| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a Department of Biological Sciences, Wayne State University, Detroit, Michigan
b Department of Biology, University of Oregon, Eugene.
Robert Arking, Wayne State University College of Science, Department of Biological Sciences, Biological Science Building, Detroit, MI 48202 E-mail: rarking{at}biology.biosci.wayne.edu.
Decision Editor: Jay Roberts, PhD
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionConclusionsReferences |
|---|
Aworking understanding of the general principles underlying the question as to why animals age and undergo senescence has been achieved in principle and can be viewed as involving trade-offs between different life-history traits (1). The details of this process, however, are another matter; the lack of agreement on them suggests that our understanding is not as complete as it should be (2). The antagonistic pleiotropy theory (3) and the mutation accumulation theory (4) are the two major alternative population genetic mechanisms in the evolution of aging (1). In the first theory, natural selection actively brings about aging in the process of increasing mean fitness. This theory suggests that selective advantages in one part of the life cycle should have compensating deleterious effects in another part of the life cycle of the same individual. The search for evidence supporting the existence of such trade-offs has involved the analysis of a number of different traits. In the second theory, late-acting mutations are the primary forces establishing aging and escape being eliminated by natural selection as a result of the limitation of their deleterious effects to only older age classes.
Several laboratories have used artificial selection to construct long-lived strains and then have used them to compare and contrast important life-history characters in the long-lived and founder populations (5)(6)(7)(8). Previous work has shown that these long-lived strains display decreased fecundity early in life and increased fecundity late in adult life as compared with their normal-lived progenitors (1)(5)(6)(9); however, this decreased fecundity is highly dependent on environmental conditions and is evidently an unstable character (10). Other studies with long-lived strains of Drosophila have shown that there is no difference in early fecundity between long-lived and normal-lived populations (8), although an increased fecundity late in the adult life of the long-lived strains was observed. Finally, similar selection pressures for long life in different sister founder lines can generate trade-offs in some lines but not in others, suggesting that the two processes are not causally connected in all cases (11).
Other life-history traits have also been examined. It has been reported that an increased development time and increased adult body size are positively correlated with increases in longevity in some strains (12) but not in others (13). The design of some selection experiments may have inadvertently selected for faster absolute (although not relative) development time, thus confounding the interpretations (7)(14). In addition, other studies comparing the effects of development time and adult body size on longevity have shown that the two are not coupled (15)(16)(17)(18)(19)(20).
The mutation accumulation theory of Medawar (4) postulated the existence of a class of pleiotropic genes that were neutral early in life but had deleterious effects late in life. Such genes would accumulate because their late deleterious effects would escape the scrutiny of natural selection. A review of the available experimental tests of this theory led Curtsinger and colleagues (2) to conclude that the then-existing evidence in favor of the theory was inconclusive. Subsequent work by this group using quantitative genetic studies on large and defined populations found evidence either failing to support the mutation accumulation model (21) or else consistent with the operation of mutation accumulation only in long-standing laboratory-adapted populations (22). In this case the process probably reflects the culling of deleterious mutations accumulated in the laboratory and not a process that would normally operate in the wild. A recent investigation by Pletcher and colleagues (23) used an ingenious experimental procedure that allowed them to compare mutational variability in animals before and after periods of mutation accumulation. They concluded that mutation accumulation was operative in their lines.
One fact that stands out from this brief review, besides the lack of any global agreement, is that most of the life-history traits examined by the various laboratories, including ours, were those belonging to the adult stage of the Drosophila life cycle. Yet there is nothing in the theories that requires that assumption. Accordingly, we set out to assay a number of developmental life-history traits in our normal-lived (R) and long-lived (L) strains, as well as in two isochromosomal strains (002 and 220) derived from them (24).
Even though we have a reproductive process in our laboratory carefully constructed so as to keep the normal-lived strains under conditions of no consistent selective pressure (7)(14), these strains have been in laboratory culture since 1979. Therefore, we believed it was important to control for the effects of long-term laboratory adaptation and test our results on another independent set of strains. Accordingly, in 1992 we used the technique of selecting for late life reproduction as previously described (7) to generate a new and independent set of normal-lived (RPH) and long-lived (LPH) strains. These strains were derived from isolines of wild caught Drosophila females trapped near Port Huron, Michigan in the fall of 1992 (Allan G. Force, personal communication). These isolines were combined and the resulting Port Huron strain was then subjected to simultaneous selection for both longevity and development time so as to yield the following four selected strains: strain 1, normal-lived fast-developing (RPHF); 2, normal-lived slow-developing (RPHS); 3, long-lived fast-developing (LPHF); and 4, long-lived slow-developing (LPHS).
Once selection was completed on these strains, they were assayed for a number of developmental indices. These Port Huron strains have not been described before. There were two goals to this experiment. The first was to independently derive a new set of long-lived strains that we could use to test whether conclusions drawn from our previously described Wayne State strains were of a general or a restricted import. The second goal was to test the hypothesis that selection for developmental speed would have a significant effect on longevity.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
68 days whereas the R strain only slightly increased to 45 days. These life-span parameters have been more or less stable in these strains since 1984 (see (25) for a discussion of the historical data). Isochromosomal strains..-- The 002 and 220 strains were derived from the parental R and L strains in a series of controlled chromosome substitution crosses, resulting in the production of a series of stable isochromosomal strains (24 for details). Briefly, the 002 strain is composed of animals whose third chromosome homologues were all derived from the L strain, whereas the remaining chromosome pairs were derived from the R strain. The 220 strain represents the opposite case, where the third chromosome homologues were all derived from the R strain, whereas the remaining chromosome pairs were all derived from the L strain. Thus, the 002 strain has a long-lived phenotype even though two of its chromosomes have been replaced with homologues from a normal lived strain; the 220 strain has a normal-lived phenotype even though two of its chromosomes have been replaced with homologues from a long-lived strain. Including these strains in the experiment allowed us to determine whether developmental time and viability correlated with the presence of the chromosome responsible for extended longevity. The L, R, and isochromosomal lines are each derived from the same progenitor stock. Therefore we will refer to them collectively as the Wayne State strains.
Port Huron strains..-- In 1992, two populations of wild-caught D. melanogaster were trapped in a pear orchard outside of Port Huron, Michigan and brought into the lab. They were subjected to a two-way hybrid cross in order to establish a base stock population containing a high degree of genetic diversity. This base stock was reproduced late in life for the PHL strain and reproduced randomly during the first 30 days of life for the PHR strain, using the same techniques as described for our R and L strains (7). In addition, each of these strains was subjected to a simultaneous selection for either fast (RPHF, LPHF) or slow (RPHS, LPHS) development. This was done by allowing only the animals that emerged during days 1–4 of eclosion (i.e., fast developers) or during days 6–10 of eclosion (i.e., slow developers) to become the next breeding generation of that stock.
All strains were maintained as randomly bred populations of 200–300 animals in pint bottles containing 50 ml of a standard yeast–sucrose–agar media supplemented with a top coating of air-dried liquid yeast. Bottles were maintained in an incubator at 25°C on a 12 h/12 h light/dark cycle with moderate but uncontrolled humidity.
Egg Collections and Measurement of Development Times
Egg collections were performed on sufficient animals (400–500) so as to obtain a number of eggs adequate for the entire study from a single 1-hour collection on a petri dish containing agar and surface-dried liquid yeast suspension. Only animals 5–10 days old were used, as such animals yield the highest levels of egg viability (9)(26). A 1- to 2-hour precollection was performed to minimize the number of eggs laid that were prefertilized and held by the female. The surface of the agar was scored to facilitate oviposition. After collection, eggs were transferred to the surface of food vials (2.5 x 9.5 cm) containing our standard medium supplemented with surface-dried liquid yeast. For the Wayne State strains, vials were set up at both low larval density (20 eggs/vial) and high larval density (100 eggs/vial). For the Port Huron strains, only the high larval density was used. All vials were maintained at 25°C as described. In order to control for unforeseen environmental variables, all of our data comes from two synchronous experiments, one involving only the Wayne State strains and the other involving only the Port Huron strains.
The recording of hatching times began at 16 hours after the midpoint of collection (hour 0), at which time any hatched eggs and their corresponding first instar larvae were removed and not included in the experiment. Such early hatches represent eggs that were fertilized and held by the female prior to egg laying, and typically represent 10% of the egg population. Hatchings were recorded every hour up to hour 25. Pupation time commenced at hour 108 and was recorded every 4 hours up to and including hour 204. The time of pupation was established as the time when the pupae first attain a tannish color, typically several hours after immobilization. Time of eclosure began at hour 204 and was recorded every 4 hours through hour 324. The times selected for recording these three developmental periods encompassed the entire developmental periods in all eight strains; there were no hatchings, pupations, or eclosions occurring outside the recorded time windows.
Developmental Viability
Once the eggs were placed in the vial at hour 0, the animals were not physically handled in any way. Viabilities were therefore obtained on the same animals used for these timing experiments.
Pupation Height
Pupation height was recorded as the distance in 1-cm increments above the surface of the food and was obtained on the same groups of animals used for the developmental timing experiment. Pupation height is a behavioral trait that reflects fitness in laboratory populations of Drosophila (27), and we measured this in order to verify the conclusions obtained from the developmental studies.
Adult Body Weight
For the Wayne State strains, body weight measurements were performed on six replicates of approximately ten animals of each sex of each strain at day 5 of adult life and averaged to the nearest 0.1 mg. For the Port Huron strains, body weight measurements were performed on six replicates of 10 animals of each sex of each strain at day 14 of adult life.
Statistical Analysis
The data were compared for statistical significance by using the SPSS for Windows, version 7, statistical package (SPSS, Inc., Chicago). The mean times for each developmental stage were compared by using the one-way analysis of variance (ANOVA) protocol such that the F test was used to ascertain statistical significace. The F values, degrees of freedom, and the p values for each of these tests are indicated in Table 1 . Those stages that yielded significant F values were then assayed with the post hoc Scheffé test, which indicated signficant intraset differences and grouped them into homogenous subsets. In Table 1 , cohorts with different superscript numbers within the same developmental stage are significantly different from one another ( p < .05). The Wayne State and Port Huron strain sets were analyzed separately. An identical analysis was done for the pupation height data of Table 4 below. The developmental mortalities of Table 2 were analyzed by using the chi-square test and the Wilcoxon paired sample sign test (28).
|
|
|
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
S 
. This finding is expected and has been shown before by numerous investigators (16). However, within each density treatment, there were consistent differences and similarities in the total developmental time between strains. Of the four HD lines, three showed no significant differences in overall development time between them (LHD, 002 HD, and 220 HD). There is a significant difference 
between the normal-lived RHD and the LHD strains (Table 1 ). At low density, the LLD strain and the 220 LD strain both develop significantly faster than do the corresponding 002 LD or RLD strains under similar density treatments (S > 5.73, p < .006; Table 1 ). In order to determine the developmental stages in which these differences in total developmental times were taking place, each developmental compartment was analyzed in a manner similar to that done for comparing total developmental time.
|
Larval development clearly showed differences in duration when compared with low- and high-density treatments; there was an increase of 18–20 hours in time spent as a larva in high-density conditions within strains (Table 1 ). Within a density treatment, there were also differences between strains such that the long-lived L strains developed significantly faster as larvae than did their same density normal-lived R strain counterparts 
; however, the long-lived 002 strain developed significantly, slower than did the same density normal-lived 220 counterparts 
. The change in larval developmental time as a function of density treatment ranged from 14% to 17% of total larval development. Thus, longevity differences are not necessarily reflected in larval development differences.
Pupal development, in contrast, was fairly constant regardless of strain or density treatment, except for the 002 HD strain, which had the shortest pupal development time and was placed in a separate group because of its signficant difference from the LHD strain on the Scheffé test 
. It appears that the trend for the L and 220 strains to develop faster than the R and 002 strains during the larval phase does not hold true for the pupal developmental period.
Longevity of the Port Huron Strains
We now take the opportunity to describe these recently selected strains. Fig. 2 and Fig. 3 show the changes in mean longevity (R vs L) when selection for development time is held constant. It is clear that selection for long life is effective regardless of the development time selection: the LPHF strain has a 22% increase in mean life span over its RPHF control, and the LPHS strain has a 36% increase over its RPHS control. These values are comparable with those observed in other selection experiments for adult longevity (7). We have demonstrated elsewhere that the extended longevity characteristic of the Wayne State strains is robustly correlated with an enhanced antioxidant gene expression and enzyme activity coupled with a reduced level of oxidative damage (13)(29)(30). We do not yet have a firm understanding of the mechanisms underlying the extended longevity of the Port Huron strains, although these experiments are in progress.
|
|
Developmental Timing for the Port Huron Strains
For each strain at each developmental stage, the mean duration, SEM, and N were calculated as described (Table 1 ). Fig. 4 depicts the mean and variance of the overall developmental time for each strain. The Scheffé tests indicates that there is no significant difference in total developmental time between the RPHF and LPHF strains, nor is there a difference between the RPHS and LPHS strains, but the two sets are not identical (Table 1 ). In order to determine the developmental stages in which these differences in total developmental times were taking place, each developmental compartment was analyzed.
|
, which themselves are statistically identical to one another 
.
There is no significant difference in larval developmental time between the RPHF and LPHF strains 
. However, there is a significant difference in this value between the RPHS and LPHS strains 
. Selection for slow development and normal longevity has markedly lengthened the larval period in the RPHS strain such that it is now the longest in duration observed among all 12 strains tested.
There is a significant difference in pupal developmental time between the four Port Huron strains (Table 1 ), reflecting the significant difference between the LPHF strain and the LPHS, RPHF, and RPHS strains (S < 3.7617, p < .02). It should be noted that the LPHF strain now has the shortest pupal development time observed among all 12 strains tested.
Developmental Viability of the Wayne State and Port Huron Strains
Because the developmental timing studies were done on known numbers of animals that were not handled once the eggs were counted and placed in vials, it was possible to obtain accurate developmental viabilities. The data are shown in Table 2 . There are statistically significant differences in the mortalities within both the Wayne State and the Port Huron strains (see notes to Table 2 ). An inspection of that data makes it quite obvious that, in every case, the normal-lived member of each normal-lived–long-lived pair has a significantly higher developmental viability. The only exception to this statement occurs in the pupal stage of the 002 LD–220 LD comparison, but even this isolated stage-specific exception is offset by the decreased embryonic and larval viability of the 002 LD strain and thus does not alter the fact of the higher overall survivability of the normal-lived member of each pair. This directional difference in the mortalities of each pair is statistically significant as assayed by the Wilcoxon paired sample test 
. The higher developmental survival is not due to density conditions, nor is it due to developmental speed. It is tightly correlated with adult longevity in both of the independently selected Wayne State and Port Huron sets of independently selected sets of strains (see Fig. 5).
|
. This finding is true at both low- and high-density conditions, with adult body weights being less in all strains raised under high-density conditions. In contrast, the data from the 14-day-old Port Huron strains show that under selection for slow development and under high-density conditions, the RPHS strains have significantly lower body weights than do their LPHS cohorts 
. However, when selected for fast development, the LPHF strain females have higher body weights than do their RPHF cohorts, although this difference is not statistically significant.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
. None had a mean longevity remotely comparable with that commonly found in long-lived laboratory strains. In our laboratory, our long-lived flies have a higher lifetime fecundity than do their normal-lived progenitors (9), and such a relative phenomenon might be expected to occur in the wild as well. The question then arises, Why is long life not a common phenotype in the wild? The mutation accumulation and antagonistic pleiotropy concepts are valuable hypotheses that address this question (1). Unfortunately, some of the correlations suggested by one or the other of these theories are not fully supported by empirical evidence: a decreased early fecundity seems not to be a constant aspect of long life in all strains; older animals may actually have a lower age-specific mortality and higher life expectancy than younger cohorts; and so forth. A number of ingenious experiments that strive to address this question have been reported (see 1,2 for references; also 23), but there has yet to be reported any evolutionarily important life-history trait that is robustly associated with selection for extended longevity across a number of independently derived strains. Of course, there is nothing in either theory that requires that all selected long-lived lines should respond with similar trait correlations. In such a case, however, it will be difficult to develop a comprehensive explanation as to why long life is not yet a common phenotype in the wild.
It is clear that normal-lived animals have a significantly higher developmental survival rate than do long-lived animals (Table 2 ). The overall survival differences range from a 3% to 18% 
advantage in favor of the normal-lived animal (Table 2 ). Differences of this magnitude are likely to be evolutionarily significant. The inverse relationship between longevity and viability is statistically significant (Fig. 5). It is possible that this inverse relationship is enhanced by simultaneous selection for slow development (see caption to Fig. 5). This differential viability is observed in almost all (34/36) of the different strain-specific developmental stages examined. It appears to reflect a general developmental physiological superiority of the normal-lived embryo, larvae, and pupae over their long-lived cohorts.
One specific indicator of this differential fitness is pupation height. It is well known that animals that pupate on the food surface or low on the side of the vial have a significantly higher mortality than do those animals that pupate higher up (see 27 for data and references). The data of Table 4 show that every normal-lived strain pupates at a higher mean height above the food than do their respective long-lived cohorts, and that this relationship holds regardless of density. This observation may explain the lower pupal viabilities noted in five of the six pairwise comparisons of Table 2 . To the extent that choice of a high pupation site leads to an increased probability of successful eclosion, this trait may reflect the increased developmental fitness of normal-lived animals.
Body weight is thought to be another indicator of fitness, although it is highly sensitive to environmental conditions (32). It is generally true that normal-lived young adults are heavier than their long-lived counterparts (Table 3 ). This may reflect a more vigorous growth during the larval phase and, to the extent that higher body weight leads to an increased survival, then this trait may also reflect the increased developmental fitness of normal-lived animals. Thus, it appears as if the overall higher developmental fitness characteristic of the normal-lived strains depends on both physiological (larval growth, resistance to diverse stress factors) and behavioral (choice of pupation site) parameters.
Finally, we have shown elsewhere (33) that adults of one of the long-lived strains described here (LHD) are significantly more sensitive to the deleterious effects of heat stress than are adults of its normal-lived (RHD) counterpart. This finding suggests that, under variable temperature conditions, the long lived strain may have a decreased viability throughout the life cycle.
Taken together, these data demonstrate that under variable temperature conditions, the normal lived animal is better able to survive in both the developmental and adult stages than is the long-lived animal. Such a decreased viability might well explain why long-lived animals are not prevalent in the wild (31). In the face of fluctuating environmental conditions likely to prevail in the field, it would seem that generalists have a better chance of surviving than do specialists.
Extended Longevity in the Port Huron Strains
The data in Fig. 2 and Fig. 3 demonstrate that selection for extended longevity is effective even when the stocks are simultaneously selected for fast or slow development. Selection for slow development modulates but does not extinguish selection for extended longevity. This observation immediately suggests that there cannot be a strong causal connection between development time and adult longevity. Nonetheless, the fact that within a particular longevity treatment, the slow developers display a decreased longevity relative to the fast developers implies some sort of limited inverse relationship between these two traits.
Relationship of Developmental Time to Selection for Extended Longevity
Wayne State strains..--
The differences in overall development time in the Wayne State strains do not appear to be robustly associated with the extended longevity phenotype as such (Fig. 1). They appear to be associated with larval density such that all cohorts raised under HD conditions take longer to develop than do their LD counterparts. Within each density treatment, there also appears to be an association with chromosome composition, as a faster development is associated with the presence of an L-type first and second chromosome (002) whereas a slower development is associated with an R-type first and second chromosome (220) (24).
The stage-specific data (Table 1 ) show that the major differences between the several strains occur in the larval stage. An alteration in larval density can bring about as much as a 17% increase in the duration of the larval phase in the same genotype. However, at any given density, there is only a 4–8% change in larval duration associated with normal- or long-lived adults.
Port Huron strains..-- The fast developing strains have statistically identical overall development times, as do the strains selected for slow development (Table 1 and Fig. 4). Not unexpectedly, the strains selected for fast development have shorter development times. Within either set of developmentally selected strains, the long-lived strain develops faster than does the normal-lived strain. However, the difference is statistically significant for the RPHS/LPHS pair but not for the RPHF/LPHF pair. There does not appear to be a robust correlation between developmental duration and adult longevity in these strains either.
We note that the RPHS animals actually showed a decrease in their mean longevity as a result of selection for slow development. The R animals are maintained such that they are under no overt selection pressure for an increased or decreased adult longevity (7). Thus the decreased longevity might well be the indirect consequence of selection for slow development.
Previous studies with Drosophila have shown that increased larval density results in a slower developmental rate (34)(35). Our current results uphold this finding, and there is no dispute that increased larval density is associated with increased developmental time. However, there is still a debate over the association between developmental time and adult longevity. Partridge and Fowler (8) have shown that both their Brighton and Dahomey long-lived "old" lines have longer developmental times as compared with their Brighton and Dahomey "young" lines. Roper and colleagues (36), using the Dahomey "old" and "young" lines of Partridge and Fowler as well as the Dahomey base stock line, not surprisingly came to the same conclusion while addressing a different aspect of aging theory. Interestingly, the work of Roper and colleagues (36) showed that the development time of the Dahomey base stock line was actually longer than either the "old" or "young" line, although this finding was based on only a single replicate value for the base stock. As such, the work of Roper and colleagues (36) showed that there is either no change or a slight decrease in the development time of their "old" line as compared with the base stock Dahomey line, a finding consistent with our current results.
Yonemura and colleagues (37) have demonstrated that extended longevity is associated with a faster developmental time in their strains. In contrast, Economos and Lints (32) have demonstrated that life span and developmental time can be uncoupled by varying the amount of yeast available to developing larvae in their experiments. Their results showed that both the slowest and fastest developing animals had lower adult life spans and that there was some optimum intermediate developmental time where life span was found to be highest. We have also shown that life span and development time can be uncoupled. By using our specially constructed isochromosomal lines 002 and 220, we have shown that the long-lived phenotype associated with the presence of the 002 genotype possesses a developmental time indistinguishable from the normal-lived R strain, whereas the normal-lived 220 strain has a development time similar to that of the long-lived L strain. These observations point to the conclusion that developmental time is not a causal factor in determining adult longevity.
|
Conclusions |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
The key result of this investigation is our finding that the life history trait most obviously associated with all populations of our independently derived long-lived Wayne State and Port Huron strains is a significantly reduced developmental viability, with the long-lived strains having as much as twice the developmental lethality as any of the normal-lived strains. This reduced fitness may well explain why long lived strains are not found in the wild (31). Our observations are supported by the report of Luckinbill (38), in which he finds that sister strains to our long-lived Wayne State strains have a decreased larval viability when raised under low-temperature conditions. Our data suggest that the decreased developmental viability is a general phenomenon and is not peculiar to any one set of developmental conditions. A complete analysis of adult longevity must include some reference to the developmental stages. Generalized theories of stress resistance as a prolongevity trait will have to be modified to take into consideration the increased mortality of the developmental stages, a phenomenon that might be caused by a decreased resistance to some as-yet-undefined stresses. In fact, our recent finding that the long-lived LHD adults are significantly less resistant to heat stress shows that their relatively decreased viability occurs in all stages of the life cycle if the adults are placed under variable temperature conditions (33).
These data suggest that extended longevity, in Drosophila at least, is not a gift but comes with real evolutionary and physiological costs. In both our Wayne State and Port Huron strains, it is instructive to note that the developmental mortality costs are paid not by the long-lived adults but rather by their nonsurviving sibs. Thus the benefits and costs of extended longevity do not necessarily accrue to the same individual, as seems to have been the common assumption. Presumably the parents are compensated for the prereproductive death of some offspring by the increased fecundity of the survivors in the protected laboratory environment. It is also reasonable to presume that this increased fecundity cannot compensate for the increased stress components of the typical nonprotected field environment. Long-lived strains are a laboratory tool.
Finally, a genetic analysis of the genes promoting extended longevity in C. elegans has identified the dauer formation (daf) genes, particulary the daf-two gene, and the age-one gene as being involved (39)(40)(41)(42). The wild-type alleles of both of these mutants code for different portions of an insulinlike signaling system. Of particular interest is the report that the long-lived age-one and daf-two mutants have a decreased brood size and a decreased developmental viability of 6–15% relative to the wild type (41), thus indicating that the inverse association between longevity and developmental viability is not restricted to flies and may be a more general phenomenon. If so, then it will be important in the future to understand if it is possible to separate the benefits of extended longevity from its costs.
|
Acknowledgments |
|---|
Received November 5, 1998
Accepted November 5, 1999
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Vettraino, S. Buck, and R. Arking Direct Selection for Paraquat Resistance in Drosophila Results in a Different Extended Longevity Phenotype J. Gerontol. A Biol. Sci. Med. Sci., October 1, 2001; 56(10): B415 - 425. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | The Gerontologist |
| Journals of Gerontology Series B: Psychological Sciences and Social Sciences | |
. The number within the box is the mean overall development time in hours; the box indicates the standard error of the mean; the error bar indicates the standard deviation; N indicates the total number of individuals that successfully completed all of development. The initial numbers are listed in 
. The number within the box is the mean overall development time in hours; the box indicates the standard error of the mean; the error bar indicates the standard deviation; N indicates the total number of individuals that successfully completed all of development. The initial numbers are listed in 
. The correlation coeficient (r) of .791 
is statistically significant (.05 > p > .02; two-tailed test). Note that if the two slow developers (RPHS and LPHS) are omitted from the analysis, then the r value jumps to .972 
. This might indicate that the two slow developers could be in a class distinguished from the others by an even faster decrease in developmental viability as a function of longevity 
. See the text for further discussion. RPHF, RPHS, normal lived, fast or slow developing, respectively; LPHF, LPHS, long lived, fast or slow developing, respectively.