Publikasjoner
NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.
2025
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Forfattere
Jing Zhou Qianyi Duan Nicole AndersonSammendrag
Accurate classification of grass seed crop species is essential for estimating seasonal field acreages, informing market strategies, promoting crop diversification, and establishing long-term cropping histories. Unlike major commodity crops, grass seed crops lack reliable datasets and mapping products. This study investigates the use of spaceborne imagery and artificial intelligence (AI)-driven computer vision to remotely classify grass seed crops. Our ground observation dataset comprises 15 grass seed species grown in Oregon, USA (2021-2023), covering over 4,000 data points. Satellite imagery was acquired from Sentinel-2 (S2) spanning January 1st to June 14th each study year. The imagery includes 12 bands across 400-2190 nm with a spatial resolution of 10 m pixel-1, collected at five-day intervals, totalling 34 time stamps. Statistical analyses identified the second and third weeks of May as the most critical temporal window for spectrally distinguishing among grass species using satellite imagery, coinciding with field inspection timing for crop purity. The near-infrared [835.1 nm (S2A) / 833 nm (S2B)], red edge [740.2 nm (S2A) / 739.1 nm (S2B)], and narrow near-infrared [864.8 nm (S2A) / 864 nm (S2B)] bands showed the highest spectral separability among major grass species. A U-Net Temporal Attention Encoder (U-TAE) model was trained to classify grass seed crop species, integrating temporal and spectral data. The overall classification accuracy - defined as the ratio of correctly classified samples to total samples - was 0.89 across all 15 grass species with high accuracies for four major species, including tall fescue (0.93) (Schedonorus arundinaceus (Shreb.) Dumort.), perennial ryegrass (0.90) (Lolium perenne L.), annual ryegrass (0.87) (Lolium perenne L. ssp. Multiflorum (Lam.) Husnot), and Kentucky bluegrass (0.83) (Poa pratensis L.) (0.83). Our findings provide actionable insights for industry stakeholders, enabling informed pricing, planting strategies, and reduced risk of cross-pollination. This work highlights the potential of AI and remote sensing in grass seed crop production, with future efforts focused at estimating field acreage and predicting production potential.
Forfattere
Corinne Butler Amber Moore Nicole AndersonSammendrag
Black soldier fly larvae (BSFL) frass, a byproduct of insect farming, can be used as a nitrogen (N) source in cereal production, although its impacts on grain protein, grain yield, and nitrate (NO3-N) leaching are not well understood. This study determined the effect of different BSFL frass and urea combinations on hard red spring (HRS) wheat grain protein, grain yield, above-ground N uptake, post-harvest soil N, and NO3-N leaching. Wheat response to three urea/BSFL frass blends (100% urea/0% frass, 67% urea/33% frass, and 33% urea/67% frass) was evaluated at two N rates (141 and 281 kg N ha-1) alongside a non-amended control in a greenhouse using a Willamette silt loam (soil NO3-N concentration of 13.6 mg kg-1). At 141 kg N ha-1, increasing frass-N from 0 to 67% of the total N in the urea-frass blends caused a linear decline in grain yield from 26 to 19 g pot-1, with grain protein only declining from 33 to 67% frass-N. At 281 kg N ha-1, yield, protein, and plant N uptake declined when frass-N increased from 33 to 67%. Replacing 33% urea-N with frass-N at 281 kg N ha-1 decreased soil NO3-N by 86%. A leaching component showed NO3-N leaching was 17x higher for 100% urea than the 33% urea/67% BSFL frass blend at 281 kg N ha-1. Gradual mineralization of organic frass-N may have limited N early in the season, decreasing yield at increased frass proportions, whereas extended N mineralization later in the season helped maintain grain protein concentrations. Continued plant N uptake and unreleased organic frass-N at season’s end likely decreased soil NO3-N accumulation with frass blends. These results indicate that substituting up to one-third of urea with BSFL frass at recommended N rates may sustain HRS wheat grain yield and grain protein while substantially reducing NO3-N leaching potential.
Forfattere
Trygve S. AamlidSammendrag
Det er ikke registrert sammendrag
Forfattere
Trygve S. AamlidSammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag