The purpose of this experiment was to determine the effects of feeding increasing levels of fresh forage (FF) as a proportion of total dry matter intake (DMI) on nutrient intake, rumen digestion, nutrient utilization, and productive performance of total mixed ration (TMR)-fed cows. Twelve dairy cows (90 ± 22 d in milk, 523 ± 88 kg of body weight, 7,908 ± 719 kg of milk production in the previous lactation) were housed in individual tiestalls and assigned to treatments according to a 3 × 3 Latin square design replicated 4 times. Treatments were 100% TMR (T100), 75% TMR plus 25% FF (T75), and 50% TMR plus 50% FF (T50). The experiment lasted 60 d, divided into 3 periods of 20 d each; the first 12 d of each period were used for diet adaptation and the last 8 d for data collection. The TMR (18.1% crude protein, 24.6% acid detergent fiber) and FF (Lolium multiflorum; 15.1% crude protein, 24.1% acid detergent fiber) were prepared and cut daily and offered to each cow individually. The highest DMI was reached in T100 and T75, which was reflected in greater intake of the different nutrients than T50. No differences were detected in the apparent total digestibility of the nutrients, mean ruminal pH, and total volatile fatty acid concentrations among treatments. Cows in T50 resulted in the lowest ruminal N-NH 3 concentration and the lowest microbial N flow to the duodenum. Milk yield was 8.5% higher from cows in T100 and T75 compared with T50, but we observed no differences for milk fat or milk protein yield among treatments. Milk fat of cows fed T50 had 8% more unsaturated fatty acids (FA) than that of cows fed T100, mostly because of a higher content of monounsaturated FA. Additionally, cows in T50 had a higher concentration of linoleic acid, vaccenic acid, and rumenic acid than T100. Meanwhile, the concentration of linoleic acid and vaccenic acid in cows fed T75 was higher than T100. The milk fat of the cows fed T50 and T75 had a lower n -6: n -3 ratio than T100. We concluded that including up to 29% of FF in the total DMI in combination with a TMR did not affect the intake or digestion of nutrients or the productive response in dairy cows and resulted in a higher concentration of desirable FA from a consumer's perspective.

The monitoring of surface-water quality followed by water-quality modeling and analysis are essential for generating effective strategies in surface-water-resource management. However, worldwide, particularly in developing countries, water-quality studies are limited due to the lack of a complete and reliable dataset of surface-water-quality variables. In this context, several statistical and machine-learning models were assessed for imputing water-quality data at six monitoring stations located in the Santa Lucía Chico river (Uruguay), a mixed lotic and lentic river system. The challenge of this study is represented by the high percentage of missing data (between 50% and 70%) and the high temporal and spatial variability that characterizes the water-quality variables. The competing algorithms implement univariate and multivariate imputation methods (inverse distance weighting (IDW), Random Forest Regressor (RFR), Ridge (R), Bayesian Ridge (BR), AdaBoost (AB), Hubber Regressor (HR), Support Vector Regressor (SVR) and K-nearest neighbors Regressor (KNNR)). According to the results, more than 76% of the imputation outcomes are considered “satisfactory” (NSE > 0.45). The imputation performance shows better results at the monitoring stations located inside the reservoir than those positioned along the mainstream. IDW was the model with the best imputation results, followed by RFR, HR and SVR. The approach proposed in this study is expected to aid water-resource researchers and managers in augmenting water-quality datasets and overcoming the missing data issue to increase the number of future studies related to the water-quality matter.

To the best of our knowledge, this paper presents the first Internet Domain Name System (DNS) queries data study from a national K-12 Education Service Provider. This provider, called Plan Ceibal , supports a one-to-one computing program in Uruguay. Additionally, it has deployed an Information and Communications Technology (ICT) infrastructure in all of Uruguay’s public schools and high-schools, in addition to many public spaces. The main development is wireless connectivity, which allows all the students (whose ages range between 6 and 18 years old) to connect to different resources, including Internet access. In this article, we use 9,125,888,714 DNS-query records, collected from March to May 2019, to study Plan Ceibal user’s Internet behavior applying unsupervised machine learning techniques. Firstly, we conducted a statistical analysis aiming at depicting the distribution of the data. Then, to understand users’ Internet behavior, we performed principal component analysis (PCA) and clustering methods. The results show that Internet use behavior is influenced by age-group and time of the day. However, it is independent of the geographical location of the users. Internet use behavior analysis is of paramount importance for evidence-based decision making by any education network provider, not only from the network-operator perspective but also for providing crucial information for learning analytics purposes.

The monitoring of surface-water quality followed by water-quality modeling and analysis is essential for generating effective strategies in water-resource management. However, worldwide, particularly in developing countries, water-quality studies are limited due to the lack of a complete and reliable dataset of surface-water-quality variables. In this context, several statistical and machine-learning models were assessed for imputing water-quality data at six monitoring stations located in the Santa Lucía Chico river (Uruguay), a mixed lotic and lentic river system. The challenge of this study is represented by the high percentage of missing data (between 50% and 70%) and the high temporal and spatial variability that characterizes the water-quality variables. The competing algorithms implemented belonged to both univariate and multivariate imputation methods (inverse distance weighting (IDW), Random Forest Regressor (RFR), Ridge (R), Bayesian Ridge (BR), AdaBoost (AB), Hubber Regressor (HR), Support Vector Regressor (SVR), and K-nearest neighbors Regressor (KNNR)). According to the results, more than 76% of the imputation outcomes are considered satisfactory (NSE > 0.45). The imputation performance shows better results at the monitoring stations located inside the reservoir than the ones positioned along the mainstream. IDW was the most chosen model for data imputation.