Seven subjects pedalled on a Monark cycle ergometer as fast as possible for approximately 7 s against four different resistances which corresponded to braking torques (TB) equal to 19, 38, 57 and 76 N.m at the crank level. Exercise periods were separated by 5-min recovery periods. Pedal velocity was recorded every 50 ms by means of a disc with 360 slots fixed on the flywheel, passing in front of a photo-electric cell linked to a microcomputer which processed the data. Every 50 ms, the time necessary to perform half a pedal revolution (t1/2) was computed by adding the 50-ms periods necessary to reach 669 slots (the number of slots corresponding to half a pedal revolution). To measure t1/2 to an accuracy better than 50 ms, this time was computed by a linear interpolation of the time-slot number relationship. Power (P) was averaged during t1/2 by adding the power dissipated against braking torque and the power necessary to accelerate the flywheel. The torque-velocity (T-v) relationship was studied during the acceleration phase of a sprint against a single TB by computing every 50 ms the relationship between v and T (N.m), equal to the sum of TB and the torque necessary to accelerate the flywheel at the same time. The T-v relationships calculated from the acceleration phase of a single all-out exercise were linear and similar to the previously described relationships between peak velocity and braking force.(ABSTRACT TRUNCATED AT 250 WORDS)

A group of 24 subjects performed on a cycle ergometer a fatigue test consisting of four successive all-out sprints against the same braking torque. The subjects were not allowed time to recover between sprints and consequently the test duration was shorter than 30 s. The pedal velocity was recorded every 10 ms from a disc fixed to the flywheel with 360 slots passing in front of a photo-electric cell linked to a microcomputer which processed the data. Taking into account the variation of kinetic energy of the ergometer flywheel, it was possible to determine the linear torque velocity relationship from data obtained during the all-out cycling exercise by computing torque and velocity from zero velocity to peak velocity according to a method proposed previously. The maximal theoretical velocity (v(0)) and the maximal theoretical torque (T(0)) were estimated by extrapolation of each torque-velocity relationship. Maximal power (P(max)) was calculated from the values of T(0) and v(0) (P(max) = 0.25v(0)T(0). The kinetics of v(0), T(0) and P(max) was assumed to express the effects of fatigue on the muscle contractile properties (maximal shortening velocity, maximal muscle strength and maximal power). Fatigue induced a parallel shift to the left of the torque-velocity relationships. The v( 0), T(0) and P(max) decreases were equal to 16.3 percent, 17.3 percent and 31 percent, respectively. The magnitude of the decrease was similar for v(0) and T(0) which suggested that P max decreased because of a slowing of maximal shortening velocity as well as a loss in maximal muscle strength. However, the interpretation of a decrease in cycling v(0) which has the dimension of a maximal cycling frequency is made difficult by the possible interactions between the agonistic and the antagonistic muscles and could also be explained by a slowing of the muscle relaxation rate.

No abstract

This paper provides an early assessment of the effects of the COVID-19 outbreak and of subsequent response measures on milk production, collection, processing, marketing and consumption in Africa. We focus on the period surrounding the first wave of the outbreak (from February to June 2020), during which the number of cases surged and many steps were taken to curb the epidemic. The paper is based on reports from four countries covered by the Africa-Milk Research Project: Burkina Faso, Kenya, Madagascar and Senegal. Data was collected primarily from nine dairy processors located in those countries. Major conclusions of the study are: (1) Dairy farmers were negatively affected by COVID-19 measures when the health crisis coincided with the peak of the milk production season, and when governments did not take steps to support milk production. (2) Small and informal milk collectors were also affected by traffic restrictions as they could not obtain traffic permits. (3) Milk powder importation remained unaffected during the outbreak. (4) Dairy processors (particularly small ones) faced many challenges restricting their operation. Travel restrictions led to temporary interruptions of milk supply, and because of employee protection and safety measures, processing costs increased. (5) Many small retailers were affected by bans on public transport and reduced their purchases of artisanal dairy products; meanwhile, spoilage of dairy products increased during long curfews coupled with poor storage conditions. Supermarkets were able to increase their market share during the pandemic thanks to their connections with industrial dairy processors and wholesalers. (6) A majority of consumers decreased their consumption of dairy products due to a decrease of purchasing power. In some cases, an increase in consumption occurred (due to Ramadan month and dry season high temperatures) and consumption shifted towards long-life dairy products. (7) Overall, the consequences of the health crisis affected more small and informal dairy supply chains than the larger ones, which are more formal, better organised and finally more resilient to face this kind of global crisis.

Deep learning provide successful applications in many fields. Recently, machines learning are involved for oceans remote sensing applications. In this study, we use and compare about eight (8) deep learning estimators for retrieval of a mainly pigment of phytoplankton. Depending on the water case and the multiple instruments simultaneously observing the earth on a variety of platforms, several algorithm are used to estimate the chlolophyll-a from marine reflectance.By using a long-term multi-sensor time-series of satellite ocean-colour data, as MODIS, SeaWifs, VIIRS, MERIS, etc…, we make a unique deep network model able to establish a relationship between sea surface reflectance and chlorophyll-a from any measurement satellite sensor over West Africa. These data fusion take into account the bias between case water and instruments.We construct several chlorophyll-a concentration prediction deep learning based models, compare them and therefore use the best for our study. Results obtained for accuracy training and test are quite good. The mean absolute error are very low and vary between 0,07 to 0,13 mg/m 3 .