Leg and arm lactate and substrate kinetics during exercise

[EN] To study the role of muscle mass and muscle activity on lactate and energy kinetics during exercise, whole body and limb lactate, glucose, and fatty acid fluxes were determined in six elite cross-country skiers during roller-skiing for 40 min with the diagonal stride (Continuous Arm + Leg) followed by 10 min of double poling and diagonal stride at 72-76% maximal O(2) uptake. A high lactate appearance rate (R(a), 184 +/- 17 micromol x kg(-1) x min(-1)) but a low arterial lactate concentration ( approximately 2.5 mmol/l) were observed during Continuous Arm + Leg despite a substantial net lactate release by the arm of approximately 2.1 mmol/min, which was balanced by a similar net lactate uptake by the leg. Whole body and limb lactate oxidation during Continuous Arm + Leg was approximately 45% at rest and approximately 95% of disappearance rate and limb lactate uptake, respectively. Limb lactate kinetics changed multiple times when exercise mode was changed. Whole body glucose and glycerol turnover was unchanged during the different skiing modes; however, limb net glucose uptake changed severalfold. In conclusion, the arterial lactate concentration can be maintained at a relatively low level despite high lactate R(a) during exercise with a large muscle mass because of the large capacity of active skeletal muscle to take up lactate, which is tightly correlated with lactate delivery. The limb lactate uptake during exercise is oxidized at rates far above resting oxygen consumption, implying that lactate uptake and subsequent oxidation are also dependent on an elevated metabolic rate. The relative contribution of whole body and limb lactate oxidation is between 20 and 30% of total carbohydrate oxidation at rest and during exercise under the various conditions. Skeletal muscle can change its limb net glucose uptake severalfold within minutes, causing a redistribution of the available glucose because whole body glucose turnover was unchanged.

Pulmonary gas exchange and acid-base state at 5,260 m in high-altitude Bolivians and acclimatized lowlanders

[EN] Pulmonary gas exchange and acid-base state were compared in nine Danish lowlanders (L) acclimatized to 5,260 m for 9 wk and seven native Bolivian residents (N) of La Paz (altitude 3,600-4,100 m) brought acutely to this altitude. We evaluated normalcy of arterial pH and assessed pulmonary gas exchange and acid-base balance at rest and during peak exercise when breathing room air and 55% O2. Despite 9 wk at 5,260 m and considerable renal bicarbonate excretion (arterial plasma HCO3- concentration = 15.1 meq/l), resting arterial pH in L was 7.48 +/- 0.007 (significantly greater than 7.40). On the other hand, arterial pH in N was only 7.43 +/- 0.004 (despite arterial O2 saturation of 77%) after ascent from 3,600-4,100 to 5,260 m in 2 h. Maximal power output was similar in the two groups breathing air, whereas on 55% O2 only L showed a significant increase. During exercise in air, arterial PCO2 was 8 Torr lower in L than in N (P < 0.001), yet PO2 was the same such that, at maximal O2 uptake, alveolar-arterial PO2 difference was lower in N (5.3 +/- 1.3 Torr) than in L (10.5 +/- 0.8 Torr), P = 0.004. Calculated O2 diffusing capacity was 40% higher in N than in L and, if referenced to maximal hyperoxic work, capacity was 73% greater in N. Buffering of lactic acid was greater in N, with 20% less increase in base deficit per millimole per liter rise in lactate. These data show in L persistent alkalosis even after 9 wk at 5,260 m. In N, the data show 1) insignificant reduction in exercise capacity when breathing air at 5,260 m compared with breathing 55% O2; 2) very little ventilatory response to acute hypoxemia (judged by arterial pH and arterial PCO2 responses to hyperoxia); 3) during exercise, greater pulmonary diffusing capacity than in L, allowing maintenance of arterial PO2 despite lower ventilation; and 4) better buffering of lactic acid. These results support and extend similar observations concerning adaptation in lung function in these and other high-altitude native groups previously performed at much lower altitudes.

The re-establishment of the normal blood lactate response to exercise in humans after prolonged acclimatization to altitude

[EN] 1. One to five weeks of chronic exposure to hypoxia has been shown to reduce peak blood lactate concentration compared to acute exposure to hypoxia during exercise, the high altitude 'lactate paradox'. However, we hypothesize that a sufficiently long exposure to hypoxia would result in a blood lactate and net lactate release from the active leg to an extent similar to that observed in acute hypoxia, independent of work intensity. 2. Six Danish lowlanders (25-26 years) were studied during graded incremental bicycle exercise under four conditions: at sea level breathing either ambient air (0 m normoxia) or a low-oxygen gas mixture (10 % O(2) in N(2), 0 m acute hypoxia) and after 9 weeks of acclimatization to 5260 m breathing either ambient air (5260 m chronic hypoxia) or a normoxic gas mixture (47 % O(2) in N(2), 5260 m acute normoxia). In addition, one-leg knee-extensor exercise was performed during 5260 m chronic hypoxia and 5260 m acute normoxia. 3. During incremental bicycle exercise, the arterial lactate concentrations were similar at sub-maximal work at 0 m acute hypoxia and 5260 m chronic hypoxia but higher compared to both 0 m normoxia and 5260 m acute normoxia. However, peak lactate concentration was similar under all conditions (10.0 +/- 1.3, 10.7 +/- 2.0, 10.9 +/- 2.3 and 11.0 +/- 1.0 mmol l(-1)) at 0 m normoxia, 0 m acute hypoxia, 5260 m chronic hypoxia and 5260 m acute normoxia, respectively. Despite a similar lactate concentration at sub-maximal and maximal workload, the net lactate release from the leg was lower during 0 m acute hypoxia (peak 8.4 +/- 1.6 mmol min(-1)) than at 5260 m chronic hypoxia (peak 12.8 +/- 2.2 mmol min(-1)). The same was observed for 0 m normoxia (peak 8.9 +/- 2.0 mmol min(-1)) compared to 5260 m acute normoxia (peak 12.6 +/- 3.6 mmol min(-1)). Exercise after acclimatization with a small muscle mass (one-leg knee-extensor) elicited similar lactate concentrations (peak 4.4 +/- 0.2 vs. 3.9 +/- 0.3 mmol l(-1)) and net lactate release (peak 16.4 +/- 1.8 vs. 14.3 mmol l(-1)) from the active leg at 5260 m chronic hypoxia and 5260 m acute normoxia. 4. In conclusion, in lowlanders acclimatized for 9 weeks to an altitude of 5260 m, the arterial lactate concentration was similar at 0 m acute hypoxia and 5260 m chronic hypoxia. The net lactate release from the active leg was higher at 5260 m chronic hypoxia compared to 0 m acute hypoxia, implying an enhanced lactate utilization with prolonged acclimatization to altitude. The present study clearly shows the absence of a lactate paradox in lowlanders sufficiently acclimatized to altitude.

Parasympathetic neural activity accounts for the lowering of exercise heart rate at high altitude

[EN] BACKGROUND: In chronic hypoxia, both heart rate (HR) and cardiac output (Q) are reduced during exercise. The role of parasympathetic neural activity in lowering HR is unresolved, and its influence on Q and oxygen transport at high altitude has never been studied. METHODS AND RESULTS: HR, Q, oxygen uptake, mean arterial pressure, and leg blood flow were determined at rest and during cycle exercise with and without vagal blockade with glycopyrrolate in 7 healthy lowlanders after 9 weeks' residence at >/=5260 m (ALT). At ALT, glycopyrrolate increased resting HR by 80 bpm (73+/-4 to 153+/-4 bpm) compared with 53 bpm (61+/-3 to 114+/-6 bpm) at sea level (SL). During exercise at ALT, glycopyrrolate increased HR by approximately 40 bpm both at submaximal (127+/-4 to 170+/-3 bpm; 118 W) and maximal (141+/-6 to 180+/-2 bpm) exercise, whereas at SL, the increase was only by 16 bpm (137+/-6 to 153+/-4 bpm) at 118 W, with no effect at maximal exercise (181+/-2 bpm). Despite restoration of maximal HR to SL values, glycopyrrolate had no influence on Q, which was reduced at ALT. Breathing FIO(2)=0.55 at peak exercise restored Q and power output to SL values. CONCLUSIONS: Enhanced parasympathetic neural activity accounts for the lowering of HR during exercise at ALT without influencing Q. The abrupt restoration of peak exercise Q in chronic hypoxia to maximal SL values when arterial PO(2) and SO(2) are similarly increased suggests hypoxia-mediated attenuation of Q.

Muscle glycogen resynthesis during recovery from cycle exercise: no effect of additional protein ingestion

[EN] In the present study, we have investigated the effect of carbohydrate and protein hydrolysate ingestion on muscle glycogen resynthesis during 4 h of recovery from intense cycle exercise. Five volunteers were studied during recovery while they ingested, immediately after exercise, a 600-ml bolus and then every 15 min a 150-ml bolus containing 1) 1.67 g. kg body wt(-1). l(-1) of sucrose and 0.5 g. kg body wt(-1). l(-1) of a whey protein hydrolysate (CHO/protein), 2) 1.67 g. kg body wt(-1). l(-1) of sucrose (CHO), and 3) water. CHO/protein and CHO ingestion caused an increased arterial glucose concentration compared with water ingestion during 4 h of recovery. With CHO ingestion, glucose concentration was 1-1.5 mmol/l higher during the first hour of recovery compared with CHO/protein ingestion. Leg glucose uptake was initially 0.7 mmol/min with water ingestion and decreased gradually with no measurable glucose uptake observed at 3 h of recovery. Leg glucose uptake was rather constant at 0.9 mmol/min with CHO/protein and CHO ingestion, and insulin levels were stable at 70, 45, and 5 mU/l for CHO/protein, CHO, and water ingestion, respectively. Glycogen resynthesis rates were 52 +/- 7, 48 +/- 5, and 18 +/- 6 for the first 1.5 h of recovery and decreased to 30 +/- 6, 36 +/- 3, and 8 +/- 6 mmol. kg dry muscle(-1). h(-1) between 1.5 and 4 h for CHO/protein, CHO, and water ingestion, respectively. No differences could be observed between CHO/protein and CHO ingestion ingestion. It is concluded that coingestion of carbohydrate and protein, compared with ingestion of carbohydrate alone, did not increase leg glucose uptake or glycogen resynthesis rate further when carbohydrate was ingested in sufficient amounts every 15 min to induce an optimal rate of glycogen resynthesis.

Metabolic and thermodynamic responses to dehydration-induced reductions in muscle blood flow in exercising humans

[EN] 1. The present study examined whether reductions in muscle blood flow with exercise-induced dehydration would reduce substrate delivery and metabolite and heat removal to and from active skeletal muscles during prolonged exercise in the heat. A second aim was to examine the effects of dehydration on fuel utilisation across the exercising leg and identify factors related to fatigue. 2. Seven cyclists performed two cycle ergometer exercise trials in the heat (35 C; 61 +/- 2 % of maximal oxygen consumption rate, VO2,max), separated by 1 week. During the first trial (dehydration, DE), they cycled until volitional exhaustion (135 +/- 4 min, mean +/- s.e.m.), while developing progressive DE and hyperthermia (3.9 +/- 0.3 % body weight loss and 39.7 +/- 0.2 C oesophageal temperature, Toes). On the second trial (control), they cycled for the same period of time maintaining euhydration by ingesting fluids and stabilising Toes at 38.2 +/- 0.1 degrees C. 3. After 20 min of exercise in both trials, leg blood flow (LBF) and leg exchange of lactate, glucose, free fatty acids (FFA) and glycerol were similar. During the 20 to 135 +/- 4 min period of exercise, LBF declined significantly in DE but tended to increase in control. Therefore, after 120 and 135 +/- 4 min of DE, LBF was 0.6 +/- 0.2 and 1.0 +/- 0.3 l min-1 lower (P < 0.05), respectively, compared with control. 4. The lower LBF after 2 h in DE did not alter glucose or FFA delivery compared with control. However, DE resulted in lower (P < 0.05) net FFA uptake and higher (P < 0.05) muscle glycogen utilisation (45 %), muscle lactate accumulation (4.6-fold) and net lactate release (52 %), without altering net glycerol release or net glucose uptake. 5. In both trials, the mean convective heat transfer from the exercising legs to the body core ranged from 6.3 +/- 1.7 to 7.2 +/- 1.3 kJ min-1, thereby accounting for 35-40 % of the estimated rate of heat production ( approximately 18 kJ min-1). 6. At exhaustion in DE, blood lactate values were low whereas blood glucose and muscle glycogen levels were still high. Exhaustion coincided with high body temperature ( approximately 40 C). 7. In conclusion, the present results demonstrate that reductions in exercising muscle blood flow with dehydration do not impair either the delivery of glucose and FFA or the removal of lactate during moderately intense prolonged exercise in the heat. However, dehydration during exercise in the heat elevates carbohydrate oxidation and lactate production. A major finding is that more than one-half of the metabolic heat liberated in the contracting leg muscles is dissipated directly to the surrounding environment. The present results indicate that hyperthermia, rather than altered metabolism, is the main factor underlying the early fatigue with dehydration during prolonged exercise in the heat.

Arterial O2 content and tension in regulation of cardiac output and leg blood flow during exercise in humans

[EN] A universal O2 sensor presumes that compensation for impaired O2 delivery is triggered by low O2 tension, but in humans, comparisons of compensatory responses to altered arterial O2 content (CaO2) or tension (PaO2) have not been reported. To directly compare cardiac output (QTOT) and leg blood flow (LBF) responses to a range of CaO2 and PaO2, seven healthy young men were studied during two-legged knee extension exercise with control hemoglobin concentration ([Hb] = 144.4 +/- 4 g/l) and at least 1 wk later after isovolemic hemodilution ([Hb] = 115 +/- 2 g/l). On each study day, subjects exercised twice at 30 W and on to voluntary exhaustion with an FIO2 of 0.21 or 0.11. The interventions resulted in two conditions with matched CaO2 but markedly different PaO2 (hypoxia and anemia) and two conditions with matched PaO2 and different CaO2 (hypoxia and anemia + hypoxia). PaO2 varied from 46 +/- 3 Torr in hypoxia to 95 +/- 3 Torr (range 37 to >100) in anemia (P < 0.001), yet LBF at exercise was nearly identical. However, as CaO2 dropped from 190 +/- 5 ml/l in control to 132 +/- 2 ml/l in anemia + hypoxia (P < 0.001), QTOT and LBF at 30 W rose to 12.8 +/- 0.8 and 7.2 +/- 0.3 l/min, respectively, values 23 and 47% above control (P < 0.01). Thus regulation of QTOT, LBF, and arterial O2 delivery to contracting intact human skeletal muscle is dependent for signaling primarily on CaO2, not PaO2. This finding suggests that factors related to CaO2 or [Hb] may play an important role in the regulation of blood flow during exercise in humans.

Muscle blood flow is reduced with dehydration during prolonged exercise in humans

[EN] 1. The present study examined whether the blood flow to exercising muscles becomes reduced when cardiac output and systemic vascular conductance decline with dehydration during prolonged exercise in the heat. A secondary aim was to determine whether the upward drift in oxygen consumption (VO2) during prolonged exercise is confined to the active muscles. 2. Seven euhydrated, endurance-trained cyclists performed two bicycle exercise trials in the heat (35 C; 40-50 % relative humidity; 61 +/- 2 % of maximal VO2), separated by 1 week. During the first trial (dehydration trial, DE), they bicycled until volitional exhaustion (135 +/- 4 min, mean +/- s.e.m.), while developing progressive dehydration and hyperthermia (3.9 +/- 0.3 % body weight loss; 39.7 +/- 0.2 C oesophageal temperature, Toes). In the second trial (control trial), they bicycled for the same period of time while maintaining euhydration by ingesting fluids and stabilizing Toes at 38.2 +/- 0.1 C after 30 min exercise. 3. In both trials, cardiac output, leg blood flow (LBF), vascular conductance and VO2 were similar after 20 min exercise. During the 20 min-exhaustion period of DE, cardiac output, LBF and systemic vascular conductance declined significantly (8-14 %; P < 0.05) yet muscle vascular conductance was unaltered. In contrast, during the same period of control, all these cardiovascular variables tended to increase. After 135 +/- 4 min of DE, the 2.0 +/- 0.6 l min-1 lower blood flow to the exercising legs accounted for approximately two-thirds of the reduction in cardiac output. Blood flow to the skin also declined markedly as forearm blood flow was 39 +/- 8 % (P < 0.05) lower in DE vs. control after 135 +/- 4 min. 4. In both trials, whole body VO2 and leg VO2 increased in parallel and were similar throughout exercise. The reduced leg blood flow in DE was accompanied by an even greater increase in femoral arterial-venous O2 (a-vO2) difference. 5. It is concluded that blood flow to the exercising muscles declines significantly with dehydration, due to a lowering in perfusion pressure and systemic blood flow rather than increased vasoconstriction. Furthermore, the progressive increase in oxygen consumption during exercise is confined to the exercising skeletal muscles.

Cardiovascular responses to dynamic exercise with acute anemia in humans

[EN] We hypothesized that reducing arterial O2 content (CaO2) by lowering the hemoglobin concentration ([Hb]) would result in a higher blood flow, as observed with a low PO2, and maintenance of O2 delivery. Seven young healthy men were studied twice, at rest and during two-legged submaximal and peak dynamic knee extensor exercise in a control condition (mean control [Hb] 144 g/l) and after 1-1.5 liters of whole blood had been withdrawn and replaced with albumin [mean drop in [Hb] 29 g/l (range 19-38 g/l); low [Hb]]. Limb blood flow (LBF) was higher (P < 0.01) with low [Hb] during submaximal exercise (i.e., at 30 W, LBF was 2.5 +/- 0.1 and 3.0 +/- 0.1 l/min for control [Hb] and low [Hb], respectively; P < 0.01), resulting in a maintained O2 delivery and O2 uptake for a given workload. However, at peak exercise, LBF was unaltered (6.5 +/- 0.4 and 6.6 +/- 0.6 l/min for control [Hb] and low [Hb], respectively), which resulted in an 18% reduction in O2 delivery (P < 0.01). This occurred despite peak cardiac output in neither condition reaching >75% of maximal cardiac output (approximately 26 l/min). It is concluded that a low CaO2 induces an elevation in submaximal muscle blood flow and that O2 delivery to contracting muscles is tightly regulated.

Hypoxia and the cardiovascular response to dynamic knee-extensor exercise

[EN] Hypoxia affects O2 transport and aerobic exercise capacity. In two previous studies, conflicting results have been reported regarding whether O2 delivery to the muscle is increased with hypoxia or whether there is a more efficient O2 extraction to allow for compensation of the decreased O2 availability at submaximal and maximal exercise. To reconcile this discrepancy, we measured limb blood flow (LBF), cardiac output, and O2 uptake during two-legged knee-extensor exercise in eight healthy young men. They completed studies at rest, at two submaximal workloads, and at peak effort under normoxia (inspired O2 fraction 0.21) and two levels of hypoxia (inspired O2 fractions 0.16 and 0.11). During submaximal exercise, LBF increased in hypoxia and compensated for the decrement in arterial O2 content. At peak effort, however, our subjects did not achieve a higher cardiac output or LBF. Thus O2 delivery was not maintained and peak power output and leg O2 uptake were reduced proportionately. These data are consistent then with the findings of an increased LBF to compensate for hypoxemia at submaximal exercise, but no such increase occurs at peak effort despite substantial cardiac capacity for an elevation in LBF.

Role of caloric content on gastric emptying in humans

[EN] 1. This study examined the effects of caloric content (caloric density and the nature of calories) on the rate of gastric emptying using the double-sampling gastric aspiration technique. Four test meals of 600 ml (glucose, 0.1 kcal ml-1; pea and whey peptide hydrolysates, both 0.2 kcal ml-1; milk protein, 0.7 kcal ml-1) were tested in six healthy subjects in random order on four separate occasions. 2. The glucose solution was emptied the fastest with a half-time of 9.4 +/- 1.2 min (P < 0.05) and the milk protein the slowest with a half-time of 26.4 +/- 10.0 min (P < 0.05); the pea peptide hydrolysate and whey peptide hydrolysate solutions had half-times of emptying of 16.3 +/- 5.4 and 17.2 +/- 6.1 min, respectively. The rates of gastric emptying for the peptide hydrolysate solutions derived from different protein sources were not different. 3. Despite the lower rate of gastric emptying for the milk protein solution, the rate of caloric delivery to the duodenum during the early phase of the gastric emptying process was higher than that for the other three solutions (46.3 +/- 6, 63.5 +/- 22, 62.5 +/- 19 and 113.8 +/- 25 cal min-1 kg-1 for the glucose, pea peptide hydrolysate, whey peptide hydrolysate and milk protein meals, respectively; P < 0.05). The caloric density of the test solutions was linearly related to the half-time of gastric emptying (r = 0.96, P < 0.05) as well as to the rate at which calories were delivered to the duodenum (r = 0.99, P < 0.001). 4. This study demonstrates that the rate of gastric emptying is a function of the caloric density of the ingested meal and that a linear relationship exists between these variables. Furthermore, the nature of the calories seems to play a minor role in determining the rate of gastric emptying in humans.

Sparql Interpreter : a component for Triskel architecture

[ES] SPARQL Interpreter es uno de los cinco componentes de la Arquitectura Triskel, una arquitectura de software para una base de datos NoSQL que intenta aportar una solución al problema de Big Data en la web semántica. Este componente da solución al problema de la comunicación entre el lenguaje y el motor, interpretando las consultas que se realicen contra el almacenamiento en lenguaje SPARQL y generando una estructura de datos que los componentes inferiores puedan leer y ejecutar.

Diseño e implementación de un convertidor numérico como aplicación Android

[ES] El Trabajo de Fin de Grado, Diseño e Implementación de un Convertidor Numérico como Aplicación Android es una aplicación desarrollada para terminales móviles con SO Android. Esta desarrollada en el entorno de desarrollo Eclipse, sobre el lenguaje de programación Java y hace uso de diferentes herramientas, las más importante de ellas el SDK TOOLS para Android. El objetivo o principal motivación por el cual he creado dicha aplicación es facilitarle al usuario final una forma más fácil y amena de acceder a toda la información proporcionada por el Servicio Web Números TIP. Esta aplicación podría tener gran uso en el campo de las enseñanzas primarias para enseñar a los niños a escribir números con letras y también podría ser de gran utilidad para las personas que no tengan un dominio extenso de nuestro idioma. La principal funcionalidad de la aplicación es realizar una consulta al Servicio Web Números TIP y luego mostrar por pantalla todos los datos devueltos. Todo el proceso de dibujo de la interfaz de usuario se realiza de manera dinámica y en tiempo de ejecución, logrando de esta manera adaptarnos a los datos que devuelva el servicio web. Para realizar la consulta al Servicio Web Números TIP el usuario introduce una ristra de caracteres sobre la cual se realizan determinadas comprobaciones en el servidor y se dibuja en la interfaz de usuario la respuesta devuelta. Esta ristra de caracteres puede contener cualquier signo, letra o número y el servicio web se encarga de devolver un error o reconocer un número, ya sea en su forma entera, fraccionaria, decimal o romana. La aplicación esta estandarizada para los cuatro tamaños generales reconocidos por Android y para sus densidades. Además se podría decir que la aplicación reconoce el idioma pre configurado en el teléfono y en base a ello solicita al servicio web las respuestas en español o en inglés.

A model and an algebra for semi-structured and full-text queries

The need for a convergence between semi-structured data management and Information Retrieval techniques is manifest to the scientific community. In order to fulfil this growing request, W3C has recently proposed XQuery Full Text, an IR-oriented extension of XQuery. However, the issue of query optimization requires the study of important properties like query equivalence and containment; to this aim, a formal representation of document and queries is needed. The goal of this thesis is to establish such formal background. We define a data model for XML documents and propose an algebra able to represent most of XQuery Full-Text expressions. We show how an XQuery Full-Text expression can be translated into an algebraic expression and how an algebraic expression can be optimized.

Data and behavioral contracts for web services

The recent trend in Web services is fostering a computing scenario where loosely coupled parties interact in a distributed and dynamic environment. Such interactions are sequences of xml messages and in order to assemble parties – either statically or dynamically – it is important to verify that the “contracts” of the parties are “compatible”. The Web Service Description Language (wsdl) is a standard used for describing one-way (asynchronous) and request/response (synchronous) interactions. Web Service Conversation Language extends wscl contracts by allowing the description of arbitrary, possibly cyclic sequences of exchanged messages between communicating parties. Unfortunately, neither wsdl nor wscl can effectively define a notion of compatibility, for the very simple reason that they do not provide any formal characterization of their contract languages. We define two contract languages for Web services. The first one is a data contract language and allow us to describe a Web service in terms of messages (xml documents) that can be sent or received. The second one is a behavioral contract language and allow us to give an abstract definition of the Web service conversation protocol. Both these languages are equipped with a sort of “sub-typing” relation and, therefore, they are suitable to be used for querying Web services repositories. In particular a query for a service compatible with a given contract may safely return services with “greater” contract.