Hopp til hovedinnholdet

Publications

NIBIOs employees contribute to several hundred scientific articles and research reports every year. You can browse or search in our collection which contains references and links to these publications as well as other research and dissemination activities. The collection is continously updated with new and historical material.

2023

To document

Abstract

The large-scale import of soybean products into the EU decreases the self-sufficiency of livestock production. The fractionation of grassland forage crops presents an opportunity to locally produce protein-rich feed for monogastrics. Two promising fractionation methods, twin-screw press juicing and leaf stripping, were evaluated in parallel in field experiments established in Norway and Sweden to compare the nutrient composition and yield of the resulting biorefined and residual fractions. The clearest delineation between the methods was in the ash-free neutral detergent fibre (aNDFom) concentration, with juicing producing a biorefined fraction with a lower aNDFom than leaf stripping. Variability in the allocation of crude protein (CP) and biomass to the biorefined fractions occurred in both methods between cuts and locations and is likely due to differing stand characteristics and inconsistency in machine functionality. Additional work is needed to understand how characteristics such as stand density, botanical composition, and plant phenological stage impact each fractionation method’s ability to allocate protein, fibre, and biomass into the resulting fractions. Future studies should focus particularly on determining standardised settings for leaf stripping machinery based on a range of stand characteristics to ensure consistency in the yield and nutrient composition of the resulting fractions.

To document

Abstract

Faced with risky yields and returns, risk-averse farmers require a premium to take risks. In this paper, we estimate individual farmers’ degrees of risk aversion to adjust for the risk premium in returns and to replace the farmers’ realized returns with their certainty equivalent returns in the production function. In that way, the effect of the inputs on returns will automatically be risk-adjusted, i.e., we obtain risk-adjusted marginal effects of inputs, which can be used in decision-making support of farmers’ input choices in production. Using farm-level data from organic basmati rice smallholders in India, we illustrate this method using nonparametric production functions. The results show that the input elasticities and returns-to-scale estimates change when the farmers’ degree of risk aversion is taken into consideration.

To document

Abstract

This paper synthesizes a five-year project (BIOWATER) that assessed the effects of a developing bioeconomy on Nordic freshwaters. We used a catchment perspective and combined several approaches: comparative analyses of long-term data sets from well-monitored catchments (agricultural, with forestry, and near pristine) across Fennoscandia, catchment biogeochemical modelling and ecosystem services assessment for integration. Various mitigation measures were also studied. Benchmark Shared Socio-economic Pathways were downscaled and articulated in dialogue with national stakeholder representatives leading to five Nordic Bioeconomy Pathways (NBPs) describing plausible but different trajectories of societal development towards 2050.These were then used for catchment modelling and ecosystem service assessment. Key findings from the work synthesized here are: (a) The monitoring results from 69 catchments demonstrate that agricultural lands exported an order of magnitude more nutrients than natural catchments (medians 44 vs 4 kg P km−2 y-1 and 1450 vs 139 kg N km−2 y-1) whilst forests were intermediate (7 kg P km−2 y-1 and 200 kg N km−2 y-1). (b) Our contrasting scenarios led to substantial differences in land use patterns, which affected river flow as well as nutrient loads in two of the four modelled catchments (Danish Odense Å and Norwegian Skuterud), but not in two others (Swedish catchment C6 and Finnish Simojoki). (c) Strongly contrasting scenarios (NBP1 maximizing resource circularity versus NBP5 maximizing short-term profit) were found to lead to similar monetary estimates of total societal benefits, though for different underlying reasons – a pattern similar across the six studied Nordic catchments. (d) The ecological status of small to medium sized rivers in agricultural landscapes benefitted greatly from an increase in riparian forest cover from 10 % to 60 %. Riparian buffer strips, constructed wetlands, rewetting of ditched peatlands, and similar nature-based solutions optimize natural biogeochemical processes and thus can help in mitigating negative impacts of intensified biomass removal on water quality.

To document

Abstract

Fire in the boreal forests emits substantial amounts of organically bound carbon (C) to the atmosphere and converts a fraction of the burnt organic matter into charcoal, which in turn is highly refractory and functions as a long-term stable C pool. It is well established that the boreal forest charcoal pool is sufficiently large to play a significant role in the global C cycle. However, there is a need for spatially representative estimates of how large proportions of the forest floor C pool are made up of charcoal across different plant communities in the boreal forest ecosystem. Thus, we have quantified the amounts of C separately in charcoal and the organic layers of the forest floor across fine spatial scales in a boreal forest landscape with a well-documented fire history. We found that the proportion of charcoal C made up an average of 1.2% of the total forest floor C, and the charcoal proportions showed a high small-scale spatial variability and were concentrated in the organic–mineral soil interface. Proportions of charcoal C decreased with increasing time since last fire. Deeper soils, denser soils, and local concave areas had the highest proportions of charcoal C, whereas historical fire frequencies and current differences in vegetation did not relate to the proportions of charcoal C.

To document

Abstract

Virtual fencing is a promising alternative to contain livestock dispersal without using physical barriers. This technology uses smart-wearable collars that deliver predictable warning tones to animals when they approach virtual boundaries paired with mild electric pulses. Virtual fencing allows for dynamic management of livestock grazing, based on site-specific variations in the quality and quantity of forages. However, several factors can affect the efficacy of virtual fencing, including the length of prior experience with virtual fencing, climatic conditions, forage availability inside and outside virtual fencing paddocks and collar configuration schedules. Lactation requirements and social interactions between collared cows and uncollared calves can also influence the efficacy of the technology. Virtual fencing trials were conducted at the New Mexico State University’s Chihuahuan Desert Rangeland Research Center from August 27 to December 21 of 2022 to evaluate the efficacy of virtual fencing to manage rangeland cows during late lactation and following weaning. Twenty-six Brangus cows previously trained to use NoFence C2 collars (NoFence, Batnfjordsøra, Norway), were monitored for 30 days during late lactation and 28 days after weaning. Collared cows and uncollared calf pairs were allocated to four virtual fence pastures in late lactation and after weaning, with pasture duration (4.2 ± 0.6 d), size (72 ± 19 ha) and perimeter (4,523 ± 352 m) varying according to forage availability and access to fresh drinking water. Audio cues, electric pulses and ratio of electric pulses to audio cues before and after weaning were compared by ANOVA in a Completely Randomized Design replicated across pre-weaning and post-weaning pastures (n = 8). The average number of electric pulses per cow was greater (P < 0.0004) for pre-weaning (3.7 ± 0.2) than for post-weaning post-weaning (1.6 ± 0.3) pastures. The number of audio warnings per cow was also greater (P < 0.0001) for pre-weaning (52 ± 3.3) than post-weaning (34 ± 3.3) pastures. Conversely, cows had decreased (P < 0.0001) ratios of electric pulses relative to audio tones on post-weaning (4.8 ± 0.5%) than pre-weaning (7.0 ± 0.8%) pastures. These results suggest that cows responded better to virtual fencing after weaning, likely because weaned cows were no longer affected by social interactions with uncollared calves. Furthermore, cows after weaning apparently relied on warning tones and fewer electric pulses to interact safely with virtual fences. However, it is important to note that sources of variation not accounted for or controlled by the present experimental design may have also affected the recorded interactions with virtual fences in the present study.

To document

Abstract

Detection of parturition of rangeland cows remotely may be possible using low cost LoRa WAN monitoring systems that are capable of logging and transmitting cow activity and position data in real time. This study evaluated candidate algorithms for early detection of parturition using longitudinal data of cow activity and position collected by GPS and triaxial accelerometers. Trials were conducted at the USDA Jornada Experimental Range from November to December 2022. Five Raramuri Criollo and five Angus x Hereford mature cows were equipped with LoRa WAN tracking collars instrumented with GPS and triaxial accelerometers and monitored through late gestation (> 7 months) while grazing rangeland pastures of 1,230 and 2,200 ha, respectively. Animal location (latitude and longitude) and activity count (Ac) obtained from GPS and accelerometers data, respectively, were collected by receiving stations that transmitted data in real time through a LoRa WAN network. Collars transmitted GPS positions at one-hour intervals and Ac data at two-minute intervals. An operator routinely inspected focal cows in herds to register parturition within approximately 12 h accuracy. Sensor data for 21 days prior to calving were processed to calculate distance traveled (m/h) and activity rate (Ac/h). For each hour interval, the adjusted activity Index IN = activity/distance (Ac/m) was computed to disentangle motion changes not associated with walking activity. Two algorithms were tested. The first considered the temporal deviation (D) of IN for a given hour (X0), compared with the average IN of the same hour in the previous seven days: D = INX0 /(INX-1+ INX-2 + …+ INX-7)/7). The second considered the normalized probability (N) of D for a given hour (X0) compared with the same hour over the previous seven days: N = (INX0-(INX-1+ INX-2 + …+ INX-7)/7)/sd.(INX-0, INX-1, …, INX-7). A threshold for high probability of calving was set when at least three consecutive hours with D >3 or N >0.95 were detected. Both algorithms correctly triggered alerts on actual calving days. Thus, lack of detection or false detections of calving indicated that the sensitivity and specificity for calving detection were both 100%. The normalized method (N) triggered delayed calving alerts in two cases. Furthermore, greater (P < 0.05) number of consecutive hours with D > 3 (5.6 ± 2.1) around actual calving time were detected vs. the number of consecutive hours with N > 0.95 (3.9 ± 1.2), suggesting that the former algorithm was also able to detect longer duration of behaviors associated with calving. Results indicate possibilities for remote detection of the onset and duration of calving behavior (parturition + first nursing hours) of beef cows managed on large rangeland pastures that impose operational challenges for visual inspection of cows during calving. Further tests with a greater number of cows and management systems would be needed to confirm this hypothesis.

To document

Abstract

Monitoring cattle on rangelands is a daunting task that can be improved by using wearable sensors that are capable of transmitting motion and position data in real time and at low cost. This study tested the performance of machine learning (ML) classifiers to discriminate among foraging activities of cows based on triaxial accelerometer data collected in real-time by LoRa WAN networks. Trials were conducted at the New Mexico State University Chihuahuan Desert Rangeland Research Center and the USDA Jornada Experimental Range in Doña Ana County, NM. A total of 24 Brangus, Brahman, Raramuri Criollo and Angus x Hereford mature cows fitted with LoRa WAN tracking collars housing GPS and triaxial accelerometers were monitored across four periods during the 2022 summer and fall seasons on desert rangeland pastures. Trackers integrated and transmitted activity count (Ac) data from accelerometers at one-minute intervals. Video recording of focal cows (n = 24) was undertaken during daylight hours (0630 to 2000 h) from a distance of ~30 m to minimize interference with natural behaviors. A total of 168 hours of video were recorded and inspected by an experienced observer to label video files according to a classification tree of four main activities: grazing (GR), walking (WA), resting (RE) and ruminating (RU), and two states: active (AC) or static (ST). Individualized activities and states were considered when cows performed the same predefined activity or state for more than 30 secs. Retrieved sensor data from collar trackers were labeled by state and activity according to labels collected from video records. This classification resulted in a dataset containing 9,222 events, including 3,928 for GR, 2,286 for WA, 2,032 for RE, and 976 for RU, as well as 6,214 labels for AC and 3,008 labels for ST. Deep learning through Multilayer Perceptron Classifiers (MLPC) were coded and implemented using a split configuration of 70% of the data for training and 30% for testing, respectively. In preliminary runs, models had reduced ability to properly discriminate among RE (F1 = 0.42) and RU (F1 = 0.43) Thus, RE and RU were merged on subsequent tests, resulting in 3,928 labels for GR, 2,286 labels for WA, and 3,008 labels for merged RE. Deep learning models successfully classified between AC vs. ST behavior with an overall F1 performance score of 0.96. Further use of the same deep learning models successfully classified among GR, WA, and RE activities with an overall F1 performance score of 0.91, suggesting satisfactory application of the trained models to monitor cattle grazing activities on desert rangeland.