Global demand for food is increasing, and use of large amounts of potentially human-edible feedstuffs for dairy cows is an important concern. The present study examined whether feeding a by-product-based concentrate combined with high-quality grass silage to high-producing dairy cows affected feed intake and milk production compared with a conventional diet, as well as the effect on efficiency of human food production. In a changeover experiment with four 21-d periods, 24 dairy cows in mid-lactation were offered 9.6 kg of dry matter per day with 1 of 4 concentrates and high-quality grass silage ad libitum. The control concentrate was based on cereal grain (wheat, oat, and barley) and soybean meal, whereas the 3 by-product-based concentrates contained sugar beet pulp in combination with mainly heat-treated rapeseed meal, distillers grain, or a mixture of both. All diets were formulated to be isoenergetic and isonitrogenous. The cows had 10-fold higher starch intake when fed the control diet than when fed the by-product-based concentrates. Silage intake (13 kg of dry matter/d) and milk production (33 kg of energy-corrected milk/d) were not affected by the change in diet. Therefore, replacing cereals and soybean meal with human-inedible by-products in a high-quality forage diet to dairy cows increased net food protein production substantially without lowering milk production.

The development of future food systems will depend on normative decisions taken at different levels by policymakers and stakeholders. Scenario modeling is an adequate tool for assessing the implications of such decisions, but for an enlightened debate, it is important to make explicit and transparent how such value-based decisions affect modeling results. In a participatory approach working with five NGOs, we developed a future food vision for the Nordic countries (Denmark, Finland, Norway and Sweden) through an iterative process of defining the scenario, modeling, and revising the scenario, until a final future food vision was reached. The impacts on food production, land use, and greenhouse gas emissions, and the resulting diets in the food vision, were modeled using a mass flow model of the food system. The food vision formulated was an organic farming system where food is produced locally and livestock production is limited to "leftover streams," i.e., by-products from food production and forage from pastures and perennial grass/clover mixtures, thus limiting food-feed competition. Consumption of meat, especially non-ruminant meat, was substantially reduced compared with current consumption in the Nordic countries (− 81%). An estimated population of 37 million people could be supplied with the scenario diet, which uses 0.21 ha of arable land and causes greenhouse gas emissions of 0.48 tCO 2 e per diet and year. The novelty of this paper includes advancing modeling of sustainable food systems by using an iterative process for designing future food visions based on stakeholder values, which enables results from multidisciplinary modeling (including agronomy, environmental system analysis, animal and human nutrition) to be fed back into the decision-making process, providing an empirical basis for normative decisions and a science-based future vision of sustainable food systems.

Ruminants can produce meat and milk from fibrous feed and byproducts not suitable for human consumption. However, high-yielding dairy cows are generally fed a high proportion of cereal grain and pulses, which could be consumed directly by humans. If high production of dairy cows could be maintained with ingredients of low human interest, the sustainability of dairy production would improve. In the present study, 37 multiparous [Holstein (n = 13) and Swedish Red (n = 24)] dairy cows were followed over a whole lactation. A low-concentrate diet of up to 6 kg concentrate per day (6kgConc) was fed to 27 cows, whereas 10 cows were fed a high-concentrate diet of up to 12 kg concentrate per day (12kgConc). The concentrate was mainly based on byproducts (sugar beet pulp, wheat bran, rapeseed meal, distiller's grain). Grass-clover silage of high digestibility was offered ad libitum. Over the whole lactation, cows on the 6kgConc diet had lower dry matter intake and higher forage intake than cows on the 12kgConc diet. Milk yield and energy balance were not influenced by dietary treatment. However, the cows on the 6kgConc diet numerically produced 2.4 kg less energy-corrected milk than cows on 12kgConc diet. The study lacked the statistical power to identify treatment effects on daily yield below 2.8 kg of milk due to low number of animals per treatment. Feed efficiency (as energy-corrected milk yield/dry matter intake or residual feed intake), body weight change, body condition change, milk fatty acid concentration in total milk fatty acids, plasma nonesterified fatty acids, glucose, β-hydroxybutyrate, and fertility measurements were not affected by diet, supporting the energy balance results. However, higher plasma concentrations of insulin-like growth factor-1 and insulin were observed in cows fed the 12kgConc diet. These findings show that cows can adapt to a high-forage diet virtually without humangrade ingredients, without compromising feed efficiency or energy balance, thereby contributing to sustainable food production.

To lower the effect of climate change from cattle production, we should aim at decreasing their enteric methane emissions per kilogram of milk or meat. Glycerol may be absorbed through the rumen epithelium and would consequently be less available to microbes in the rumen. Glycerol could thus supply dairy cows with energy for milk production without contributing much to methane production. This study evaluated the effect of replacing wheat starch with glycerol on milk production, feed intake, and methane emissions. Twentytwo Swedish Red cows in mid lactation were used in a switch-back, change-over experiment with 3 periods of 21 d. The 2 dietary treatments consisted of a total mixed ration based on (g/kg of dry matter) grass silage (605), rapeseed meal (120), and barley (70) and either wheat starch or refined glycerol (200) fed ad libitum. The glycerol diet resulted in higher dry matter intake (21.6 vs. 20.1 kg/d) and methane emissions (482 vs. 423 g/d) compared with the diet containing wheat starch, whereas no difference was found in energy-corrected milk yield (28.4 vs. 29.7 kg/d). These results indicate that when glycerol is mixed with the feed, it is available to rumen microbes to a larger extent than initially assumed. Compared with wheat starch, adding refined glycerol (200 g/kg of dry matter) to the feed of dairy cows does not seem to have the potential to decrease enteric methane emissions.

Using radio telemetry, the present study simulated the escape of 48 adult rainbow trout Oncorhynchus mykiss from a net-cage fish farm in the Lake Ovre Fryken, Sweden. The post-release dispersal of O. mykiss was fast, showed long-range dispersal behaviour, low winter survival and lacked the ability to find suitable spawning habitats. Thus, the present study suggested that reproducing for the first time may be an obstacle to the establishment of escaped farmed O. mykiss.