Intermolt development reduces oxygen delivery capacity and jumping performance in the American locust (Schistocerca americana)
- First Online:
- Cite this article as:
- Kirkton, S.D., Hennessey, L.E., Duffy, B. et al. J Comp Physiol B (2012) 182: 217. doi:10.1007/s00360-011-0615-x
- 246 Views
Among animals, insects have the highest mass-specific metabolic rates; yet, during intermolt development the tracheal respiratory system cannot meet the increased oxygen demand of older stage insects. Using locomotory performance indices, whole body respirometry, and X-ray imaging to visualize the respiratory system, we tested the hypothesis that due to the rigid exoskeleton, an increase in body mass during the intermolt period compresses the air-filled tracheal system, thereby, reducing oxygen delivery capacity in late stage insects. Specifically, we measured air sac ventilation frequency, size, and compressibility in both the abdomen and femur of early, middle, and late stage sixth instar Schistocerca americana grasshoppers. Our results show that late stage grasshoppers have a reduced air sac ventilation frequency in the femur and decreased convective capacities in the abdomen and femur. We also used X-ray images of the abdomen and femur to calculate the total proportion of tissue dedicated to respiratory structure during the intermolt period. We found that late stage grasshoppers had a lower proportion of their body dedicated to respiratory structures, especially air sacs, which convectively ventilate the tracheal system. These intermolt changes make oxygen delivery more challenging to the tissues, especially critical ones such as the jumping muscle. Indeed, late stage grasshoppers showed reduced jump frequencies compared to early stage grasshoppers, as well as decreased mass-specific CO2 emission rates at 3 kPa PO2. Our findings provide a mechanism to explain how body mass changes during the intermolt period reduce oxygen delivery capacity and alter an insect’s life history.
The evolutionary success of insects has been attributed, in large part, to their lightweight, air-filled tracheal respiratory system (Harrison 2009b). Atmospheric air enters through spiracular openings in the body wall and then passes into tracheal tubes that develop from invaginations of the exoskeletal integument. In some insects, enlarged portions of the tracheae have thinner cuticular linings and develop into flexible air sacs that function as bellows, helping to ventilate the tracheal system. Air sacs are compressed by muscular contractions of the abdominal cavity (Harrison 1997) or changes in hydrostatic pressure in the surrounding hemolymph (Wasserthal 1996), and, thereby, convectively drive air into progressively smaller tracheal tubes. From tracheal tubes, O2 diffuses into the tracheoles (diameter from 1.0 to 0.1 μm). Tracheoles surround or even penetrate aerobically active cells; thus, O2 can diffuse across the tracheolar membrane into respiring cells (reviewed in Harrison 2009a).
Albeit structurally simple, the functional capacity and flexibility of the tracheal system are astounding. For example, O2 delivery through the tracheal system allows flying insects to have the highest mass-specific metabolic rates in the animal kingdom (Harrison and Roberts 2000; Suarez 1996). In addition, many insects can maintain their normal resting metabolic rates even in extremely low levels of hypoxia (<2 kPa PO2; reviewed in Harrison et al. 2006). Furthermore, the tracheal system is flexible enough to respond to the varying O2 demands of different tissues. For example, unsclerotized tracheoles can grow and increase branching to improve O2 delivery to aerobically demanding tissues during the intermolt period (Wigglesworth 1959, 1983). However, larger sclerotized tracheal tubes can only grow and be re-organized to increase and improve O2 delivery when an insect molts (Wigglesworth 1954, 1983).
While the respiratory system’s flexibility is astounding in most cases, at certain critical points in an insect’s development, the O2 delivery capacity is unable to meet O2 demands. During the intermolt periods when insect body mass can double (Davis et al. 1988; Greenlee and Harrison 2004; Matsuki et al. 1994; Nijhout 1979; Stockhoff 1992; Tammaru and Esperk 2007), O2 demand surpasses the mostly fixed tracheal system supply. For example, during the fifth instar, juvenile Schistocerca americana grasshoppers increase their body mass by 90%, but mass-specific metabolic rate drops by 15% (Greenlee and Harrison 2004), suggesting a mismatch between O2 demand and delivery. Indeed, it has been estimated that total tracheal system conductance decreases by 20–35% during the intermolt period (Greenlee and Harrison 2004). This decrease in O2 delivery during the intermolt period is not always compensated for by behavioral changes that help ventilate the tracheal system, such as increased abdominal pumping (Greenlee and Harrison 2004).
Greenlee and Harrison (2004) hypothesized that since insects are enclosed in an exoskeleton that can increase only during molting, an increase in body mass during the intermolt period compresses the air-filled tracheal system, reducing O2 delivery in late stage insects. In support of this hypothesis, juvenile grasshopper’s whole-body respiratory volume significantly decreases during the intermolt period, as measured using either water displacement (Clarke 1957) or inert gas (Lease et al. 2006). However, neither of these techniques provides insight into how intermolt development alters specific tracheal system morphology or respiratory dynamics in a living insect. Both topics can be addressed using synchrotron X-ray imaging (Socha et al. 2007, 2010). X-ray imaging of abdominal air sacs in early and late stage fifth instar S. americana found no difference in volume or ventilation rate with intermolt development (Greenlee et al. 2009). This contradiction with findings from non-X-ray techniques may be due to the examination of only one air sac region, a small sample size (n = 6), or a reliance on only visual cues to age the grasshoppers (Greenlee et al. 2009). Understanding how the functionality of the tracheal system may be reduced during the intermolt period is important because a decrease in O2 delivery should significantly reduce aerobic locomotory performance in late stage insects.
Grasshoppers are an excellent model in which to address O2 delivery and locomotory performance relationships, because they have the best-studied respiratory system of any insect (reviewed in Harrison 1997) and their jump performance during ontogeny has been extensively examined (Gabriel 1985a; Katz and Gosline 1993; Kirkton and Harrison 2006; Kirkton et al. 2005; Queathem 1991). S. americana molt through six flightless juvenile instars before becoming volant adults. Increasing body mass, muscle mass, and femoral exoskeleton content within the sixth instar may explain why single, maximal jump performance decreases near the end of the intermolt period (Queathem and Full 1995). However, juvenile grasshoppers more likely rely on repeated jumping. During repeated jumping, aerobic metabolism supplies 66% or more of the ATP necessary for locomotion (Kirkton et al. 2005). Therefore, we hypothesized that late stage sixth instar grasshoppers would have reduced jumping performance, because a decrease in respiratory volume would reduce O2 delivery to the jumping muscle. However, it is unclear whether intermolt development results in decreased respiratory volume throughout the entire body or whether high levels of O2 delivery are maintained to vital areas, such as the legs. In this study, we used synchrotron X-ray imaging to visualize respiratory structures at multiple body locations of different stage sixth instar grasshoppers to test the hypothesis that growing tissue during intermolt development constricts the air-filled tracheal system and reduces O2 delivery. From the X-ray images, we measured specifically how the proportion of respiratory structures, air sac ventilation rate, and air sac convective capacity vary in the abdomen, proximal metathoracic femur, and distal metathoracic femur of different stage sixth instar grasshoppers and compared these respiratory changes with jumping performance during the intermolt period.
Schistocerca americana were reared from eggs in culture at Union College, Schenectady, NY, USA. Grasshoppers were fed a diet of organic lettuce and sulfa drug-treated bran as previously described (Badman et al. 2007). We used only female sixth instar grasshoppers to control for sexual dimorphism.
Morphological changes during development
To investigate morphological changes during the intermolt period, we measured body mass, femur mass, extensor tibia muscle mass, and femoral exoskeleton mass for grasshoppers within the sixth instar. Grasshopper age was determined by painting the thorax of newly molted sixth instar individuals daily with a unique color of Tetors acrylic paint (Rockford, IL, USA). Depending on rearing conditions, the sixth instar can last 8–12 days (Kirkton et al. 2005; Queathem and Full 1995). We recorded morphological data on different groups of grasshoppers every 2 days of the sixth instar (days 2, 4, 6, 8, 10, and 12; n = 58). Body mass and femur mass were measured (±0.1 mg) using a Mettler AE 240 Balance (Columbus, OH, USA). Femur length was measured to the nearest 0.01 mm with VWR digital calipers (Batavia, IL, USA). To obtain jumping muscle mass, we sliced the femur longitudinally into half and placed it into 0.35 mol l−1 NaOH for approximately 24 h for tissue digestion, as previously described (Kirkton et al. 2005; Marden 1988). After the tissue was digested, the femoral exoskeleton was washed in distilled water, blotted dry, and re-weighed. The difference between the wet femur mass and exoskeleton mass was the wet tissue mass. We calculated the extensor tibia muscle mass to be 66% of the measured wet tissue mass (Hartung et al. 2004).
Measurements of jumping performance
To compare jump performance during the instar, we designated three relative age groups similar to those used previously to study maximal jump performance in sixth instar S. americana (Queathem and Full 1995). With n = 10 in each group, “early” stage individuals weighed <700 mg (649.2 ± 7.8 mg); “middle” stage individuals weighed from 700 to 1,200 mg (945.6 ± 43.0 mg); and “late” stage individuals weighed >1,200 mg (1,425.0 ± 28.6 mg).
Grasshoppers were removed from the colony on the day of the experiment. Individuals were weighed and isolated with food at 35°C. Grasshoppers were encouraged to jump by physical prodding for 10 min on a cotton bedsheet (310 × 80 cm) that was divided into a 5-cm numbered grid system within a 35°C temperature controlled room (Kirkton et al. 2005). As a grasshopper jumped between squares during the trial, the grid number was recorded to an Olympus VN-960PC digital voice recorder (Center Valley, PA, USA). Distances were measured from the center of each square. From the voice recorded data, we calculated jump frequency and mean distance per jump for each minute of the trial as indices of jumping performance.
Female sixth instar grasshoppers were shipped overnight from Union College to Argonne National Laboratory (Argonne, IL, USA), where they were given ad libitum access to lettuce and water until used in the experiment. Immediately, before placement in the chamber, grasshoppers were weighed and femur lengths were measured. Grasshoppers were categorized by relative intermolt stage: early, middle, and late using the same body mass criteria as above. There was no significant difference between the body mass or femur length of the grasshoppers within each age group used during the jumping analysis and those used in the X-ray studies at Argonne National Laboratories, allowing us to make valid comparisons between jumping performance and tracheal system function.
The grasshopper respiratory system was visualized using synchrotron phase-contrast X-ray imaging with a 2× objective lens, and a Cohu-cooled charge-coupled device video camera as previously described (Socha et al. 2007). Digital video was recorded using a TRV 900 camcorder (Sony, Tokyo, Japan). A 400-mesh metal grid was placed in the beam as a spatial scale for calibration of the X-ray images.
Since each grasshopper was considerably larger than the X-ray field of view (2 mm × 2 mm), a remote-controlled stage was used to move the insect within the beam, allowing us to scan different parts of the body. Exposure to X-ray radiation in a typical experiment was no longer than 15 min. Previous work has shown that radiation exposure for nearly 30 min did not affect grasshopper CO2 emission or body temperature (Greenlee et al. 2009; Socha et al. 2007).
The general experimental protocol was as follows: (1) individual grasshoppers were placed in an unsealed Plexiglas and Kapton respirometry chamber for at least 20 min to become familiar with the chamber and allow for reliable respirometry data (Greenlee and Harrison 1998); (2) the chamber was sealed and flushed with normoxic air (21 kPa PO2) without exposure to the beam for approximately 5 min to serve as a baseline; (3) the X-ray beam was turned on and air sacs in the abdomen, proximal metathoracic femur, and distal metathoracic femur were recorded for approximately 2 min each under normoxic conditions while simultaneously collecting CO2 emission rates; and (4) the grasshopper was exposed to hypoxia (3 kPa PO2) and the same air sacs were then recorded again for 2 min each while also simultaneously collecting CO2 emission rates.
CO2 emission rates
Grasshoppers were placed in relatively air-tight chambers allowing for high-speed flow-through respirometry to measure CO2 emission, as previously described (Greenlee and Harrison 2004; Greenlee et al. 2009). Dry CO2-free air mixtures (21 kPa PO2 for normoxia or 3 kPa PO2 for hypoxia) were created from compressed air and nitrogen tanks, scrubbed using Drierite and Ascarite scrubbers, and delivered via a Sable Systems MFC-4 mass-flow controller (Las Vegas, NV, USA) and Sierra Instruments mass-flow meters (Monterey, CA, USA) to the chamber. The flow rate was maintained at 4 l min−1 for both O2 treatments, such that individual breaths could be resolved. Excurrent air was dried with MgClO4 before being sent to a Li-Cor 6251 CO2 analyzer (Lincoln, NE, USA). The CO2 analyzer was adjusted daily using a two-point calibration (CO2-free air and a gas mix containing 99 ppm CO2:balance N2).
Grasshoppers were only included in the analysis if both CO2 emission data and X-ray video data were useable, resulting in the following sample sizes: early (n = 10), middle (n = 9), and late (n = 9). Data files were analyzed using Expedata v. 1.1.15 (Sable Systems, Las Vegas, NV, USA). Chamber baseline CO2 values were subtracted from animal measurements. Then, the mean CO2 fraction was calculated for each treatment (pre-beam normoxia, normoxia in the beam, and hypoxia in the beam). We calculated the CO2 emission rate by multiplying the mean fraction of expired CO2 by the flow rate in ml min−1. Absolute CO2 emission rates were converted from ml min−1 to μmol h−1, and mass-specific CO2 emission rates were calculated for each individual.
Air sac ventilation frequency and convective capacity
Because the examined air sacs compress equally in all dimensions, changes in air sac area should approximate changes in air sac volume. To determine the relative convective ability of an air sac, we measured the maximum and minimum area of an air sac during each of three consecutive breaths using ImageJ software (NIH, Bethesda, MD, USA). The initial breath of the sequence was selected using a random number generator. We then calculated both the mean change in air sac area (ΔA) as an indicator of the amount of air moved per breath, and the percent difference in air sac area [(maximum area−minimum area)/maximum area] as a relative indicator of an individual air sac’s compressibility. This procedure was repeated for the air sac at each of the three anatomical locations in each grasshopper and in both normoxia and hypoxia.
Projected area of respiratory system
In all cases, differences among mean values were considered significant when p < 0.05, and values are reported as mean values ± SEM. Changes in body mass, femur length, femoral exoskeleton mass, and extensor tibia muscle mass were analyzed using regression analysis. Jump frequency and the mean distance per jump for each minute of repeated jumping were analyzed using repeated-measures ANOVA (JMP software v.8.1; SAS, Cary, NC, USA).
CO2 emission and all other respiratory parameters were analyzed using repeated-measures ANOVA (PASW Statistics v. 17; SPSS, Chicago, IL, USA). In the analyses, O2 and/or air sac location was treated as within-subjects factors and stage within an instar was treated as a between-subjects factor. Maximum area, minimum area, and ΔA had unequal variances (Mauchly’s test of sphericity, p < 0.001) and zero points that made log transformation impossible. Therefore, each variable was transformed using the equation log (y + 1), where y = dependent variable. This transformation equalized the variance for maximum area, but not for minimum area or ΔA. To account for unequal variances, we used lower-bound corrected degrees of freedom in the repeated-measures ANOVA. Air sac compressibility percent, percent air sacs, percent tracheae, and percent non-respiratory structure were arcsine transformed before repeated-measures ANOVA. We used linear regression to determine whether body mass correlated with changes in percent air sacs, tracheae, or non-respiratory structure.
Scaling of metathoracic femur mass, extensor tibia muscle mass, femoral exoskeleton mass, and metathoracic leg length with body mass during the intermolt period of the sixth instar of S. americana (n = 58)
Upper (95% CI)
Lower (95% CI)
Femur mass (mg)
Extensor tibia muscle mass (mg)
Femoral exoskeleton mass (mg)
Leg length (mm)
The mean jump distance decreased throughout the trial regardless of stage (F9,19 = 9.8, p < 0.0001; Fig. 3b). There was no overall effect of stage on mean jump distance during the entire trial. Mean jump distance drops from minute 1 to 2 by 18% in early stage and 33% in middle stage grasshoppers; however, then, it is sustained throughout the rest of the trial. Compared to the first minute mean jump distance, the tenth minute value was 19% lower in early stage, 37% lower in middle stage, and 50% lower in late stage grasshoppers. However, when compared to the second minute values, the tenth minute mean jump distance was less than 1% lower in early stage, 6% lower in middle stage, and 34% lower in late stage grasshoppers (Fig. 3b).
Within subjects, grasshopper mass-specific CO2 emission varied differently with oxygen treatment depending on the instar stage (O2 × stage interaction, F3,78 = 6.996, p < 0.001; Fig. 4b). All instar stages showed no difference between mass-specific CO2 emission in pre-beam normoxia and normoxia in the beam. However, early and middle stage grasshoppers increased mass-specific CO2 emission rates during hypoxia. Mass-specific CO2 emission of late stage animals did not vary significantly with any treatment (Fig. 4b).
Air sac convective capacity
Minimum air sac area varied differently with anatomical location depending on the stage of the grasshopper (anatomical location × stage interaction, F0.541,25 = 13.36, p < 0.001); however, it did not vary with oxygen treatment (Fig. 6b). In both normoxia and hypoxia, early and middle stage grasshoppers had similar minimum air sac areas that were also significantly greater than those of late stage grasshoppers in both normoxia (sevenfold greater in abdominal air sac; tenfold greater in distal femoral air sac) and hypoxia (4-fold greater in abdominal air sac; 24-fold greater in distal femoral air sac; Fig. 6b). Early stage grasshoppers had the largest minimum air sac area in the proximal femur of any stage (fivefold greater than late stage in normoxia; fourfold greater than late stage in hypoxia; Fig. 6b).
The estimated amount of air moved per breath by an air sac (ΔA) was significantly affected by instar stage (F1,25 = 6.1, p < 0.01) and air sac location (F2,25 = 66.2, p < 0.007; Fig. 6c). In the abdomen, early and middle stage grasshoppers had a twofold greater ΔA than late stage grasshoppers in normoxia, but similar ΔA in hypoxia (Fig. 6c). In the proximal femur, middle stage grasshoppers had a fourfold greater ΔA than early or late stage grasshoppers. In the proximal femur during hypoxia, early and middle stage grasshoppers had similar ΔA and were fivefold greater than late stage grasshoppers. The distal femur ΔA was similar for all stages in normoxia and significantly reduced with instar stage in hypoxia (Fig. 6c).
Relative air sac compressibility varied differently depending on the instar stage and anatomical location of the air sac (stage × anatomical location interaction, F2,22 = 7.4, p < 0.01) and also varied differently with stage and O2 treatment (stage × O2 interaction, F2,22 = 4.44, p < 0.03; Fig. 6d). The abdominal air sacs of late stage grasshoppers compressed more in both normoxia (twofold greater than early stage) and hypoxia (threefold greater than early stage; Fig. 6d). In the proximal femoral air sac, middle stage grasshoppers had the greatest compressibility regardless of O2 treatment (Fig. 6d). In addition, proximal femoral air sacs of early grasshoppers had a greater compressibility than late stage air sacs in hypoxia but not in normoxia (Fig. 6d). Distal femoral air sac compressibility was significantly reduced by hypoxia in middle and late stage grasshoppers (Fig. 6d).
Proportion of body dedicated to respiratory structures
Femoral air sac ventilation
Unlike abdominal pumping, ventilation in the femur cannot be observed externally. Thus, this study is the first to examine functioning femoral air sacs in an intact insect. Femoral air sacs in early stage grasshoppers were clearly visible and ventilated the leg throughout the trial. In contrast, many late stage grasshoppers had no visible femoral air sacs, supporting our hypothesis that air sacs become compressed as tissue mass increases during the intermolt period. The leg may be more affected by intermolt tissue growth because the femoral exoskeleton is more sclerotized and rigid than the abdominal exoskeleton (Chapman 1998; Hendricks and Hadley 1983), and it significantly increases in both mass (Table 1) and thickness (Gabriel 1985b) during the intermolt period. Indeed, late stage grasshoppers had lower ventilation frequencies (Fig. 5), smaller air sacs (Fig. 6), and a lower proportion of their femur dedicated to air sacs than early stage grasshoppers (Figs. 7, 8). The less pliable femoral exoskeleton also reduces the hypoxic response in late stage grasshoppers, as evidenced by the lack of an increase in femoral ventilation frequency in hypoxic late stage grasshoppers (Fig. 5). The mechanisms underlying femoral air sac ventilation, including whether airflow down the leg is tidal or unidirectional, are currently unknown (Hartung et al. 2004; Kirkton 2007); however, femoral air sacs could be ventilated by abdominal pumping, femoral muscle contractions, and/or hemolymph pumping.
Abdominal air sac ventilation
In this paper, we demonstrate that during intermolt development, O2 delivery capacity in the abdomen is significantly reduced even though O2 demand increases. This could be problematic for developing orthopterans, since abdominal ventilation is the primary source of convective O2 delivery in juvenile grasshoppers (Harrison 1997). The changes in late stage grasshoppers were due mostly to the decrease in air sac area and ΔA (Fig. 6), since abdominal ventilation frequencies were similar to those of early stage grasshoppers (Fig. 5a). In addition, in the abdomen, late stage grasshoppers have a greater proportion of non-respiratory tissue and a relatively lower proportion of air sacs than early stage animals (Figs. 7, 8). Interestingly, late stage grasshoppers have a higher proportion of abdominal tracheae than early stage animals. This finding is likely the result of a change in the size and shape of air sacs due to developing body mass, such that they now appear to resemble tracheae (Fig. 2).
Similar to our data, an earlier X-ray study of fifth instar S. americana found that abdominal air sac ventilation frequency did not vary with stage (Greenlee et al. 2009). Although Greenlee et al. (2009) found no difference in air sac compressibility, we found that relative compressibility increased over twofold at the end of the instar (Fig. 6d). One possible explanation for the discrepancy is that the previous study estimated changes in volume, while this study uses changes in area as an index of air sac compressibility. However, we feel our use of area as an index of volume changes is justified because when representative grasshoppers are viewed both laterally and dorsoventrally, the mean depth-to-width ratio of each air sac was not significantly different across stages (abdomen: 1.1; proximal femur: 1.2; distal femur: 0.9) and did not vary significantly when expanded or compressed. The more likely explanations for the differences between these two studies are the previous study’s small sample sizes (n = 6) and use of only visual cues to determine grasshopper age. Alternatively, there may be inherent differences between fifth and sixth instar respiratory systems. Our findings more closely match those from external, non-X-ray studies of adults that had body masses similar to those of our sixth instars. Early stage adults (850 mg body mass) had 49% greater abdominal tidal volumes than late stage adults (1,450 mg body mass; Greenlee and Harrison 2004), suggesting that early stage animals are able to deliver more O2 per abdominal compression.
A reduction in respiratory system volume during the intermolt period of juvenile grasshoppers has been shown using both water displacement (90% lower respiratory volume in sixth instar Locusta migratoria; Clarke 1957) and inert gas (80% reduced respiratory volume in sixth instar S. americana; Lease et al. 2006). Grasshoppers in the present study also decreased in respiratory system capacity during the intermolt period, with the proportion of the body dedicated to air sacs decreasing by 72% in the abdomen and 98% in the femur. Our estimates differ slightly because both of these other techniques provide an indication of whole-body respiratory system volume changes while our data examine the changes in the abundance of specific respiratory structures at exact anatomical locations. Furthermore, both of these other techniques may overestimate the tracheal volume by either over-inflating the air sacs with water (water displacement) or having excess helium dissolved in insect body tissues or trapped under the wings/hair (inert gas), as reviewed by Lease et al. (2006). Therefore, X-ray imaging of the entire body of early and late stage insects will be necessary to accurately determine how both total tracheae and air sac abundances change during the intermolt period.
Morphology, O2 delivery, and jump performance
Unlike adults, juvenile grasshoppers lack functional wings and must jump to find food, escape predators, or disperse (Kuitert and Connin 1952). Consequently, because of their size and lack of flight ability, sixth instar juveniles suffer higher predation rates than younger juveniles or adult grasshoppers (Stower and Greathead 1969). Therefore, jumping ability is extremely important in sixth instar grasshoppers. We hypothesize that the reduced jumping performance in late stage grasshoppers is attributed to both morphological and O2 delivery changes during the intermolt period.
Late stage grasshoppers have over twofold greater body mass, a relatively lower muscle mass:body mass ratio (Table 1), significantly more femoral exoskeleton (Table 1; Gabriel 1985b), and stiffer tibia (Katz and Gosline 1994) than early stage grasshoppers. These morphological changes suggest that the jumping muscles of late stage grasshoppers have to work harder to bend their relatively stiffer cuticular springs to store energy prior to a jump, as well as propel a significantly greater load. Indeed, during single maximal jumps, late stage sixth instar grasshoppers do not jump as far as middle stage grasshoppers because of a lower muscle mass:body mass ratio and a possible decrease in cuticular stiffness (Queathem and Full 1995). Interestingly, early and late stage grasshoppers show similar maximal jump distances. The reduced maximal distance in early stage grasshoppers is due to their relatively small, developing jumping muscle and non-tanned femoral cuticle (Queathem and Full 1995). Similar to single maximal jumps, our repeated jumping results show that compared to early stage animals, late stage grasshoppers had similar mean jump distances during most of the trial (Fig. 3a). However, late stage grasshoppers also had increased fatigue rates and reduced jump frequency during the trial (Fig. 3b). The decline in jump frequency with equivalent jump distance suggests that although it takes longer between jumps for late stage grasshoppers to bend cuticular springs, the energy storage system allows them to travel the same distance. The reduced jumping performance of late stage grasshoppers may also be the result of a behavioral modification. For example, although we varied the physical prodding throughout the jumping trial, it is possible that late stage grasshoppers habituated more to the stimulation than early stage animals and chose not to jump as frequently.
Beyond changes in muscle mass and exoskeleton characteristics, increased jumping fatigue in late stage grasshoppers may be compounded by the reduction in whole-body abdominal and femoral O2 delivery. While a grasshopper does not appear to pump its abdomen during a jump, abdominal pumping rate increases significantly between jumps and after repeated jumping (Krolikowski and Harrison 1996). Because aerobic metabolism supplies 66% of the ATP during the first 2 min of jumping and nearly all of the ATP after 2 min (Kirkton et al. 2005), the decrease in O2 delivery capacity during the intermolt period has serious implications for jumping grasshoppers. In support of this idea, acute hypoxia does not alter maximum force-generating capacity in mammalian skeletal muscle, but it increases fatigue rates (Degens et al. 2006; reviewed in Perry and Rupp 2009). Reduced O2 delivery during the intermolt period may play a similar role in the grasshopper jumping muscle by not altering the maximal force generating capacity of the muscles or energy storage system, but by increasing muscle fatigue, explaining the similarity between early and late stage jump lengths while resulting in a lower jump frequency. Muscle fatigue is thought to occur from an accumulation of energy metabolites (inorganic phosphate, Ca+2, H+, or lactate; Juel 1996; Kirkton et al. 2005; Perry and Rupp 2009; Westerblad et al. 2002); however, it is unknown whether late stage grasshoppers have a greater accumulation of fatigue-inducing energy metabolites during repeated jumping than early stage animals.
Implications of decreased tracheal system function during intermolt growth
Our results are the first to demonstrate that decreased abundance of respiratory structures and declining air sac function are the mechanisms for the decreased safety margins for O2 delivery within an instar. For critical tissues, such as the jumping muscle, these intermolt changes in O2 delivery may have serious life history implications. It may be more difficult for late stage animals to find food, escape predators, or disperse. Thus, late stage animals would be predicted to have higher predation rates than early staged juveniles.
Another way in which respiratory system structure and function may influence life history is by limiting insect body size. As insects increase in body size, the amount of tracheae required to provide oxygen delivery to the leg becomes larger than the space available in the coxae (Kaiser et al. 2007). Current scaling models suggest that larger than extant insects would require an unrealistic proportion of their body devoted to respiratory structures (Greenlee et al. 2009). This study provides another possible mechanism for the limitation on insect body size. As insects grow within an instar, the decreased oxygen delivery capacity may require them to molt at a smaller size to maintain adequate tissue PO2 levels. Conceivably, insects could respond to hypoxia by tolerating a decrease in the PO2 gradient from atmosphere to mitochondria, but it is unclear how tissue PO2 actually varies. Studies of tissue PO2 of early and late stage grasshoppers in hypoxia could be conducted using electron paramagnetic resonance oximetry (Kirkton 2007).
Lastly, decreases in O2 delivery capacity within an instar may be a trigger for molting. While it has long been known that molting can be stimulated by abdominal stretch receptors or upon attainment of a critical body mass (Nijhout 1975, 1979), the initial trigger for molting remains unclear. Molting is tightly controlled by a hormonal cascade that begins when pre-ecdysis triggering hormone (PETH) is released from Inka cells. Recent work has shown that PETH and its receptors are present in the tracheal system of several insect groups, an intriguing finding (Roller et al. 2010). Since body mass and O2 consumption are correlated, the idea that O2 availability ultimately controls molting is plausible. Indeed, Tenebrio molitor insects reared in hypoxia show a decrease in intermolt period (Greenberg and Ar 1996) and late stage insects have lower safety margins for O2 delivery (Greenlee and Harrison 2004, 2005). Future studies on the relationship between PETH and O2 availability will help to elucidate the role of O2 in molting.
Support for this project was provided by the Union College Faculty Research Fund (SDK), National Science Foundation IOS-0953297 (KJG), and National Institutes of Health 2P20RR0l5566 from the National Center for Research Resources (KJG). The contents of this study are solely the responsibility of the authors and do not necessarily reflect the views of the NIH. Use of the Advanced Photon Source at Argonne National Laboratory was supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract DE-AC02-06CH11357. We would also like to thank Leah Pepe and Kathryn Jackson for assistance with data collection. We would like thank the three anonymous reviewers for their helpful comments in improving the manuscript.