Ontogenetic effects on aerobic and anaerobic metabolism during jumping in the American locust, Schistocerca americana

In vertebrate locomotion, older animals have both increased power outputs(lizards – Garland,1984; carp – Wakeling et al., 1999; humans – Doré et al., 2000) and greater endurance (lizards – Garland, 1984; Garland and Else, 1987; snakes and anurans – Pough, 1977, 1978; Pough and Kamel, 1984). These vertebrate locomotory performance improvements with age appear to be facilitated by increased aerobic (squirrels – Chappell and Bachman, 1995;sparrows – Chappell et al.,1999) and anaerobic metabolic rates (fish – Somero and Childress, 1980;humans – Taylor et al.,1997). For insects, some authors have suggested that larger sizes may cause increased problems with oxygen delivery due to slower diffusion rates through longer tracheae (Graham et al., 1995). This raises the question, do insects show improvements in locomotory performance and metabolic rates with age, as seen in vertebrates, or do gas exchange constraints produce a different pattern in tracheated organisms?

Among invertebrates, developmental effects on locomotory performance have been best studied in jumping grasshoppers. During ontogeny, the power output during single, maximal jumps increased sixfold in Schistocerca gregaria grasshoppers (Gabriel,1985a; Queathem,1991; Katz and Gosline,1993). This increase in single jump power output was first attributed to a greater proportion of jumping muscle mass and improved energy storage in the exoskeleton of the femur (Gabriel, 1985a,b). However, a second study showed that adults have stiffer cuticular springs but may have the same proportion of muscle as juveniles, leaving the mechanism of increased power output of maximal jumps during development unresolved(Katz and Gosline, 1993). No studies to date have examined the effect of ontogeny on sustained locomotory performance in grasshoppers or other insects.

Ontogenic effects on aerobic and anaerobic metabolism of active insects are also not known. Adult grasshoppers produce lactate at relatively high rates during jumping (Zebe and McShan,1957; Gade, 1984; Harrison et al., 1991). If juveniles do not produce lactate during jumping, this could at least partially explain the increase in single-jump power output in grasshoppers, and such a pattern would be consistent with the suggestion that larger insects experience more problems with oxygen delivery (Graham et al., 1995). Resting mass-specific metabolic rates fall with age in grasshoppers, but the safety margin for oxygen delivery increases,suggesting an improved oxygen delivery system in larger/older grasshoppers(Greenlee and Harrison, 2004). Morphological measures indicate that mitochondrial content and tracheal oxygen delivery capacities within the jumping leg of grasshoppers increase strongly with age (Hartung et al.,2004). Thus, the data to date suggest that larger/older grasshoppers do not experience increased problems with oxygen delivery that might reduce locomotory performance. In the present study, we used a wide range of developmental stages (2nd, 4th, 6th instars and adults), representing a 30-fold increase in body mass, to determine how ontogeny affected power output, endurance, lactate concentrations, oxygen consumption and carbon dioxide production during repeated jumping in S. americana.

If older/larger grasshoppers experience more problems with oxygen delivery due to longer tracheae, as suggested by Graham et al.(1995), then their locomotion may be more oxygen sensitive than juveniles. For example, if older/larger grasshoppers experience oxygen limitations during jumping, we would predict that their jump performance would be stimulated by hyperoxia, and strongly inhibited by hypoxia; if smaller/younger grasshoppers have excessive oxygen delivery capacities, their locomotory performance should be relatively oxygen insensitive. As noted above, evidence from resting grasshoppers suggests that older S. americana grasshoppers have improved oxygen delivery relative to younger animals, probably due to improved convective ventilation(Greenlee and Harrison, 2004). However, metabolic rates and critical oxygen partial pressure(P_O2) values for insects during flight are much higher than measured for insects at rest(Chadwick and Gilmour, 1940; Davis and Fraenkel, 1940; Joos et al., 1997; Harrison and Lighton, 1998; Greenlee and Harrison, 2004). If body size imposes limits on tracheal oxygen delivery, these are most likely to be evident during locomotion, when gas exchange requirements are high.

S. americana Drury were reared from eggs in culture at Arizona State University as previously described(Harrison and Kennedy, 1994). To determine a grasshopper's age in days within an instar, the thorax of newly molted individuals was marked with a unique color of Testors acrylic paint(Rockford, IL, USA). The paint was shed with the exoskeleton during molting,so it was necessary to re-paint the grasshoppers at each instar. We used animals from the latter two-thirds to three-quarters of an instar because maximal jump performance was reduced during the initial days after molting(Queathem and Full, 1995). Sex could not be distinguished in the 2nd and 4th instars, but in both 6th instars and adults only the larger female grasshoppers were used to maximize the body size contrast across instars. We utilized female adults up to day 21, because they had not yet laid egg pods that dramatically alter body mass (S.D.K., K. J. Greenlee and J.F.H., unpublished).

We measured body mass and femur mass (±0.1 mg) using a Mettler Analytical AE 240 Dual Range Balance (Hightstown, NJ, USA). Femur length was measured to the nearest 0.01 mm with a Mitutoyo Digimatic CD-6 digital micrometer (Kawasaki, Japan). We then sliced the femur longitudinally in half and placed it into 0.35 mol l^–1 NaOH for approximately 24 h for tissue digestion (Marden,1988). After the tissue was removed, the femoral exoskeleton was washed in distilled water, blotted dry and re-weighed. The femoral exoskeletons of 6th instars and adults were weighed on the analytical balance and those of the 2nd and 4th instars were weighed on a Cahn C-33 microbalance(±0.01 mg; Cerritos, CA, USA). Wet tissue mass was calculated as the difference between the exoskeleton mass and wet femur mass. We calculated the extensor tibia muscle mass to be 66% of the measured wet tissue mass(Hartung et al., 2004).

Grasshoppers were removed from the colony on the day of the experiment and kept in a group with food at 35°C. Approximately 1 h before a jumping trial, we weighed the individual to be jumped and removed the distal third of the adult wings to prevent flight. Grasshoppers were encouraged to jump by physical prodding for up to 5 min or until fatigued, defined as a 30-s pause between subsequent jumps with continued prodding. The method of physical stimulation was varied throughout the trial to motivate the grasshopper. Grasshoppers were jumped on a cotton bed-sheet (310×80 cm) that was divided into a 10 cm (adult) or 5 cm (juvenile) numbered grid system within a 35°C temperature controlled room. As grasshoppers jumped between squares during the trial, the grid number was called out and recorded onto an audiotape. Distances were measured from the center of each square.

Grasshoppers similar in age to the ones used in the jumping performance experiment were jumped in a 100 liter Plexiglas gloved box at 35°C for 0,15, 30, 60, 120, 300 or 600 s (N=8–10 at each time). After jumping, animals were frozen in liquid nitrogen and stored at –20°C. Later, the body mass, femur mass, femur length and lactate concentrations were measured.

Frozen femurs were diluted 9× with chilled 0.6 mol l^–1 perchloric acid and pulverized with a ground-glass-tissue homogenizer immersed in ice-water. After centrifugation (Beckman Microfuge E,Palo Alto, CA, USA) for 5 min at 15 850 g, the supernatant was removed and diluted 9× with a 25 mmol l^–12-amino-2-methyl-1-propanol buffer. Lactate was measured in the supernatant using a fluorometric assay (Passonneau and Lowry, 1993) with a Standard Curve Filter Fluorometer (Optical Technology Devices Inc., Elmsford, NY, USA). Each sample was run six times [three with lactate dehydrogenase (LDH) and three without LDH to account for background noise].

Gas exchange for jumping S. americana was measured using Plexiglas chambers for all ages except the 2nd instars, which were jumped in a portion of a 60 cm³ syringe. Each age group (N=7 per group) was tested using a different-sized chamber and flow rates so that the 95%equilibrium time was approximately 45 s in each case(Lasiewski et al., 1966): 2nd instars, 10 ml chamber, 35.0 ml min^–1 flow rate; 4th instars,39.5 ml chamber, 138.4 ml min^–1; 6th instars, 116.4 ml chamber, 377.8 ml min^–1; adults, 246.3 ml chamber, 563.1 ml min^–1. Each jumping chamber was air tight except for incurrent and excurrent air-flow ports. The smooth bottom surface of the chamber was covered with coarse sand paper to facilitate jumping. Grasshoppers were encouraged to jump using the end of a small paint brush attached to a wire that was inserted through a rubber septum.

To reduce fluctuations in the background oxygen level, dry,CO₂-free air (Balston purge-gas generator; Havervill, MA, USA) was generated and stored in a gas cylinder, which was warmed with heating pads to improve mixing. The purge gas generator was then disconnected, and air flow from the cylinder to the respirometry system was controlled by a Brooks 5878 mass flow controller and three 5850i Brooks mass flow meters (Brooks Instruments, Hatfield, PA, USA). One meter controlled air flow to the respirometry chamber and the grasshopper, the second meter was used to flush the respirometry chamber between trials, and the third meter controlled a line that flowed to the reference channel of the gas analyzers(Fig. 1). Using three-way valves, we could switch between recording either the baseline O₂and CO₂ levels (while the animal chamber was being flushed) or the grasshopper's gas exchange.

Both the reference stream and the excurrent stream from the jumping chamber were dried using MgClO₄ (except for the 2nd instars, for whom this step was omitted to improve temporal resolution). Next, the air in each line emptied into a syringe open to the atmosphere, from which air was sub-sampled(at 10 ml min^–1 for 2nd instars and 50 ml min^–1 for the other ages) using a Li-Cor pump (Li-Cor,Lincoln, NE, USA) and pulled through a Li-6252 Li-Cor CO₂ analyzer and then an Oxilla O₂ analyzer (Sable Systems, Las Vegas, NV, USA). Both gas analyzers were used in the differential mode to improve resolution. CO₂ and water were removed from each stream between the analyzers using Ascarite® and Drierite®, respectively. The flow rates through the analyzers were controlled by FMA-A2405 Omega mass flow controllers(0–500 ml min^–1; Stamford, CT, USA) located between the O₂ analyzer and the Li-Cor pump. The output of both gas analyzers was digitized and recorded (Sable Systems).

After recording a starting baseline CO₂ and O₂ level for 5 min, the three-way valves were switched to record the animal's resting CO₂ production and O₂ consumption for 10 min. Next, the jumping metabolic rate was measured for 5 min by turning on a light in the incubator and stimulating the grasshopper to jump with the brush. After 5 min,the light was turned off and the grasshopper's whole-body CO₂production and O₂ consumption during recovery were measured for approximately 10 min before switching back to the bypass and recording baseline values for the last 5 min of the trial. In control tests, there was no evidence of leaks when switching between the different experimental streams nor when manipulating the brush in the chamber. The signal-to-noise resolution of the respirometry system ranged from 12 (resting 2nd instars) to 95 (jumping adults) for O₂, and from 337 (resting 2nd instars) to 1554 (jumping adults) for CO₂. Immediately following the trial, the animal was weighed on the analytical balance (±0.1 mg).

To quantify how the aerobic ATP production of the jumping muscles may change during ontogeny, we calculated the locomotory rate of oxygen uptake(V̇_O2) of the jumping muscle by comparing the V̇_O2 values for animals stimulated to locomote before and after autotomizing their jumping legs. Each grasshopper was encouraged to autotomatize its hind femurs by gently rubbing the femoral joints with warmed scissors. Autotomy caused minimal damage to the animal because there are no direct muscles in the autotomical plane, and after the leg is detached a membrane closes the wound,minimizing haemolymph bleeding (Arbas and Weidner, 1991). Also, the thoracic muscles responsible for moving the femur do not show denervation for at least 3 days(Personius and Arbas, 1998) or muscular atrophy for 7 days after the autotomy(Arbas and Weidner, 1991). The hindleg-less grasshopper was returned to a 35°C container with food and water. The metathoracic hind femurs were weighed on an analytical balance, and extensor tibia muscle mass was calculated as previously described. O₂ consumption and CO₂ emission were re-measured exactly as described above for juveniles on the following day and for adults after 2 days.

Injections of room air were used for each chamber and flow rate to correct for lag time from the chamber to the analyzers (approximately 1 min) and between analyzers. The rates of whole-body O₂ consumption and CO₂ emission(Ṁ_O2,wb and Ṁ_CO2,wb,respectively; μmol g^–1 h^–1) were calculated similarly to Greenlee and Harrison(2004) and converted from ml g^–1 min^–1 to μmol g^–1h^–1. There were no statistically significant differences in jump frequencies between the grasshoppers forced to jump in the respiratory chamber and those jumped in the temperature-controlled room, so these values are not reported.

Oxygen sensitivity during jumping was measured in a 100 liter Plexiglas gloved box at 35°C. The gloved box was perfused at 6 l min^–1 with artificial oxygen atmospheres (5 kPa, 12 kPa, 21 kPa or 45 kPa – balance nitrogen) created by mixing oxygen and nitrogen supplied from compressed tanks via needle valves with mixtures confirmed using an S-3A/I AEI Technologies Oxygen Analyzer (Pittsburgh, PA,USA). Within the gloved box, a fan promoted air circulation, a thermocouple recorded air temperature, and the bottom was covered with a soft paper to provide traction for the jumping grasshoppers. Ṗ_O2 did not vary from the reported mix in the empty gloved box.

Individual grasshoppers were placed in the adjoining airlock for 30 s to equilibrate to the oxygen tension before being released into the chamber. Grasshoppers were forced to jump for up to 10 min or until fatigued (defined as when 30 s passed between jumps). Each animal was tested in a single oxygen atmosphere (N=5–8 per oxygen tension).

The effects of age and time during the trial on jump performance data were analyzed using repeated-measures analysis of variance (ANOVA) with Systat 10.2(Systat Software Inc., Richmond, CA, USA). We utilized ANOVA with Bonferroni-corrected post-hoc comparisons to compare the lactate concentrations among different aged grasshoppers at specific minutes during the trial (Sokal and Rohlf,1995). The Ṁ_O2, Ṁ_CO2 and oxygen sensitivity data were analyzed using ANOVA with Bonferroni-corrected post-hoc comparisons. In all cases, our level of significance was 0.05. Unless otherwise noted, all reported values are the means ±standard errors.

During development in S. americana, femur mass and muscle mass grew isometrically relative to body mass(Table 1). In contrast to muscle development, femurs from larger animals were proportionally longer and had a greater amount of exoskeleton (Table 1). For the grasshoppers used in the jumping experiment, body mass increased ∼30-fold as animals developed from the 2nd instars to adults(2nd instars, 60±1.6 mg; 4th instars, 262±6.5 mg; 6th instars,1219±37.3 mg; adults, 1782±62.0 mg).

Jump performance was significantly affected by age. The mean number of jumps per minute declined more rapidly in older animals(Fig. 2; repeated-measures ANOVA, instar × time, F_12,227=3.7, P<0.001). The body mass-specific power output decreased during the trial regardless of age; however, power output decreased more rapidly in older animals (Fig. 3;repeated-measures ANOVA, instar × time, F_12,227=13.2, P<0.001).

Lactate concentrations in the extensor tibia muscle increased across developmental stages and with time spent jumping(Fig. 4). There was no detectable lactate in the muscles of 2nd instars during the trial (assay detection limit was 0.3 mmol kg^–1). Muscle-specific lactate levels measured at time 0 increased with age, rising from 1.7±0.3 mmol kg^–1 in 4th instars to 3.1±0.2 mmol kg^–1 in 6th instars and 3.7±0.8 mmol kg^–1 in adults (Fig. 4; ANOVA, F_3,28=13.4, P<0.001). During 2 min of jumping, muscle lactate increased more rapidly in older animals (Fig. 4; ANOVA, F_3,33=157.7, P<0.001). After 2 min of jumping,lactate levels reached a peak for each age group (6.0±0.3 mmol kg^–1 in 4th instars, 7.2±0.4 mmol kg^–1 in 6th instars and 12.4±0.7 mmol kg^–1 in adults) and did not increase during the rest of the trial. No lactate was detected in the haemolymph.

During jumping, maximal (30 s period) Ṁ_O2,wb and Ṁ_CO2,wb increased significantly with age (O₂: Fig. 5A; ANOVA, F_3,24=38.2, P<0.001;CO₂: Fig. 5B; ANOVA, F_3,24=33.1, P<0.001). Ṁ_O2,wb and Ṁ_CO2,wb of jumping adults were double that of the 2nd instars(Fig. 5). The scope in Ṁ_O2,wb (jumping relative to resting values) increased with age (2nd instars, 2.0; 4th instars,2.4; 6th instars, 3.4; adults, 4.2). After 2 min of recovering, there was no effect of age on Ṁ_O2,wb(Fig. 6A), but Ṁ_CO2,wb was inversely related to age (Fig. 6B; ANOVA, F_3,24=13.8, P<0.001). Similarly, during jumping, age had a significant effect on the amount of O₂ consumed by the jumping muscle(Fig. 7A; ANOVA, F_3,24=6.6, P<0.01) and CO₂ produced(Fig. 7B; ANOVA, F_3,24=4.6, P<0.05), with both increasing with ontogeny.

Regardless of activity state (resting, jumping or recovery), grasshopper age did not significantly affect the respiratory exchange ratio (RER, the ratio of CO₂ produced to O₂ consumed), so RER values were combined across ages. RER did not change between rest and thirty seconds of jumping (0.94) but significantly increased to 1.0 when jumping ended(Fig. 8; repeated-measures ANOVA, F_6,19=10.3, P<0.001). The RER was greatest after 2 min of recovery (1.2) and then decreased back to values observed at rest.

Within an age group, hypoxia affected first-minute jump rates. Extreme hypoxia (5 kPa O₂) significantly reduced first-minute jump rate for grasshoppers of all ages, while moderate hypoxia (12 kPa O₂)significantly reduced first-minute jump rate only in the 2nd instars(Fig. 9; ANOVA, F_3,20=4.8, P<0.05) and adults(Fig. 9; ANOVA, F_3,30=28.6, P<0.001). When compared with normoxia (21 kPa), there was no effect of hyperoxia (45 kPa) on first-minute jump rate for 2nd and 4th instar grasshoppers of any age; however, hyperoxia increased the first-minute jump rate in 6th instars(Fig. 9; ANOVA, F_3,21=20.3, P<0.001) and adults(Fig. 9; ANOVA, F_3,30=28.6, P<0.001).

Age did not significantly affect endurance in the gloved box(Fig. 10), perhaps due to overall lower jump rates in the glove box. Moderate hypoxia (12 kPa O₂) reduced endurance in older grasshoppers (6th instars –ANOVA, F_3,21=70.0, P<0.001; adults –ANOVA, F_3,30=22.5, P<0.001) but not the younger animals (2nd and 4th instars; Fig. 10). Endurance was strongly reduced in extreme hypoxia (5 kPa) for all ages. Decreasing the oxygen level reduced jumping endurance for all juveniles, with hypoxic effects increasing with age in juveniles(Fig. 10). Endurance of the largest juveniles (6th instars) was significantly affected by moderate hypoxia while the endurance of 2nd instars was only significantly affected at extreme hypoxia (Fig. 10). The jumping endurance of adults was significantly affected by moderate hypoxia but proportionally less so than that of 6th instars(Fig. 10).

Our data indicate that, as generally shown for vertebrates, most measures of locomotory and metabolic performance increase with age in jumping grasshoppers. Mass-specific power output, aerobic metabolic rate, and lactate concentrations during jumping all increase dramatically and steadily throughout ontogeny. The increase in mass-specific aerobic metabolism during development seems inconsistent with the hypothesis that larger size and longer tracheae necessarily lead to increased problems with oxygen delivery in insects. However, the rise in lactate concentrations, the decrease in endurance and the increase in the oxygen sensitivity of locomotion do indicate that the safety margin for oxygen delivery is reduced in older jumping grasshoppers. However, these safety margins decrease with age due to increases in mass-specific power outputs (and ATP requirements) rather than because mass-specific oxygen delivery capacities fall.

Similar to other developmental studies of locomotory V̇_O2(Full, 1987; Chappell and Bachman, 1995; Chappell et al., 1999), the factorial aerobic scope increased with ontogeny. The whole-body aerobic scope increased from 2.1 in 2nd instars up to 4.2 in adults. Compared with other types of insect locomotion, the aerobic scope calculated during jumping is low(running ectothermic insects, 8; running endothermic insects, 26; flying insects, 129; Full, 1997). While the aerobic scopes are low, the amount of oxygen consumed during jumping in S. americana adults (7.2 ml g^–1h^–1) was greater than the mean oxygen consumption of running insects (2.4–4.5 ml g^–1 h^–1; Harrison and Roberts, 2000),suggesting that the relatively low scopes are due to elevated resting rates in this study. Our measured values for resting Ṁ_O2 were approximately double those reported by Greenlee and Harrison(2004), perhaps because the animals could be much more active in the larger chamber we used (we did not monitor activity during the prejump, resting Ṁ_O2measurements). If the resting Ṁ_O2 from Greenlee and Harrison (2004) is used,aerobic scopes increase to approximately 10 in adults, a value well within the range for an ectothermic running insect.

The high oxygen consumption of jumping S. americana compared with insect runners is consistent with the high mitochondrial contents of the muscle (Hartung et al., 2004). The only other reported V̇_O2max for grasshopper jumping (Melanoplus bivittatus; 1.3 ml g^–1 h^–1, 58 μmol g^–1h^–1; Harrison et al.,1991) was ∼5-fold lower than measured in this study. M. bivittatus may not have such a well-developed respiratory system as S. americana since peak gas exchange rates occurred after, rather than during, jumping in M. bivittatus, and oxygen consumption rates returned to resting rates much more slowly(Harrison et al., 1991). Strong support for the validity of this surprising interspecific difference is provided by the observation that traveling speeds during jumping are 4-fold higher in S. americana than in M. bivittatus(Harrison et al., 1991).

Unlike the situation for most vertebrates, older animals had reduced endurance in S. americana grasshoppers (Figs 2, 3). These trade-offs in performance variables may relate to life history. While juvenile grasshoppers utilize jumping as their primary mechanism for escape from predators and migration (Ellis, 1951), the powerful single jumps of adults are thought to be necessary to achieve the take-off velocity required for flight(Katz and Gosline, 1993).

The stimulation method used during this study seems unlikely to be the cause of the decreased endurance with development. Based on data from running vertebrates (Heglund et al., 1974) and invertebrates (reviewed in Full, 1997), one might expect the jump frequency of grasshoppers to scale with body mass raised to the–0.14 to –0.25. Within the first minute of our experiment, when grasshoppers needed no prodding, the scaling of jump frequency with body mass was –0.06, suggesting that larger grasshoppers jumped more than expected based on vertebrate data. During the second minute of repeated jumping, the scaling coefficient was –0.12. However, by the third minute, the scaling coefficient (–0.21) was similar to values reported for running animals. This suggests that the increased jump frequency of older grasshoppers in the first minute was due to increased reliance on aerobic metabolism in larger animals, rather than the method of stimulation. After two minutes of repeated jumping, when aerobic ATP production dominates, the scaling of jump performance matches running animals.

The reduced endurance of older grasshoppers is correlated with the increased utilization of anaerobic metabolism during the initial minutes when jump frequencies and power outputs are very high compared with younger animals(Fig. 4). During the first two minutes of jumping, the contribution of anaerobic ATP production ranged from 0% in 2nd instars to approximately 40% for older animals(Table 2). The constant lactate levels in the muscle after two minutes of jumping, and lack of appearance of lactate in the haemolymph, indicate that net lactate production is negligible after the first few minutes and suggest that prolonged hopping is completely supported by aerobic metabolism. A predominant reliance on aerobic ATP production during repeated jumping has also been shown in adults of the two-striped grasshopper (Harrison et al.,1991).

Potentially, the increased usage of anaerobic metabolism due to inadequate oxygen delivery during the first minutes of jumping may cause the reduced endurance with age. The stimulation of first-minute jump rate in older animals by hyperoxia (Fig. 9) and the increased lactate concentration in older animals suggest that the Pasteur effect is stimulating lactate production. Lactic acid has long been considered to be an agent of muscle fatigue, primarily due to effects on muscle pH(Juel, 1996). Leg Ṗ_CO2 also rises dramatically during jumping of adult grasshoppers(Krogh, 1913; Krolikowski and Harrison,1996), which will further reduce muscle pH, potentially causing fatigue. However, the lack of an effect of hyperoxia on endurance argues against this hypothesis. If lactate production is caused by low muscle Ṗ_O2, then hyperoxia should have reduced lactate production and increased endurance in older animals. More recent studies have suggested that muscle fatigue during burst locomotion may be due to accumulation of inorganic phosphates (reviewed in Westerblad et al., 2002). Adult Locusta migratoria use arginine phosphate to power the initial seconds of jumping (Hitzemann,1979; Schneider et al.,1989). Potentially, the use of arginine phosphate increases with age, and increased accumulation of inorganic phosphate explains the reduced endurance with age.

One of the mysteries of fatigue in jumping grasshoppers is the apparent temporal dissociation between the metabolic events thought to affect fatigue and the organism-level changes in performance. Arginine phosphate utilization is completed within the first five jumps (<15 s) in adult grasshoppers(Hitzemann, 1979; Schneider et al., 1989), and lactate concentrations are similar after two minutes(Fig. 4), but power output falls steadily over many minutes (Fig. 3). These data suggest that other processes (e.g. muscle glycogen depletion, neurotransmitter depletion, changes in hormone levels, behavioral habituation) may contribute to long-term fatigue in grasshoppers.

Although older S. americana have increased aerobic capacities and tracheal oxygen delivery capacities(Hartung et al., 2004), it does appear that there is an increasing tendency for oxygen limitation of jump performance with age. Hyperoxia improved the jump rate in the first minute(Fig. 9) only for larger/older grasshoppers, suggesting that oxygen delivery is inadequate during the first minute only in adults and 6th instars. In many tissues, inadequate oxygen concentrations can cause lactate production (Pasteur effect), and so one hypothesis is that inadequate oxygen delivery during the first minutes of jumping causes the increased lactate levels in older grasshoppers. Therefore,the usually high jump rates of 6th instar grasshoppers at 12 kPa and their rapid fatigue may be due to increased lactate production. The effect of lactate increasing initial jump frequency is supported by the finding that anaerobic and aerobic metabolism correlate with different scaling coefficients during repeated jumping. Measurements of the effect of hyperoxia on lactate production rates during jumping would test this hypothesis. Also, the increased sensitivity of endurance to hypoxia in 6th instars and adults is consistent with a reduced safety margin for oxygen delivery during ontogeny for jumping grasshoppers.

Previous research has suggested that the increased power output during jumping of older grasshoppers was attributed to increased muscle mass:body mass ratio (Gabriel, 1985a). However, our more extensive data set clearly showed that muscle mass developed isometrically with changes in body mass in S. americana(Table 1). In addition, when all developmental stages are included from Gabriel's study on S. gregaria (Gabriel,1985a), the proportion of extensor tibia muscle to body mass appears nearly isometric (1st instars 6.1%, 2nd instars 4.7%, 3rd instars 5.0%, 4th instars 4.3%, 5th instars 5.5%, adults 6.3%; N=5 for each age). Thus, it seems likely that isometric development of jumping muscle mass occurs throughout Schistocerca, as in the related African migratory locust (L. migratoria migratoriodes; Duarte, 1938).

Originally, the increased femoral exoskeleton thickness was thought to explain the greater power output in older grasshoppers by improving cuticular energy storage (Gabriel,1985b). However, it was later reasoned that thicker femoral walls would make it more difficult for grasshoppers with similar proportions of muscle to bend the cuticle to store energy, so a stiffer spring could not, on its own, explain improved jump performance(Katz and Gosline, 1992). These estimates of energy storage assumed constant muscle properties across instars (Gabriel, 1985a,b; Katz and Gosline, 1993). However, adult leg muscles have double the mitochondrial content and a 12-fold greater tracheal diffusing capacity than second instars(Hartung et al., 2004). Our data showed that the oxygen consumption rate of the jumping muscle increased 6-fold during development (during the first 2 min; Table 2). In addition, the muscles of older grasshoppers produced ATP anaerobically at much higher rates(Fig. 4; Table 2). The 10-fold increase in total leg-muscle specific ATP production rates exceeds the greater than 2-fold increase in body-mass specific power output, perhaps because of complexities in the mechanism of power production from a near-isometrically contracting muscle and a spring system(Bennet-Clark, 1975). Nonetheless, it is now clear that increases in the capacity of the jumping muscle to produce ATP (Table 2)allow older/larger grasshoppers to utilize their stiffer leg springs to produce greater power.

We thank Robby Roberson, Ron Rutowski, Glenn Walsberg, Wayne Willis, Brenda Rascón, Joanna Henry, Rekha Nair, Michelle Fay, Sydella Blatch and especially Kendra Greenlee for suggestions regarding this manuscript. We thank Manfred Grieshaber for providing access to K. Hitzemann's thesis. Ty C. M. Hoffman and Kendra Greenlee provided valuable insight and help with the respirometry set-up. This research was supported by the National Science Foundation through award IBN-9985857 to J.F.H. and award IBN-0104959 to S.D.K. and J.F.H.