedo.A Kruskal-Wallis non-parametric test was also used to examine the change in albedo within the five-year study period, where the dependent variables were surface albedo and chlorophyll nitrogen content.Mean albedo by year and site was compared using Tukey's HSD test with treatment effects considered significant at p < 0.05.Finally, a generalized linear mixed effects model was developed to investigate land use effects on albedo, where year was treated as a random factor and crop species treated as fixed effects.
Change in albedo (∆ α ) in this study was determined by the calculating the albedo of a specific crop less than albedo from a reference crop: where ∆ α is the local change in albedo, α agroecosysten is the agroecosystem albedo, and α reference is the reference bare ground albedo.Annual RF from albedo change was then determined using bare ground as the reference ecosystem, as it represented the majority of land use in the Midwest after the winter season, as shown in (3): where RF α is the change in net radiative flux from the surface driven by surface albedo (W m −2 ) at the top of the atmosphere, Sw ↓ is local incoming solar radiation incident to the surface (W m −2 ), T k is the upwelling transmittance derived from estimating thermal radiant fluxes within the environment (Campbell and Norman 2012), N is the number of days, and ∆ α is the albedo difference between our agroecosystem and Climate impacts from changes in RF were also assessed using GWI within life cycle analyses for the derivation of climate warming or cooling potentials.For a period of one hundred years (GWI: kg CO 2 e m −2 yr −1 ), the impact of albedo-induced RF is calculated as shown in (4): where, RF α is the RF from changes in albedo TOA in equation ( 2) (W m −2 ), S is the local area subjected to albedo change (normalized to 1 m 2 ), AF is percentage of human-emitted CO 2 that remains in the atmosphere after a period of time from anthropogenic sources, rf CO2 is the derived RF from 1 kg of CO 2 , and TH is the time horizon for 100 years (TH = 100).This enabled us to examine the importance of albedo relative to its climate impact on different landscapes, relative to common farming practices and species effectiveness.Negative values of GWI α indicate a CO 2 e mitigation impact, i.e. a localized cooling effect.

Temporal changes in albedo
Albedo of the sampling plots ranged from 0.11 to 0.27 during the growing season.Within our study sites CO-Soybean was observed to have the highest average surface reflectivity (0.186 ± 0.03), followed by RI-Soybean (0.182 ± 0.03), and then Switchgrass (0.179 ± 0.04) (figure 2).RI-Corn had the lowest average albedo (0.154 ± 0.02) of the systems measured.Within the CSW rotations, RI-Corn had the lowest albedo overall, as well as within its CSW treatment (0.154 ± 0.02), while wheat had the lowest albedo within the annual cropping systems (0.168 ± 0.03).Soybean performed particularly well within both the CO and RI treatments (CO-α: 0.186 ± 0.03; RI-α: 0.182 ± 0.03), with the largest surface reflectivity over the study period.Interannually, variability in albedo was high, with differences in surface reflectivity ranging between 0.012-0.019based on the year (S figure 3).In 2019 perennial switchgrass had the highest albedo (0.178 ± 0.03), while annual wheat in both conventional (CO-α: 0.172 ± 0.04) and reduced input (RI-α: 0.169 ± 0.02) had lower albedo (figure 3).In 2020 annual CO-Corn had the highest albedo (0.179 ± 0.06), slightly higher than perennial switchgrass (0.173 ± 0.04).Annual CO-Soybean had the highest albedo in 2021 (0.186 ± 0.03), while switchgrass had the highest albedo in the following two years (0.191 ± 0.02).In 2023, no-till corn had highest albedo (0.179 ± 0.03).Between 2019 and 2020 prairie strips absorbed the most radiation, contributing to the lowest recorded albedo within those years (0.148 ± 0.02), but then increased its albedo in 2021 to be on par with other perennials (0.169 ± 0.03), and then increased again to 0.172 ± 0.02 in the following two years.

Agronomic practices and climate
The planting of unique crop species led to distinct patterns in albedo.Albedo was observed to increase at the beginning of the growing season during the months of May and early June, where differences between perennials and annual row crops were most apparent during the early establishment period (figure 4).These observed differences between crops diminished during the peak growing season in July and converged around maturity, where crop canopies were homogenous, leaves were dense, and surface reflectivity was similar between all crop species.After harvest-which occurred between the months of August and October, depending on the crop phenology-albedo varied depending on the agronomic management practice.
Species composition led to significant differences in observed albedo on the landscape and distinct visual patterns from agronomic practices.Effects  on albedo were stronger on bare landscapes or sparsely vegetated lands early in the growing season.Perennials increased their canopy coverage and cooling effects early in the growing season, compared to their annual counterparts, which had a barren landscape through much of May.By the middle of the growing season, all crops regardless of species had a defined leaf canopy.These differences were also observed in the variances in surface reflectivity between conventional, no-till, and reduced input CSW rotations, as well as the planting of a winter cover crop (figure 5).Differences between CO and RI rotations of CSW showed that conventional cropping systems developed faster during the growing season and greener leaves from higher measured chlorophyll content resulting from conventional inputs of fertilizer.When a cover crop was planted at the end of the growing season within the RI sites of CSW, differences in albedo between cover crop and bare ground (CO) CSW landscapes after harvest were noticeable, with bare ground having lower albedo during months of September to October before the onset of winter.Variability in annual albedo due to agricultural land use showed that the interaction between year and crop had a significant effect on surface reflectivity (S table 3).Inter-annual differences from unique land usage showed that species composition also played a significant effect in albedo.

Chlorophyll content effect on albedo
Within our study, the highest chlorophyll concentrations were found in corn (CO: 44.20;NT: 40.99;RI: 43.72) and the lowest in wheat (CO: 22.92; RI: 27.85) (S figure 4).We observed that, as chlorophyll content increased, surface reflectivity increased in RI-Soybean (p < 0.05), RI-Wheat and Early Successional landscapes, while in all corn sites (CO-Corn and NT-Corn: p < 0.05) and CO-Wheat (p < 0.01) chlorophyll concentrations increased as surface reflectivity decreased (figure 6).Among different agronomic managements of corn, conventional had the highest chlorophyll content (CO: 44.20), followed by RI, then NT with the lowest.During the peak growing season, chlorophyll content was highly affected by climate and agronomic management, as shown in high swings of SPAD measurements in annual row crops compared to more stable observations of chlorophyll throughout the course of the growing season in perennial grasslands (S figure 5).The Kruskal-Wallis test showed that that pairwise comparisons of land use across species and years were significant, and when delving deeper into our ten landscapes, three sites were statistically significant and two were marginally significant (S tables 1 and 2).In our perennial sites, the decrease in albedo was less steep, with neutral to increasing surface reflectivity over the course of the five-year study period.

Perennials provided higher cooling potential compared to annual crops
When converting from bare ground to either an annual row crop or a perennial grassland, localized cooling effects of various intensities were observed (table 2).Overall, perennials (excluding restored prairie) on average provided higher cooling effects (Switchgrass: −0.291 kg CO 2 e m −2 yr −1 ; E Successional: −0.251 kg CO 2 e m −2 yr −1 ) than their annual crop counterparts (avg: −0.257 kg CO 2 e m −2 yr −1 ).Within each unique landscape, soybeans in both conventional and reduced input agronomic practices had the highest cooling effects, while RI-Corn was almost neutral, trending towards warming impacts.Restored prairie  had low cooling effects on the landscape over the entire study period (PR: −0.135 kg CO 2 e m −2 yr −1 ), but when considering the first two-year establishment period in 2019-2020 and observing cooling effects from 2021 to 2023, prairie displayed higher cooling potentials (−0.259 kg CO 2 e m −2 yr −1 ), on par with switchgrass and early successional landscapes.

Discussion
Field measurements provided in-situ observations of surface reflectivity, which can be used to determine potential warming or cooling effects of albedo on climate.Differences in albedo were observed due to plant phenology, crop species, seasonality, and agronomic practices.

Potential cooling effects from increased surface reflectivity
The substantial RF provided by enhancing the albedo of agroecosystems have made recent efforts towards solar radiation management particularly attractive.Overall, converting a barren or marginal landscape to another candidate ecosystem with higher albedo led to substantial decreases in RF and localized climate cooling, with the degree of cooling dependent on ecosystem inputs and plant species.The peak average in annual mean RF between annual row crops and perennial grasslands averaged −0.265 kg CO 2 m −2 , with soybean, switchgrass, and early successional vegetations having the highest cooling potentials.These results were comparable to Miller et al (2016) who used albedo measurements to compare switchgrass and a conventional rotation of corn-soybean (CS) and found that their CS and switchgrass sites had a cooling potential of −2.67 Wm −2 and −3.37 Wm −2 , respectively.
Plant phenology can also have a significant effect on surface reflectivity.Contemporary corn, wheat, and soybean crops tend to not vary much across fields, nor from one year to another, in planting density, root systems, leaf structure, canopy cover, plant height, and plant composition.Perennials had a higher albedo due to perennials having a longer growing season and a denser, greener, and more homogenous canopy, compared to annual row crops.Albedo had a positive relationship with crop cover throughout the growing season and peaked during peak growing season, declined during senescence (e.g.brown crops, open plant canopies, exposed soil), and then varied at harvest with residue, cover crop planting, or bare soil.This finding can be attributed to the nature of monoculture perennial (e.g.switchgrass) and polyculture perennial (e.g.successional, prairie) grasslands compared to annual row crops (e.g.maize and wheat), as less soil is exposed over the course of the growing season.These results were also observed in Sciusco et al (2020), Sciusco et al (2022) and Lei et al (2023), where surface reflectivity varied across crop species and the choice of crop was shown to directly affect shortwave irradiance and reflectance at both seasonal and annual time scales.These agroecosystems also are highly affected by external factors such as available water and soil moisture, fertilizer, and pesticide inputs.These were specifically observed in comparisons between our conventional and reduced input CSW rotations, where RI treatments had lower albedo and a lower chlorophyll content overall compared to their conventional counterparts.However, in the case of the perennial sites, agronomic contributions may have provided a smaller effect than plant phenology as perennials have a longer growing season that starts as soon as the landscape thaws after the winter freeze, are more resistant to drought (Hawkes and Kiniry 2018) than annual crops, and are not as dependent on the application of fertilizer in their critical stages of growth.

Effects of agronomic management
From seeding to emergence, maturity, senescence, and harvest, farmer actions can significantly affect plant growth and surface reflectivity.Cirilo and Andrade (1994) found that the late planting of corn decreased its potential growth and development due to less solar radiation utilized during emergence and throughout the silking stage.Nielson (2000) also found that leaf stage development is highly correlated to planting date, which can vary based on extreme variances in surface temperature.Both would likely influence surface reflectance.Thus, changes in land use and agronomic practices can modify surface albedo on a large scale, and agronomic management has a strong influence on albedo at the localized level.
Management efforts and landowner preferences for a specific crop affect farmers' crop selection, as do environmental impacts unrelated to climate mitigation (Robertson et al 2017).The establishment of restored prairie had an interesting effect on the landscape during our study, where it showed low albedo during the first two years and then increased by 6% after establishment.In Lei et al (2021), restored prairie's biomass yield was found to be highly variable on the landscape within its first two years, after which it provided high carbon sequestration, high stable yield, and a homogeneous canopy after roots were established.Schramm (1978) similarly found that native grass species such as big bluestem and indiangrass will quickly establish themselves within an ecosystem with large grass stands within two years.This is an appealing finding, as the planting of restored prairie perennials can have a significant effect on localized cooling of the landscape after just two years, compared to the continuous planting of annual row crops, or leaving the landscape barren.As such, planting perennial grasses in marginalized low-efficiency strips on a landscape can lead to more carbon capture and storage over longer periods, and provide cooling effects after the initial establishment period.
Within our study, no-till rotations of CSW had a higher albedo compared to conventional CSW counterparts.Under a no tillage management, albedo was observed to be approximately 0.01-0.03higher than conventional means, providing cooling differences.This was also observed in Sieber et al (2022b), where residue provided higher reflectivity than ploughed soils.Practices such as double cropping, which target keeping the soil surface covered by vegetation throughout the year, and allowing stover from harvest to persist on the landscape can have a beneficial impact on climate via changes in surface reflectivity and increases in carbon sequestration (Hirsch et al 2018, Lugato et al 2020).With the modification of available plant matter on the landscape via tillage practices and planting density, we concluded that surface reflectivity can be affected by human intervention by up to two weeks at the beginning of the growing season, and by up to three weeks at harvest.

The relationship between surface-induced albedo and chlorophyll concentration
One of the principal factors defining the magnitude of plant albedos is the strong absorption of photosynthetically active radiation by chlorophyll in the leaves of plants.In our study for all annual row crops, surface reflectivity decreased with increasing chlorophyll content.As vegetation started to green up in spring, surface albedos rapidly decreased.With plant senescence, reflectance started to increase due to less absorption by now-degraded chlorophyll pigments from browned leaves, silked corn, and dying stalks.Shrestha et al (2018) observed that a change in the planting period of annual row crops could affect the proper arrangement of leaves and the canopy required for better interception of sunlight promoting photosynthesis and other metabolic processes.Hollinger et al (2010) also noted that there can also be a significant relationship between foliage nitrogen and shortwave albedo reflected by the top of the canopy.Within our perennial systems, the decrease in albedo was less steep, with neutral to increasing surface reflectivity from varying chlorophyll content.This shows that increased albedo within the upper boundaries of homogeneous plant canopies can provide higher reflectance compared to the lower levels of the plant canopy, where solar radiation effects from scattering and leaf dynamics are smaller.Thus, our results agree with those of Genesio et al (2021) and Richter et al (2008) who found that crops that have less chlorophyll may still provide a sustainable and relatively simple solution to increasing surface reflectance of local to global regions via changes in surface albedo, while not inhibiting other important agricultural factors provided by perennials such as yield, biodiversity, and productivity.
As the usage of nitrogen input in ecosystems also serves to increase canopy foliar nitrogen content, canopies with higher concentrations of chlorophyll in their leaves have both a higher potential capacity to remove CO 2 from the atmosphere and absorb less incidental shortwave radiation compared to landscapes that have reduced fertilizer applied.Studies by Bray and Sanger (1961), Bray (1962), and Bray et al (1966) have proposed that the use of albedo as an index of chlorophyll content could provide policymakers and farmers critical landscape information in relation to crop water use, fertilization needs and crop yield.Our conventional CSW rotations had typical rates of nitrogen fertilizer, which presumably facilitated faster development of the plants, more homogenous canopies, greener leaves, and greater soil coverage over the course of the growing season.Although there is compelling evidence from this study for a canopy-scale relationship between surface-induced albedo and chlorophyll concentrations, the underlying mechanism remains to be fully understood.Bray et al (1966) and Bartlett et al (2011) did not observe a canopy-level association between increasing canopy nitrogen content and increasing canopy albedo, in direct contrast to the findings of Ollinger et al (2008) where canopy N concentration were strongly correlated with shortwave surface albedo.Notably, all of the aforementioned studies examined tree species: there is little information on the effect of nitrogen on albedo in agricultural grasslands and croplands.Our study bridges this knowledge gap.
As we focused on surface-induced albedo change and its effects on RF and GWI, we did not consider other variables that also affect the direction and magnitude of the energy balance within agroecosystems, which can include latent and sensible heat, outgoing longwave radiation, and plant canopy growth.These factors control fluxes of energy, water and aerodynamic movement between the surface and the atmosphere, which subsequently influence climate on local to global scales (Pielke et al 2011).However, our inclusion of agronomic practices alongside relationships between albedo and canopy nitrogen content provides an additional step towards understanding how seasonality and management changes can affect landscape surface reflectivity.Strategies to enhance surface albedo are premised on the fact that the reflectance properties of leaves vary among different crop species and change seasonally.Overall, our results emphasize the importance of including albedo when evaluating cropping system impacts on RF and GWI, and point to albedo management strategies for cropping systems that could contribute significantly to climate mitigation.

Conclusions
Direct measurements of albedo for five years in perennial and annual agroecosystems in southwest Michigan USA showed the cumulative effects of seasonality and agronomic management on the potential for localized climate cooling.Increased albedo and therefore net climate cooling was observed for all land uses compared to a bare ground, varied between annual and perennial systems, within annual crops, among systems that differed in tillage practices, cover crops, and canopy nitrogen content.Our results show that cropping system changes may either cool or warm the landscape due to spatial and temporal changes in albedo.Understanding the potential magnitudes of the irradiance of cropland surfaces can help develop accurate calculations of albedo in future modeling and climate assessments of agricultural land use change and management.

Figure 1 .
Figure 1.(a) Plot layout showing replicate block (R1-R4) and location of each cropping system (color-coded to each treatment) within selected experimental plots at the MCSE, with albedo and chlorophyll sampling stations indicated by red circles.(b) A closer-view diagram of one plot with all sampling locations.(c) Field personnel Cheyenne Lei (right) and Pietro Sciusco (left) walking between sampling stations in 2020.Photo credit: Jiquan Chen.Map modified from https://lter.kbs.msu.edu/research.

Figure 2 .
Figure 2. Average albedo for all land use types (conventional, no-till and reduced input CSW annual row crop rotations, and perennials switchgrass, early successional and restored prairie) at MCSE.Asterisks indicate statistical significance between crop types at p < 0.05.Error bars represent ±1SD.

Figure 3 .
Figure 3. Mean albedo from 2019 to 2023 (greyscale) for all land use types (conventional, no-till and reduced input CSW annual row crop rotations, and perennials switchgrass, early successional, and restored prairie).Crops within the CSW rotations are depicted using stripes (wheat), crosshatch (corn), and blank (soybean) for each respective year.CSW-No Till was only measured in 2023 during its corn rotation.Error bars represent ±1SD.

Figure 4 .
Figure 4. Stages of growth development of study sites from early growing period to peak growing season.Photos were taken during the summer field visit in 2023 and represent the typical stages of growth for intensely managed agricultural lands.Not included are photographic observations after harvest and planting of a cover crop.

Figure 5 .
Figure 5. Albedo observations for agroecosystems during the growing season May to October.Colored vertical lines and boxes indicate regions for early establishment differences between crops (red), flowering stage between perennials and annual row crops (green), maturity phase where crop canopies are homogenous (grey), harvest (black), and cover crop establishment (yellow), where landscapes are planted with a winter crop, or left bare after harvest.

Figure 6 .
Figure 6.Relationship between mean chlorophyll content to measure surface reflectance for each site from 2019 to 2023.Points indicate each measured sample, blue lines provide the average correlation, and grey boundaries show the standard error.

Table 1 .
Site characteristics of each crop system used in the study, as well as their abbreviations, species composition, fertilization rates, and approximate planting and harvest dates.Early successional sites do not receive any fertilizer input.Conventional and Reduced-Input CSW rotations were divided into each specified annual crop.Species composition for treatments at early successional and restored prairies at the KBS experiment are available at http://lter.kbs.The value of RF α is a representation of the daily local power that would be reflected back to the atmosphere(Carrer et al 2018).By converting albedo into RF, the change in the energy available to the climate caused by a change in surface reflectivity can be parameterized (He et al 2018).Thus, comparisons of global warming impact (GWI) for climate mitigation of greenhouse gases fluxes from ecosystems can be assessed in terms of RF, so that the changes in surface albedo of a given area can be converted into CO 2 -equivalents when surface irradiance and its seasonal changes are known(Betts 2001,  Bright et al 2011).
msu.edu/research/long_term_experiments.bEarly successional grasses and restored prairie strips are not harvested; they were applied with controled burns annually in early March to promote healthy growth and diversity of native species.cNitrogen was applied as ammonium nitrate (N-P-K content: 34-0-0, or 19-17-0, or 19-19-19) and urea ammonium nitrate (UAN;28% N) at planting and/or side-dressing.For wheat, a split application of UAN was applied.d Soybeans within our sites perform biological N fixation, and are not provided any N fertilization, but instead phosphate (0-46-0) at reference landscape.