Shoulder,elbow, and wrist joint angle excursions vary by gesture during touchscreen interaction

Repeated gesturing on touchscreen computing devices has become part of professional, personal, or school use by persons of all ages. Few studies have compared kinematics among joint motions and gestures during touchscreen interaction. We aimed to quantify the relative contributions of the shoulder, elbow and wrist to completion of several gestures to aid understanding of touchscreen ergonomics. Joint angles of the shoulder, elbow, and wrist were recorded for 22 seated participants while they interacted with a 10.1″ tablet computer held on an easel. Joint excursions at the shoulder, elbow, and wrist were all on average ≤20° during touchscreen interaction. The greatest excursion measured was shoulder rotation for swipe right with a mean of 15.5(±6.0)°. Index finger tap on a touchscreen was completed by participants with less than 5° of mean joint excursion at the shoulder, elbow and wrist. Tap, pinch and stretch gestures demonstrated significantly more wrist flexion/extension (p < 0.05) than shoulder flexion/extension, ab/adduction and rotation. Also, swipe left, right and up involved more shoulder rotation (p < 0.05) than wrist flexion/extension. These results suggest that when gestures are repeated frequently, the relative risk of overuse injury at the shoulder, elbow, or wrist may depend on the gesture being repeated.     

Peculiarities of Activation of Muscles of the Shoulder Belt in Voluntary Two-Joint Movements of the Upper Limb

We studied coordination of central motor commands (СMCs) coming to muscles of the shoulder and shoulder belt in the course of single-joint and two-joint movements including flexion and extension of the elbow and shoulder joints. Characteristics of rectified and averaged EMGs recorded from a few muscles of the upper limb were considered correlates of the CMC parameters. Special attention was paid to coordination of CMCs coming to two-joint muscles that are able to function as common flexors (m. biceps brachii, caput breve, BBcb) and common extensors (m. triceps brachii, caput longum, TBcl) of the elbow and shoulder joints. Upper limb movements used in the tests included planar shifts of the arm from one spatial point to another resulting from either simultaneous changes in the angles of the shoulder and elbow joints or isolated sequential (two-stage) changes in these joint angles. As was found, shoulder muscles providing movements of the elbow with changes in the angle of the elbow joint, i.e., BBcb and TBcl, were also intensely involved in the performance of single-joint movements in the shoulder joint. The CMCs coming to two-joint muscles in the course of two-joint movements appeared, in the first approximation, as sums of the commands received by these muscles in the course of corresponding single-joint movements in the elbow and shoulder joints. Therefore, if we interpret the isolated forearm movement performed due to a change in the angle of the elbow joint as the main motor event, while the shoulder movement is considered the accessory one, we can conclude that realization of a two-joint movement of the upper-limb distal part is based on superposition of CMCs related to basic movements (main and accessory). Neirofiziologiya/Neurophysiology, Vol. 41, No. 1, pp. 48–56, January–February, 2009.     

Kinetic chain of overarm throwing in terms of joint rotations revealed by induced acceleration analysis

This study investigated how baseball players generate large angular velocity at each joint by coordinating the joint torque and velocity-dependent torque during overarm throwing. Using a four-segment model (i.e., trunk, upper arm, forearm, and hand) that has 13 degrees of freedom, we conducted the induced acceleration analysis to determine the accelerations induced by these torques by multiplying the inverse of the system inertia matrix to the torque vectors. We found that the proximal joint motions (i.e., trunk forward motion, trunk leftward rotation, and shoulder internal rotation) were mainly accelerated by the joint torques at their own joints, whereas the distal joint motions (i.e., elbow extension and wrist flexion) were mainly accelerated by the velocity-dependent torques. We further examined which segment motion is the source of the velocity-dependent torque acting on the elbow and wrist accelerations. The results showed that the angular velocities of the trunk and upper arm produced the velocity-dependent torque for initial elbow extension acceleration. As a result, the elbow joint angular velocity increased, and concurrently, the forearm angular velocity relative to the ground also increased. The forearm angular velocity subsequently accelerated the elbow extension and wrist flexion. It also accelerated the shoulder internal rotation during the short period around the ball-release time. These results indicate that baseball players accelerate the distal elbow and wrist joint rotations by utilizing the velocity-dependent torque that is originally produced by the proximal trunk and shoulder joint torques in the early phase.     

System and modelling errors in motion analysis: implications for the measurement of the elbow angle in cricket bowling

The system and modelling errors of two fundamentally different motion capture systems (opto-reflective vs. video-based) were tested under various conditions, to determine their ability to accurately measure flexion-extension of the elbow angle in cricket bowling. A mechanical arm was used for all testing, that enabled known elbow flexion-extension and abduction ("carry") angles to be manually set. The root mean squared (RMS) error of 0.6 degrees for the opto-reflective system (Vicon-612) was more accurate in reconstructing a known angle than the video-based system (Peak Motus) (RMS error 2.3 degrees ) in the laboratory, when the same mathematical procedure (model) was applied to calculate the elbow flexion-extension angle. When different models were applied to the raw marker trajectories collected using the video-based system, RMS was lowest for the external marker segmental cluster models (2.3 degrees ) compared with 9.4 degrees for the vector and 4.5 degrees for the projected vector approaches, where joint centres were visually approximated. Real world, field-based comparisons using the video-based system showed that occluding the arm and therefore the shoulder, elbow and wrist joint centre locations by placing a shirt on the arm, increased RMS error for both vector (7.8 degrees -9.0 degrees ) and projected vector (4.3 degrees -5.1 degrees ) modelling approaches.     

Effects of the Racket Polar Moment of Inertia on Dominant Upper Limb Joint Moments during Tennis Serve

This study examined the effect of the polar moment of inertia of a tennis racket on upper limb loading in the serve. Eight amateur competition tennis players performed two sets of 10 serves using two rackets identical in mass, position of center of mass and moments of inertia other than the polar moment of inertia (0.00152 vs 0.00197 kg.m²). An eight-camera motion analysis system collected the 3D trajectories of 16 markers, located on the thorax, upper limbs and racket, from which shoulder, elbow and wrist net joint moments and powers were computed using inverse dynamics. During the cocking phase, increased racket polar moment of inertia was associated with significant increases in the peak shoulder extension and abduction moments, as well the peak elbow extension, valgus and supination moments. During the forward swing phase, peak wrist extension and radial deviation moments significantly increased with polar moment of inertia. During the follow-through phase, the peak shoulder adduction, elbow pronation and wrist external rotation moments displayed a significant inverse relationship with polar moment of inertia. During the forward swing, the magnitudes of negative joint power at the elbow and wrist were significantly larger when players served using the racket with a higher polar moment of inertia. Although a larger polar of inertia allows players to better tolerate off-center impacts, it also appears to place additional loads on the upper extremity when serving and may therefore increase injury risk in tennis players.     

The contribution of the wrist, elbow and shoulder joints to single-finger tapping

We aimed to determine the role of the wrist, elbow and shoulder joints to single-finger tapping. Six human subjects tapped with their index finger at a rate of 3 taps/s on a keyswitch across five conditions, one freestyle (FS) and four instructed tapping strategies. The four instructed conditions were to tap on a keyswitch using the finger joint only (FO), the wrist joint only (WO), the elbow joint only (EO), and the shoulder joint only (SO). A single-axis force plate measured the fingertip force. An infra-red active-marker three-dimensional motion analysis system measured the movement of the fingertip, hand, forearm, upper arm and trunk. Inverse dynamics estimated joint torques for the metacarpal-phalangeal (MCP), wrist, elbow, and shoulder joints. For FS tapping 27%, 56%, and 18% of the vertical fingertip movement were a result of flexion of the MCP joint and wrist joint and extension of the elbow joint, respectively. During the FS movements the net joint powers between the MCP, wrist and elbow were positively correlated (correlation coefficients between 0.46 and 0.76) suggesting synergistic efforts. For the instructed tapping strategies (FO, WO, EO, and SO), correlations decreased to values below 0.35 suggesting relatively independent control of the different joints. For FS tapping, the kinematic and kinetic data indicate that the wrist and elbow contribute significantly, working in synergy with the finger joints to create the fingertip tapping task.     

Coordination dynamics of trajectory formation

 The present study aims to understand the neurally based coordination dynamics (multistability, loss of stability, transitions, etc.) of trajectory formation in a simple task. Six subjects produced two spatial patterns of coordination in the xy plane by alternating the abduction-adduction and flexion-extension motions of their right index finger. Each pattern was characterized by a unique temporal ratio between the x and y directions of motion: (1) a figure zero, a 1:1 temporal pattern; and (2) a figure eight, a 2:1 temporal pattern. The patterns were produced rhythmically and movement frequency was scaled across ten frequency plateaus, with ten cycles of motion per step. As movement frequency increased, switching from a figure eight to a figure zero was observed at critical cycling frequencies. The switch from pattern (2) to pattern (1) was identified in the spatial trajectory and power spectra of x(t) and y(t). En route to the transition, enhancement of fluctuations was observed in the Fourier amplitudes of x(t) and y(t), specifically at f ₀ (the metronome frequency) and 2f ₀ (the first harmonic of f ₀). Interestingly, there was no difference in the spatial variability of the two patterns. Overall, the data demonstrate that spatial patterns of coordination can be characterized in terms of the temporal relationship between the spatial components of the trajectory itself. We discuss the experimental findings in relation to other end-point planning and multijoint control strategies, as well as the much more general problem of temporal synchronization in many interlimb and intralimb coordination tasks. Received: 24 February 1995/Accepted in revised form: 8 August 1995     

Investigating the effects of opioid drugs on electrocortical activity using wavelet transform

 Fetal electrocortical activity (ECoG) is characterized by two distinct patterns: HVSA (high voltage, slow activity) and LVFA (low voltage, fast activity). Using the wavelet transform (WT), we recently reported that the frequency characteristics of these two ECoG patterns undergo significant maturational changes prior to birth (Akay et al. 1994a). We now report that fetal ECoG can also be significantly affected by pharmacological agents. In this paper, we compared the effects of two opioid drugs (morphine and [D-Pen², D-Pen⁵]enkephalin, DPDPE) on fetal ECoG, using the chronically instrumented fetal lamb model. Morphine was infused intravenously (i.v.) at 2.5 mg/h, while DPDPE was infused into the lateral cerebroventricle (i.c.v.) at 30 μg/h. The ECoG was analyzed using WT. We performed multiresolution decomposition for four sets of parameters D _{2 j} where −1<j<−4. The four series WTs represent the detail signal bandwidths: (1) 16–32 Hz, (2) 8–16 Hz, (3) 4–8 Hz, (4) 2–4 Hz. The data were subjected to statistical analysis using the Kolmogorov–Smirnov (KS) test. Both morphine and DPDPE resulted in a significant increase in power in the first wavelet band, while power was reduced in the second, third and fourth wavelet bands. In addition, both drugs resulted in a disruption of the normal cyclic pattern between the two ECoG patterns. There was a difference in the time course of action between morphine and DPDPE. This is the first occasion in which continuous ECoG has been subjected to rigorous statistical analysis. The results suggest that the WT-KS method is most suitable for quantitating changes in the ECoG induced by pharmacological agents. Received: 21 January 1994/Accepted in revised form: 15 September 1994     

Morphology and morphometrics of the appendicular musculature in geckoes with different locomotor habits (Lepidosauria)

 In this study a ground-dwelling (Eublepharis macularius) and a highly specialised climbing (Gekko gecko) lizard were chosen as study objects. The fore- and hindlimbs of two individuals of each species were dissected, and muscle masses, mean fibre lengths, cross-sectional areas and moment arms were determined. Special attention was paid to general muscle architecture (origin, insertion, fibre orientation, etc.) and pennation angles. Using these variables (cross sectional areas and moment arms), maximal moments exertable across the shoulder/hip, elbow/knee and wrist/ankle were calculated for both species. In accordance with the biomechanical predictions related to the preferred locomotor substrate of each species (i.e. level running for E. macularius and climbing for G. gecko), the results of this study indicate that climbers such as G. gecko generally possess powerful retractor muscles crossing the shoulder and hip joints. Additionally, the specialised climber is able to exert higher flexion moments across the elbow, which prevents the animals from falling backwards. However, G. gecko appears to be constrained in its ankle extension capabilities by the presence of the adhesive toe pads. The level-running species, on the other hand, shows a relatively stronger development of the extensor muscles in the lower limbs, allowing these lizards to run in an erect posture. In general, both species show large similarities on a gross morphological level as expected when considering their phylogenetic relatedness. Adaptations to their preferred locomotor substrate only become apparent when considering the functional properties (i.e. joint moments) of the appendicular musculature. Accepted: 30 November 1998     

Upper limb kinematics and the role of the wrist during stone tool production

Past studies have hypothesized that aspects of hominin upper limb morphology are linked to the ability to produce stone tools. However, we lack the data on upper limb motions needed to evaluate the biomechanical context of stone tool production. This study seeks to better understand the biomechanics of stone tool‐making by investigating upper limb joint kinematics, focusing on the role of the wrist joint, during simple flake production. We test the hypotheses, based on studies of other upper limb activities (e.g., throwing), that upper limb movements will occur in a proximal‐to‐distal sequence, culminating in rapid wrist flexion just prior to strike. Data were captured from four amateur knappers during simple flake production using a VICON motion analysis system (50 Hz). Results show that subjects utilized a proximal‐to‐distal joint sequence and disassociated the shoulder joint from the elbow and wrist joints, suggesting a shared strategy employed in other contexts (e.g., throwing) to increase target accuracy. The knapping strategy included moving the wrist into peak extension (subject peak grand mean = 47.3°) at the beginning of the downswing phase, which facilitated rapid wrist flexion and accelerated the hammerstone toward the nodule. This sequence resulted in the production of significantly more mechanical work, and therefore greater strike forces, than would otherwise be produced. Together these results represent a strategy for increasing knapping efficiency in Homo sapiens and point to aspects of skeletal anatomy that might be examined to assess potential knapping ability and efficiency in fossil hominin taxa. Am J Phys Anthropol 143:134‐145, 2010. © 2010 Wiley‐Liss, Inc.     

Relationships between the chromosomes of Aegilops umbellulata and wheat

 A comparative genetic map of Aegilops umbellulata with wheat was constructed using RFLP probes that detect homoeoloci previously mapped in hexaploid bread wheat. All seven Ae. umbellulata chromosomes display one or more rearrangements relative to wheat. These structural changes are consistent with the sub-terminal morphology of chromosomes 2 U, 3 U, 6 U and 7 U. Comparison of the chromosomal locations assigned by mapping and those obtained by hybridization to wheat/Ae. umbellulata single chromosome addition lines verified the composition of the added Ae. umbellulata chromosomes and indicated that no further cytological rearrangements had taken place during the production of the alien-wheat aneuploid lines. Relationships between Ae. umbellulata and wheat chromosomes were confirmed, based on homoeology of the centromeric regions, for 1 U, 2 U, 3 U, 5 U and 7 U. However, homoeology of the centromeric regions of 4 U with wheat group-6 chromosomes and of 6 U with wheat group-4 chromosomes was also confirmed, suggesting that a re-naming of these chromosomes may be pertinent. The consequences of the rearrangements of the Ae. umbellulata genome relative to wheat for gene introgression are discussed. Received: 10 July 1997 / Accepted: 19 September 1997     

Kinetic coefficients for simultaneous reduction of sulfate and uranium by Desulfovibrio desulfuricans

 Previously it was demonstrated that bacteria are capable of transforming soluble uranyl ion, U(VI), to insoluble uraninite, U(IV); however, the rate for this transformation has not been determined. We report the kinetic coefficients for Desulfovibrio desulfuricans DSM 1924 grown in a continuous-flow chemostat where pyruvate was the electron donor and sulfate was the electron acceptor. The medium was supplemented with 1 mM uranyl nitrate, and the chemostat flow rate ranged from 1.12 ml/h to 4.75 ml/h with incubation at 28°C. The maximum rate of pyruvate utilization (k) was determined to be 4.7 days^-1, while the half-velocity constant (K _s) was 127 mg/l. The yield coefficient (Y) of cells per mole of pyruvate oxidized was calculated to be 0.021 g, while the endogenous decay coefficient (k _d) was determined to be 0.072 days^-1. More than 90% of U(VI) was transformed to U(VI) in the chemostat under the conditions employed. Received: 7 September 1995/Received last revision: 10 January 1996/Accepted: 5 February 1996     

Tropical zooplankton in the highly-enclosed lagoon of Taiaro Atoll (Tuamotu Archipelago, French Polynesia)

 Nocturnal zooplankton assemblages around Taiaro Atoll were sampled over six nights during February 1994. Replicate zooplankton samples were collected at windward and leeward locations in the enclosed lagoon and adjacent ocean with a metered net (85 cm diameter, 500 μm mesh) towed for 15 min at 5 m depth. The zooplankton community in the lagoon was very different from that in the ocean. Oceanic samples contained 50 mostly holoplanktonic taxa (diversity index, H′=2.62; evenness index, J′=0.67). Lagoonal samples contained 19 mostly meroplanktonic taxa (H′=1.54, J′=0.52) with three taxa (decapod larvae; an ostracod, Cypridina sp.; a copepod, Acartia fossae) contributing >90% of the individuals. Unlike the ocean, zooplankton distributions in the lagoon were not homogenous; instead spatial patterns were apparently formed by the interaction between hydrodynamic processes and species-specific behaviour. Accepted: 28 August 1997     

A dynamical model for reflex activated head movements in the horizontal plane

 We present a controls systems model of horizontal-plane head movements during perturbations of the trunk, which for the first time interfaces a model of the human head with neural feedback controllers representing the vestibulocollic (VCR) and the cervicocollic (CCR) reflexes. This model is homeomorphic such that model structure and parameters are drawn directly from anthropomorphic, biomechanical and physiological studies. Using control theory we analyzed the system model in the time and frequency domains, simulating neck movement responses to input perturbations of the trunk. Without reflex control, the head and neck system produced a second-order underdamped response with a 5.2 dB resonant peak at 2.1 Hz. Adding the CCR component to the system dampened the response by approximately 7%. Adding the VCR component dampened head oscillations by 75%. The VCR also improved low-frequency compensation by increasing the gain and phase lag, creating a phase minimum at 0.1 Hz and a phase peak at 1.1 Hz. Combining all three components (mechanics, VCR and CCR) linearly in the head and neck system reduced the amplitude of the resonant peak to 1.1 dB and increased the resonant frequency to 2.9 Hz. The closed loop results closely fit human data, and explain quantitatively the characteristic phase peak often observed. Received: 15 April 1996 / Accepted in revised form: 1 July 1996     

Genetic variation for in vitro sesame pollen germination and tube growth

  In vitro pollen germination and tube length studies are valuable in elucidating mechanisms (germination capacity and rate, tube growth rate) possibly associated with genetic differences in male transmission. On each of two collection dates, the percentage germination and tube length of the binucleate pollen grains from five diverse sesame (Sesamum indicum L.) genotypes were determined at eight times (30, 60, 90, 120, 150, 180, 240, 300 min) after inoculation on a semisolid medium containing 10% (100 g l^-1) sucrose (C₁₂H₂₂O₁₁), 0.4% (4 g l^-1) purified agar (Fisher Lot 914409), 0.1% (1 g l^-1) calcium nitrate [Ca(NO₃)₂ ⋅ 4H₂O] and 0.01% (100 mg l^-1) boric acid (H₃BO₃). Before heating, the pH of the medium was adjusted to 7.0 with a 0.1 N potassium hydroxide (KOH) solution. Over the five genotypes, 5% germination was found 30 min after inoculation and a maximum of 37% germination 120 min after inoculation with no significant changes thereafter. As indicated by the highly significant genotype×time after inoculation interaction, the genotypes differed in the time at which germination was initiated and maximum germination attained. Over all five genotypes, the tube length was 91 μm 30 min after inoculation, reaching a maximum of 1000 μm 300 min after inoculation. As shown by the highly significant genotype×time after inoculation interaction, the genotypes differed in the time at which tube length was observed and the maximum tube length was attained. Little or no relationship between percent germination and tube length was observed among the genotypes. For both percent germination and tube length, the statistical significance of collection date and its interactions with genotype and time after inoculation indicated that environment in the form of collection date was also an influencing factor. These results indicated that genetic differences among genotypes were present for in vitro germination capacity, germination rate and tube growth rate and that these factors singly or in combination could alter male transmission of genetic elements. Received: 5 February 1997 / Accepted: 23 June 1997     

Peculiarities of Activation of the Shoulder Belt and Shoulder Muscles in Generation of Different-Direction Isometric Efforts by the Forearm

We studied central motor commands, CMCs, coming to the muscles that flex and extend the shoulder and elbow joints in the course of generation of voluntary isometric efforts of different directions by the forearm; the efforts were initiated according to a visual signal. Amplitudes of EMGs recorded from the muscles of the shoulder belt and shoulder and subjected to full-wave rectification and low-frequency filtration were considered correlates of the CMC intensity. An effort of the preset direction was developed within the operational space of the horizontal plane with angles 30 deg in the shoulder joint (external angle with respect to the frontal plane) and 90 deg in the elbow joint. We plotted sector diagrams of the logarithmic coefficient of the intensity increment of EMGs of the above muscles for the entire set of directions of generated efforts with a 15- or 20-deg step. Orientations of the maxima of EMG activity of the given muscles were rather close to the directions of the maxima of the force moments generated by these muscles. In most cases, a shift of the direction by one gradation with respect to the EMG maximum in the respective muscle resulted in a significant decrease in the level of EMG activity. It is shown that preferential activation of the muscles agonistic with respect to the examined direction of the generated effort was, as a rule, accompanied by coactivation of the antagonist muscles. When “two-joint” isometric efforts are formed, realization of the socalled synergic muscle tasks (where prevailing contractions of the muscles of the same functional direction for both joints coincide, i.e., flexion-flexion or extension-extension) is organized in a simpler manner. The programs of “nonsynergic” contractions (flexion of one joint and extension of another one, or vice versa) are more complex. In different subjects, considerably dissimilar patterns of EMG activity in muscles influencing these joints could be observed.     

Use of an industrial effluent as a carbon source for denitrification of a high-strength wastewater

Denitrification of a high-strength synthetic wastewater (150 g NO^- ₃ l^-1) was carried out using a wine distillery effluent as an example of an industrial carbon source (22.7 g chemical oxygen demand l^-1). Two configurations were tested: one consisted of an acidogenesis reactor followed by a denitrifying reactor and the other was a single reactor directly fed with the raw effluents. In both cases, denitrification was achieved at a nitrate load of 9.54 g NO^- ₃ l^-1 day^-1 (2.19 g N as NO^- ₃ l^-1 day^-1) with good specific reduction rates: 32.6 mg and 35.2 mg N as NO _x  g volatile suspended solids h^-1, calculated on a single day, for the two-step and the one-step process respectively. Dissimilatory nitrate reduction to ammonium did not occur, even in the one-step process. Received: 26 October 1995/Received revision: 15 February 1996/Accepted: 20 February 1996     

CMS sources in sunflower: different origin but same mechanism?

 The presence of orfH522, orfH708 and orfH873 in the mtDNA, as well as the expression of mitochondrially encoded proteins, were investigated for 28 sources of cytoplasmic male sterility (CMS) and HA89, a fertile line of Helianthus annuus. The whole 5-kb insertion, found in PET1, is also present in all PET1-like CMS sources. However, with regard to the 11-kb inversion ANO1 demonstrated a different organization at the cob locus from the other PET1-like CMS sources. Only orfH873 gave hybridization patterns in all investigated cytoplasms. For the fertile cytoplasm, as well as ANN4, ANN5, ANL1, ANL2, ARG2 and MAX1, hybridizations obtained with orfH708 were highly polymorphic. Hybridization signals with orfH522 were only detectable in the PET1-like CMS sources and MAX1. Comparing the mitochondrially encoded proteins of the CMS sources characteristic patterns could be detected for seven cytoplasms in addition to the PET1-like CMS sources expressing the 16-kDa protein. For ANN1 and ANN3 three CMS-associated proteins of 16.3 kDa, 16.9 kDa and 34.0 kDa could be identified among the in organello translation products. Also ANT1 expressed three additional proteins of 13.4 kDa, 17.8 kDa and 19.7 kDa, respectively. In ARG3 and RIG1 one protein of 17.5 kDa was missing and instead a new protein of 16.9 kDa appeared. In addition, in GIG1 and PET2 a unique protein of 12.4 kDa could be identified. These results indicate that certain types of cytoplasmic male sterility are preferentially present in sunflower. Received: 15 June 1998 / Accepted: 13 July 1998     

Selection of an adhesive phenotype of Streptococcus salivarius subsp. thermophilus for use in fixed-bed reactors

 Streptococcus salivarius subsp. thermophilus was cultivated in a chemostat in order to obtain an adhesive phenotype of this strain. When the system was operated at low dilution rates (D<0.2 h^-1) for about 4 weeks, the strain formed a visible film on the surface of the culture vessel. The biofilm cells were not washed out even when dilution rates were increased (D=6.9 h^-1), and this resulted in a high biomass productivity (P=4.1 g l^-1h^-1). On the other hand, when the culture was grown at dilution rates faster than 0.2 h^-1, only the free suspended cells were present in the culture broth, and were washed out at velocities of about 1.0 h^-1. The biomass productivity was consequently lower (P=1.33 g l^-1h^-1) than in the previous case. The selected adhesive phenotype was grown on different glass beads and the possibility of lactate fermentation in a continuous and semicontinuous mode was demonstrated. Received: 16 August 1995/Received revision: 18 March 1996/Accepted: 25 March 1996