HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kirk, K. L. Search for Related Content PubMed PubMed Citation Articles by Kirk, K. L.

Dietary Restriction and Aging

Comparative Tests of Evolutionary Hypotheses

^a Department of Biology, New Mexico Tech, Socorro

Kevin L. Kirk, Department of Biology, New Mexico Tech, Socorro, NM 87801 E-mail: klkirk{at}nmt.edu.

Decision Editor: John A. Faulkner, PhD

	Abstract

Top Abstract Methods Results Discussion References

 
Dietary restriction (DR) increases life span in many types of animals. The response to chronic DR may be an adaptation to environments with variable food levels. This study uses the comparative method to test evolutionary predictions about the origin of the response to DR, using data from 10 species of rotifers. Most species, but not all, responded to DR by increasing mean life span, maximum life span, reproductive life span, mortality rate doubling time, and initial mortality rate. Interspecific comparisons did not show the predicted correlations between the strength of the response to DR and either reproductive life span, age of first reproduction, or total reproduction. There was support for the idea that the response to chronic DR is associated with changes in reproductive allocation during short-term periods of starvation: species that reduced reproduction when starved increased their life spans under DR, whereas species that continued to reproduce when starved decreased their life spans under DR.

DIETARY restriction (DR) increases life span and slows aging in many species of animals ⁽¹⁾. It is the only known environmental manipulation that consistently prolongs life in mammals ⁽²⁾. In rodents, DR slows age-related deterioration and delays the onset of many diseases ⁽³⁾⁽⁴⁾. DR also extends life in several other animal taxa, including fish, cladocerans, spiders, nematodes, and rotifers ^{(5)(6)(7)(8)(9)}. In many animals, DR not only increases mean and maximum life span but also slows the rate of increase in mortality rate; that is, it increases the mortality rate doubling time.

Although a great deal of experimental work has been done on the physiological and molecular mechanisms for the antiaging action of DR (reviewed in refs. 1–4), relatively little work has addressed the evolutionary reasons for the response. There have been a few theoretical attempts to explain the evolution of the response to DR. Harrison and Archer ⁽¹⁰⁾ proposed that the response evolved as an adaptation to variable food levels. The ability to slow aging when food levels are low should postpone reproductive senescence and therefore allow reproduction when food levels become high. Harrison and Archer predicted that the response to DR should be strongest in species with short reproductive life spans, because these species should be more likely to encounter periods of low food that are longer than their normal reproductive life spans. In contrast, Phelan and Austad ⁽¹¹⁾ noted that reproductive senescence may be uncommon in nature and suggested that the increase in life span in response to DR is not an adaptation to variable food levels but is instead a "secondary consequence of its effect in delaying age at maturity and decreasing the subsequent rate of reproduction." They predicted that the response to DR should be strongest in those species with early maturities and high reproductive rates.

Holliday ⁽¹²⁾ returned to the idea that the antiaging action of DR is an adaptation to life in environments with variable food levels. Species that stop reproduction during episodes of starvation can instead allocate resources to maintenance activities, and thus slow the rate of aging. Thus, a change in allocation strategy may be the most important adaptation, and life span may be extended because of allocation-based trade-offs between fecundity and survival. The effects of this trade-off on the evolution of aging and life span are known as the disposable soma theory of aging ^(13)(14). Masoro and Austad ⁽¹⁵⁾ elaborated on Holliday's ideas and suggested that the response to long-term DR evolved as an adaptation to short-term periods of food shortage. They predicted that the response to chronic DR should be strongest in species that experience more frequent episodes of short-term food shortages in nature. Thus, theories for the evolutionary origin of the antiaging effect of DR predict that there should be interspecific variation in the strength of the response to DR. Comparative studies have been useful in addressing questions about the evolution of aging (e.g., Ref. ⁽¹⁶⁾), and they should also be useful in addressing questions about the evolution of the antiaging effect of DR.

Despite broad interest in the effects of DR, none of the evolutionary theories for its origin have been tested. The ecological literature contains a great deal of information that can be used in comparative tests of evolutionary hypotheses. Indeed, knowledge of the physiological and demographic responses to food levels forms the mechanistic basis for understanding resource competition, one of the most important organizing concepts in ecology ⁽¹⁷⁾. Although ecological experiments are seldom conducted with life span or aging in mind, they provide a useful and largely unexploited set of data on the effects of DR. In this paper, I quantitatively review the effects of DR on 10 species of rotifers, which are small multicellular aquatic animals. I use original and published data, primarily from the ecological literature, to search for patterns in how species respond to DR and to test hypotheses regarding the evolutionary origin of the response to DR. Specifically, I test the hypotheses that interspecific comparisons should reveal (a) a negative correlation between reproductive life span and change in life span under DR, (b) a negative correlation between age of first reproduction (age at maturity) and change in life span under DR, (c) a positive correlation between total reproduction and change in life span under DR, and (d) an association between the response to chronic DR and the response to short-term periods of starvation.

	Methods

Top Abstract Methods Results Discussion References

 
The comparative data set was assembled from two sources: original experimental data on the responses of two rotifer species to DR, and data from published studies. Original data were collected for Brachionus calyciflorus and Synchaeta pectinata. Cohort life table experiments were conducted as described in detail elsewhere ⁽¹⁸⁾. Cohorts were initiated by using 30–82 neonates that were <12 hours old. Each individual was contained in a well of a tissue culture plate containing 0.25 ml of a defined medium plus a known concentration of phytoplankton food. Animals were kept at 20°C and transferred to new medium and food daily; survival and fecundity were noted twice daily. Data concerning survival from birth have not been previously published. Fecundity data are also new, with the exception of fecundity data for the highest food level for both species; these data have been previously published ⁽¹⁸⁾ and are shown here for comparison. As with most rotifers, these species are able to reproduce parthenogenically, so experiments were conducted solely on females. The cohorts were initially intended as controls for comparison to starved cohorts, and some control cohorts were discontinued before all of the individuals had died ⁽¹⁸⁾. Because of this, the data for B. calyciflorus could not be used to determine mean or maximum life span, but the data were used to estimate initial mortality rate and mortality rate doubling time. Three other studies provided life span data for B. calyciflorus (see text that follows).

Published studies reported on cohort life table experiments similar to those described here. In all cases, cohorts were exposed to different food levels from birth. Experiments were conducted at temperatures ranging from 19–25°C. Publications were found on the effects of different food levels on the demography of 10 rotifer species: Asplanchna brightwelli ⁽⁹⁾, A. girodi ⁽¹⁹⁾, B. calyciflorus (Ref. ⁽²⁰⁾, short-spined morph; ⁽²¹⁾, 22, experiment 1), B. patulus ⁽²³⁾, B. plicatilis ⁽²⁴⁾, Encentrum linnhei ⁽²⁴⁾, Euclanis dilatata (ref. 25, fed Chlamydomonas), Keratella testudo (Ref. ⁽²⁶⁾, unspined morph), Philodina acuticornis ⁽²⁷⁾, and S. pectinata (Ref. 22, experiments 5 and 6).

Some studies only provided data on mean life span (LS), maximum life span (LS_max, the age of the oldest individual when it died), or both. Most studies also provided data on age-specific survival (l_x) as a function of age (x), from which other parameters were calculated. The Gompertz equation, a useful way to describe the increase in age-specific mortality rate (m_x) with age, takes the form

where A is the initial mortality rate (IMR) and G is the mortality rate coefficient ⁽²⁾. One method for estimating the parameters of the Gompertz equation is to fit the equation to demographic data by using regression techniques (e.g., Ref. 2). However, regression methods can result in biased parameter estimates when cohort size is small, say, <100 ^(28)(29). In the studies considered here, cohort size ranged from 10 to 82, so regression methods were not used. Instead, maximum likelihood methods were used to estimate the Gompertz parameters; these methods provide parameter estimates that are much less biased than those provided by standard regression methods when cohort size is small ^(28)(29). Maximum likelihood estimates of Gompertz parameters were calculated by using WinModest software ⁽²⁸⁾. Data were entered in the form of the number of deaths during each age interval (d_x), which was calculated from l_x data by using the equation

where n₀ is the initial cohort size. The maximum likelihood method correctly accounts for censored individuals, such as the case in which a cohort was terminated prior to the death of all individuals (e.g., the original B. calyciflorus data from this study). The mortality rate doubling time (MRDT) is a useful way of expressing the rate of senescence and was calculated as MRDT = ln 2/G ⁽²⁾.

It is important to know whether the increased life span under DR results mainly from a delay in juvenile development and thus represents merely an extension of prereproductive life span, or whether reproductive or postreproductive life spans also increase. Most studies provided data on age-specific fecundity (b_x). The duration of the prereproductive (juvenile) period was estimated as the age of first reproduction (AFR), defined here as the age when b_x first became greater than zero. Because rotifers have little or no postreproductive life span, reproductive life span (LS_rep) was calculated as LS_rep = LS - AFR. The total fecundity of each species was defined as the net reproductive rate, the mean total number of offspring produced per female, R₀ = (l_x b_x). For l_x and b_x data to be obtained from published studies, figures were digitized and data were read by using a calibrated graphics program.

Most of the rotifer species studied are herbivorous suspension-feeders, so food level was controlled by varying food concentration. I have converted all food concentrations to units of micrograms of dry mass of phytoplankton prey per milliliter. Asplanchna spp. are carnivorous and were fed protozoan or rotifer prey. A. girodi was fed defined rations of prey ⁽¹⁹⁾, with units of micrograms of dry mass of prey eaten per day.

Studies were included in the comparative analysis only if they included food levels that corresponded to near ad-libitum ingestion rates, as well as food levels that corresponded to significant DR. Ingestion rates (rations) were given only for A. girodi. Food concentrations were therefore interpreted relative to the ecological requirements of each species. This was done by using the concept of numerical response, borrowed from the ecology of resource competition. The numerical response is the relationship between the rate of exponential population growth (r, a measure of fitness) and food concentration (R); a hypothetical example is given in Fig. 1. There is a threshold food concentration (R*) where r = 0. Because population growth is severely limited, food concentrations near R* must be associated with ingestion rates that are far below ad-libitum levels. Population growth rate reaches a maximum (r_max) gradually, so it is difficult to estimate the food concentration where r is maximized. Therefore, the food concentration (R₉₀) where r is 90% of r_max was taken to indicate a food concentration likely to be associated with ingestion rates close to ad-libitum rates. An intermediate food concentration is given by K_s, where r is 50% of r_max (Fig. 1). Because r is food limited at K_s, I assumed that food concentrations near K_s corresponded to significant DR. The shapes of the numerical response curves of many rotifer species are known ^(30)(31), allowing R*, K_s, and R₉₀ to be estimated directly. When the numerical response of a species was unknown, R*, K_s, and r_max were estimated by using allometric relationships between the parameters and body mass ⁽³¹⁾. With the assumption that the numerical response curve takes the form of a Monod function with a threshold, R₉₀ could then be estimated. The allometric regressions between each parameter and body mass were not perfect (correlation coefficients for R*, K_s, and r_max were 0.82, 0.87, and 0.73, respectively; Ref. 31), so some parameter estimates are approximate. However, numerical response parameters were used only to interpret experimental food levels in relation to the approximate ad-libitum and DR food levels, and to discard studies that did not consider a sufficient range of food levels.

View larger version (19K): [in this window] [in a new window]   Figure 1. An illustrative numerical response curve, showing the relationship between population growth rate (r) and food level (R). Here rmax is the maximum population growth rate; the inverted triangles on the x axis show the three food levels indicated on Fig. 3 for each rotifer species: R* (threshold food level where r = 0), Ks (food level where r = 50% rmax), and R90 (food level where r = 90% rmax).

	Results

Top Abstract Methods Results Discussion References

 
The effects of food level on the age-specific survival and fecundity of B. calyciflorus and S. pectinata are shown in Fig. 2. For both species, AFR (the age when fecundity first becomes greater than zero) increased at low food levels (Fig. 2). Net reproductive rate (R₀) increased monotonically as food level increased. For B. calyciflorus, R₀ increased from zero to 5.8 offspring/female as food concentration increased from 1µg/ml to 14 µg/ml. For S. pectinata, R₀ increased from 0.6 to 7.7 offspring/female as food concentration increased from 1 µg/ml to 16 µg/ml.

View larger version (30K): [in this window] [in a new window]   Figure 2. Age-specific survival (lx, proportion of cohort surviving) and age-specific fecundity (bx, offspring/female/day) for two species of rotifers, Brachionus calyciflorus and Synchaeta pectinata, at different food concentrations (µg/ml).

View larger version (45K): [in this window] [in a new window]   Figure 3. Response of 10 species of rotifers to dietary restriction. Parameters shown are mean and maximum life span (days), mortality rate doubling time (MRDT; days) and initial mortality rate (IMR; day-1). Food levels are given as either food concentration (µg/ml) or ration (µg/day). Sources of data are indicated in the text in Methods. Reference numbers are noted where data were taken from more than one reference for a given species; T = original data from this study. Inverted triangles on life span x axes denote food levels as defined in Fig. 1: R*, Ks, and R90 (left to right).

There were exceptions to the general pattern of DR's increasing life span. Both mean and reproductive life span of B. plicatilis decreased under DR, but no data are available on the response of MRDT. In K. testudo, DR decreased mean life span, but maximum life span did not change and MRDT increased (Fig. 3 and Table 1 ). S. pectinata was the only species to show a consistent response opposite to the general pattern. Original data from this study show that DR decreased mean life span, maximum life span, reproductive life span, and MRDT (Fig. 2 and Fig. 3, and Table 1 ). In contrast, a previous study ⁽²²⁾ found little effect of food on the mean life span of S. pectinata.

Nearly all species for which data were available showed an increase in IMR under DR (Fig. 3), and this change was often large (19–902%; Table 1 ). Only A. brightwelli decreased IMR under DR, but IMR was extremely small (<10^-4 day^-1) in this species at all food levels (Fig. 3). In A. girodi, the response of IMR to food level was unimodal; IMR increased as food level declined from high to moderate levels, and then decreased as food level declined to very low levels (Fig. 3).

The response to DR was often nonlinear, with the greatest changes occurring at low food levels. This pattern was observed for mean life span in A. girodi, and for MRDT in B. calyciflorus (data from this study), B. patulus, E. dilatata and K. testudo (Fig. 3). A nonlinear response to declining food levels was also observed for IMR in B. calyciflorus, B. patulus, E. dilatata, K. testudo, and S. pectinata (Fig. 3). Some species showed a nonmonotonic, unimodal response to food level. For example, in three separate studies, the mean or maximum life span of B. calyciflorus increased as food declined from high to moderate levels and then decreased at the lowest food levels (Fig. 3). The same pattern was observed for MRDT in B. calyciflorus (data from this study) and A. girodi (Fig. 3).

Interspecific comparisons were used to test theoretical predictions concerning the evolution of the response to DR. Contrary to the first hypothesis (see the first several paragraphs of this paper), there was not a negative correlation between reproductive life span (LS_rep) and the strength of the response to DR. There were no significant correlations between LS_rep and the effect of DR on either mean life span, maximum life span, or MRDT (Table 2 ). Contrary to the second hypothesis, there was not a negative correlation between AFR and the strength of the response to DR. Instead, there was a significant positive correlation between AFR and change in mean life span (Table 2 ). Contrary to the third hypothesis, there was not a positive correlation between total reproduction (R₀) and the strength of the response to DR (Table 2 ).

	Discussion

Top Abstract Methods Results Discussion References

The responses of rotifers to DR differed in some respects from the responses of rodents. Finch ⁽²⁾ quantitatively reviewed 14 studies of DR in rats. The average increase in maximum life span was 33%, whereas the average increase in MRDT was 74%. Thus, the magnitude of the response to DR was greater for some rotifer species than for rats. In addition, mice showed a roughly linear increase in life span as food intake decreased ⁽³²⁾, whereas some rotifer species showed strong nonlinearities in their response to reductions in food level.

There were exceptions to the general pattern of increasing life span and MRDT under DR. Three species (B. plicatilis, K. testudo, and S. pectinata) responded to DR by decreasing either life span, MRDT, or both. Very low food levels, near or below R*, may result in malnutrition. Thus, malnutrition may have contributed to the reductions in mean life span for K. testudo. However, malnutrition cannot account for the reductions in life span or MRDT for B. plicatilis or S. pectinata. In these two species, life span or MRDT declined even as food concentration declined from high to moderate levels.

For most rotifer species, IMR was close to zero at high food levels and increased greatly at lower food levels. The increase in IMR may have been due to an increase in juvenile mortality at low foods. For example, in S. pectinata (data from this study), juvenile mortality (percentage of juveniles that died before AFR) increased from 6% at a food level of 16 µg/ml to 84% at 1 µg/ml. In addition, the response of IMR was stronger in most rotifer species than in rats. DR increased the IMR of rats by an average of 110% ⁽²⁾, whereas DR increased the IMR of rotifers by 19–902%. One possible reason for the stronger response of rotifers is that rotifers were subjected to DR from birth, whereas DR was initiated after puberty in the rat studies. In A. girodi, the response of IMR to food level was unimodal; the mechanism for this unimodal response is not known.

In their theories for the evolution of the response to DR, Harrison and Archer predicted a negative correlation of reproductive life span and the response to DR ⁽¹⁰⁾, whereas Phelan and Austad predicted a negative correlation between AFR and the response to DR ⁽¹¹⁾. These two hypotheses are not mutually exclusive. Also, because comparative studies of many animal taxa often show a positive correlation between life span and AFR ⁽³³⁾, the two hypotheses may not be independent. However, in the present study, there was no significant correlation between reproductive life span and AFR (Pearson correlation coefficient r_p = -.03; p = 0.94, n = 10), so the two hypotheses are considered separately.

To my knowledge, this study reports on the first comparative tests of theories for the evolution of the antiaging effect of DR. Most of these theories were initially framed in terms of rodents, not rotifers, but there is no reason to expect that the predictions should be restricted to rodents. Most of the specific predictions of evolutionary theories were not supported. The comparative data did not support the hypothesis of Harrison and Archer ⁽¹⁰⁾ that species with shorter reproductive life spans should have the strongest responses to DR. Harrison and Archer's hypothesis was based on the assumption that short-lived species should be more likely than long-lived species to encounter periods of scarce food that are longer in duration than their reproductive life spans. Perhaps this assumption is incorrect. We do not know enough about the durations of food shortages in nature, and it is possible that they are typically longer than (or shorter than) the reproductive life spans of all rotifer species.

The predictions of Phelan and Austad ⁽¹¹⁾ were based on the idea that the response to DR is simply a function of reduced juvenile development rate and reduced reproduction. The comparative rotifer data did not support their hypothesis that species with early maturity (low AFR) should have a stronger response to DR. A more direct test of the idea that a prolonged juvenile period is responsible for the increased life span under DR is possible by looking for a positive correlation between changes in AFR and changes in life span. There was not a significant correlation between percent change in AFR and percent change in either mean life span (r_p = .62, n = 7, p = .14), maximum lifespan (r_p = .60, n = 5, p = .29) or MRDT (r_p = .33, n = 5, p = .60). In addition, although DR prolonged the juvenile period, it also increased the reproductive life span of most species. Phelan and Austad's other hypothesis, that the response to DR should be positively correlated with total reproduction, was also rejected.

To test the hypothesis that the response to chronic DR is associated with the response to short-term episodes of starvation ^(12)(15), we need to know how rotifers react to food deprivation. Kirk ⁽¹⁸⁾ tested the response of nine species of rotifers to starvation. Animals were deprived of food, beginning when they were young adults. Starvation time varied greatly between species, ranging from 0.4 to 5.0 days. Allometric theory, based on the relationship between body mass, respiration rate and energy storage, predicts that larger species should have longer starvation times. However, there was no correlation between body mass and starvation time. Instead, the primary factor responsible for interspecific variation in starvation time was whether or not the species curtailed reproduction when starved. Species that reduced or eliminated egg production when starved had relatively long starvation times, whereas species that maintained egg production had shorter starvation times. Two species in the genus Synchaeta increased reproductive rates when starved; these species had very low starvation times (0.4–0.7 days) and effectively reproduced themselves to death. Interspecific comparisons showed that starvation time was negatively correlated with the rate of reproduction during starvation (relative to fed controls). This is an example of an interspecific trade-off between reproduction and survival. Such trade-offs are often easiest to detect under stringent conditions such as food shortage or starvation ⁽³³⁾.

Data are available on the response to both starvation and chronic DR for two rotifer species, B. calyciflorus and S. pectinata. When starved, B. calyciflorus reduced its reproductive rate to 44% of that of fed controls, whereas S. pectinata increased its reproductive rate to 160% of that of fed controls ⁽¹⁸⁾. As predicted, the response to starvation was associated with the response to chronic DR. B. calyciflorus, which decreased its reproductive allocation when starved, increased its life span and MRDT in response to DR (Fig. 3 and Table 1 ). In contrast, S. pectinata, which increased its reproductive allocation when starved, reduced its life span and MRDT in response to DR. These data provide the first comparative evidence in support of the idea that how species respond to starvation, particularly in terms of reproductive allocation, can predict how species will respond to chronic DR ^(12)(15).

There are additional differences in the way that B. calyciflorus and S. pectinata respond to changes in food level. B. calyciflorus reduced the size of its eggs in response to chronic DR ⁽³⁴⁾ and reduced its respiration rate when starved ⁽³⁵⁾. In contrast, S. pectinata did not change egg size or respiration rate in response to DR or starvation. This provides further evidence that the allocation patterns of S. pectinata are relatively inflexible.

Evolutionary theories that seek to explain the response to DR as an adaptation to variable food levels make sense only if food levels actually vary in nature. The primary food sources for many species of planktonic rotifers are small, flagellated phytoplankton cells such as cryptomonads ⁽³⁶⁾, and the abundance of cryptomonads can vary rapidly ⁽³⁷⁾. Field experiments, in which naturally available foods were supplemented with additional foods, have shown that rotifers were nearly always food limited ^(38)(39). The intensity of food limitation was variable over time, with large changes occurring on time scales of one to a few weeks. Food limitation was sometimes intense, indicating that rotifers may face periods of starvation in nature, when food level or nutritional quality are less than physiological requirements. Thus, many rotifers experience chronic DR in nature, food levels may change drastically within the life span of a rotifer, and rotifers may experience episodes of starvation. Unfortunately, we do not yet know enough about the temporal patterns of food limitation of rotifer species to test Masoro and Austad's ⁽¹⁵⁾ hypothesis that the response to DR should be strongest in those species that experience more frequent episodes of food shortage in nature.

In addition to differences in the variability of food levels, it is possible that differences in the response to DR might be related to other ecological differences between species. Three ecological variables that may be relevant are habitat, food, and reproductive mode. Most of the species considered are planktonic and swim in the open water, except E. dilatata and P. acuticornis, which attach themselves to solid substrates in shallow waters ^(25)(40). Most species occur in freshwaters, except B. plicatilis and E. linnhei, which occur in brackish waters ^(24)(40). Most species ingest phytoplankton, bacteria, or both, except for the two Asplanchna spp., which are carnivorous ^(36)(40). Except for P. acuticornis, which always reproduces by parthenogenesis, most species reproduce by cyclical parthenogenesis, in which females usually produce clonal female offspring by parthenogenesis but occasionally produce males and undergo sexual reproduction ⁽⁴⁰⁾. None of these ecological variables has any relationship to the strength of the response to DR.

This comparative study does not control for phylogenetic effects. Because closely related species are likely to be similar to one another, species are not independent statistical replicates. This results in an artificially high degree of freedom in statistical tests and increases the chance of finding a significant correlation when in fact none exists ⁽⁴¹⁾. Some believe that phylogenetic control is essential ⁽⁴¹⁾, although others disagree (e.g., Refs. ⁽⁴²⁾,⁽⁴³⁾). In any case, phylogenetic control was impossible in this study because the phylogeny of rotifer species is not known. Because this study compares results that were obtained from different laboratories, some of the variation in results may be due to between-laboratory differences in experimental methods.

Another way to explore the evolution of the antiaging action of DR is to use theoretical modeling. Shanley and Kirkwood ⁽⁴⁴⁾ formulated a dynamic programming model that determined the optimal allocation of resources to reproduction and somatic maintenance in mice fed various food levels. The model predicted that, in environments where animals are subjected to episodes of famine, optimal allocation to maintenance, and thus life span, will increase under DR. This supports the hypothesis that the response to chronic DR is an adaptation to life in environments with variable food levels ^(12)(15). Our future understanding of DR will be enhanced by additional quantitative theories of the evolution of the antiaging effect of DR, by further comparative and experimental tests of these theories, and by knowledge of how food levels vary in nature.

	Acknowledgments

 
This work was funded by National Science Foundation Grant DEB-9407241. I thank Thomas Kieft and an anonymous reviewer for comments on the manuscript. Scott Pletcher (Max Planck Institute for Demographic Research) kindly provided his WinModest software.

Received July 25, 2000

Accepted October 23, 2000

	References

Top Abstract Methods Results Discussion References

 

1. Weindruch R, Walford RL, 1988. The Retardation of Aging and Disease by Dietary Restriction Charles C Thomas, Springfield, IL.
2. Finch CE, 1990. Longevity, Senescence, and the Genome University of Chicago Press, Chicago.
3. Masoro EJ, 1993. Dietary restriction and aging. J Am Geriatr Soc 41:994-999. [Medline]
4. Sohal RS, Weindruch R, 1996. Oxidative stress, caloric restriction, and aging. Science 273:59-63. [Abstract]
5. Comfort A, 1963. Effect of delayed and resumed growth on the longevity of a fish (Lebistes reticulatus Peters) in captivity. Gerontologia (Basel) 8:150-155.
6. Lynch M, Ennis R, 1983. Resource availability, maternal effects, and longevity. Exp Gerontol 18:147-165. [Medline]
7. Austad SN, 1989. Life extension by dietary restriction in the bowl and doily spider, Frontinella pyramitela. Exp Gerontol 24:83-92. [Medline]
8. Klass MR, 1977. Aging in the nematode Caenorhabditis elegans: major biological and environmental factors influencing life span. Mech Ageing Dev 6:413-429. [Medline]
9. Sawada M, Carlson JC, 1987. Association between lipid peroxidation and life-modifying factors in rotifers. J Gerontol 42:451-456. [Medline]
10. Harrison DE, Archer JR, 1988. Natural selection for extended longevity from food restriction. Growth Dev Aging 52:65
11. Phelan JP, Austad SN, 1989. Natural selection, dietary restriction, and extended longevity. Growth Dev Aging 53:4-6. [Medline]
12. Holliday R, 1989. Food, reproduction and longevity: is extended life span of caloric-restricted animals an evolutionary adaptation?. Bioessays 10:125-127. [Medline]
13. Kirkwood TBL, 1977. Evolution of ageing. Nature 270:301-304. [Medline]
14. Kirkwood TBL, Holliday R, 1979. The evolution of ageing and longevity. Proc R Soc London Ser B 205:531-546. [Medline]
15. Masoro EJ, Austad SN, 1996. The evolution of the antiaging action of dietary restriction: a hypothesis. J Gerontol Biol Sci 51A:B387-B391. [Abstract]
16. Kirkwood TBL, 1985. Comparative and evolutionary aspects of longevity. Finch CE, Schneider EL, , ed.Handbook of the Biology of Aging 2nd ed. 27-44. Van Nostrand Reinhold, New York.
17. Grover JP, 1997. Resource Competition Chapman and Hall, New York.
18. Kirk KL, 1997. Life-history responses to variable environments: starvation and reproduction in planktonic rotifers. Ecology 78:434-441.
19. Robertson JR, Salt GW, 1981. Responses in growth, mortality, and reproduction to variable food levels by the rotifer, Asplanchna girodi. Ecology 62:1585-1596.
20. Stemberger RS, 1990. Food limitation, spination, and reproduction in Brachionus calyciflorus. Limnol Oceanogr 35:33-44.
21. Guisande C, Galindo MD, Galan FM, Oliveros F, 1993. The cost of reproduction in the rotifer Brachionus calyciflorus Pallas. Int Rev Hydrobiol 78:493-499.
22. Gilbert JJ, 1996. Effect of food availability on the response of planktonic rotifers to a toxic strain of the cyanobacterium Anabaena flos-aquae. Limnol Oceanogr 41:1565-1572.
23. Sarma SSS, Rao TR, 1991. The combined effects of food and temperature on the life history parameters of Brachionus patulus Muller (Rotifera). Int Rev Hydrobiol 76:225-239.
24. Schmid-Araya JM, 1991. The effect of food concentration on the life histories of Brachionus plicatilis (O. F. M.) and Encentrum linnhei Scott. Arch Hydrobiol 121:87-102.
25. King CE, 1967. Food, age, and the dynamics of a laboratory population of rotifers. Ecology 48:111-128.
26. Stemberger RS, 1988. Reproductive costs and hydrodynamic benefits of chemically-induced defenses in Keratella testudo. Limnol Oceanogr 33:593-606.
27. Meadows ND, Barrows CH, 1971. Studies on aging in a bdelloid rotifer. I. The effect of various culture systems on longevity and fecundity. J Exp Zool 176:303-314. [Medline]
28. Pletcher SD, 1999. Model fitting and hypothesis testing for age-specific mortality data. J Evol Biol 12:430-439.
29. Promislow DEL, Tatar M, Pletcher S, Carey JR, 1999. Below-threshold mortality: implications for studies in evolution, ecology and demography. J Evol Biol 12:314-328.
30. Stemberger RS, Gilbert JJ, 1984. Body size, ration level, and population growth in Asplanchna. Oecologia 64:355-359.
31. Stemberger RS, Gilbert JJ, 1985. Body size, food concentration, and population growth in planktonic rotifers. Ecology 66:1151-1159.
32. Weindruch R, Walford RL, Fligiel S, Guthrie D, 1986. The retardation of aging by dietary restriction in mice: longevity, cancer, immunity and lifetime energy intake. J Nutr 116:641-654.
33. Stearns SC, 1992. The Evolution of Life Histories Oxford University Press, Oxford.
34. Kirk KL, 1997. Egg size, offspring quality and food level in planktonic rotifers. Freshw Biol 37:515-521.
35. Kirk KL, Ellis J, Taylor J, 2000. Physiological responses to variable environments: storage and respiration of starving rotifers. Freshw Biol 42:637-644.
36. Bogdan KG, Gilbert JJ, 1987. Quantitative comparison of food niches in freshwater zooplankton: a multi-tracer cell approach. Oecologia 72:331-340.
37. Stewart AJ, Wetzel RG, 1986. Cryptophytes and other microflagellates as couplers in planktonic community dynamics. Arch Hydrobiol 106:1-19.
38. González MJ, Frost TM, 1992. Food limitation and seasonal population declines of rotifers. Oecologia 89:560-566.
39. Merriman JL, Kirk KL, 2000. Temporal patterns of resource limitation in natural populations of rotifers. Ecology 81:141-149.
40. Wallace RL, Snell TW, 1991. Rotifera. Thorp JH, Covich AP, , ed.Ecology and Classification of North American Freshwater Invertebrates 187-248. Academic Press, New York.
41. Harvey PH, Pagel MD, 1991. The Comparative Method in Evolutionary Biology Oxford University Press, Oxford.
42. Charnov E, 1993. Life History Invariants Oxford University Press, Oxford.
43. Brown JH, 1995. Macroecology University of Chicago Press, Chicago.
44. Shanley DP, Kirkwood TBL, 2000. Calorie restriction and aging: a life history analysis. Evolution 54:740-750. [Medline]

This article has been cited by other articles:

				K. Hopkin Dietary Drawbacks Sci. Aging Knowl. Environ., February 26, 2003; 2003(8): ns4 - 4. [Abstract] [Full Text] [PDF]

				E. J. Masoro Subfield History: Caloric Restriction, Slowing Aging, and Extending Life Sci. Aging Knowl. Environ., February 26, 2003; 2003(8): re2 - 2. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kirk, K. L. Search for Related Content PubMed PubMed Citation Articles by Kirk, K. L.