Genetic factors leading to extinction include evolutionary changes in small populations. If a species with a small surviving population is in close proximity to more abundant congeners with which it can interbreed, a real i, *
Fig. 13.6 Extinction via hybridization. Here a rare species (white circles) has its range surrounded by another species (black circles) with which it can hybridize, reducing its ability to replace itself. Interbreeding occurs with close-neighbours, and eventually, all the reproduction of the rare species is via hybridization.
risk of extinction comes from interspecific hybridization. There are several reasons why this may be a threat to the less abundant species (Levin 2000). First, crossing with other species reduces its potential to replenish its numbers because some reproductive potential is diverted to hybrid offspring rather than pure-bred offspring (Figure 13.6). Because the small population is less effective as a reproductive donor than the large population, the effect on its reproductive potential is much greater. Second, the hybrids might compete successfully against the parents reducing their fitness. Again, the proportional effect is greater in smaller populations. Hybridization can also increase pressure from natural enemies. Many plant hybrids are more susceptible to pest exploitation than their parents and hence can support large pest populations in proximity to their parent species (Levin 2000).
Finally, genetic introgression may be so one-way that the smaller population simply becomes absorbed into the larger one. One likely future extinction from hybridization comes from the Catalina Mahogany tree, Cercocarpus traskiae (Rosaceae) which occurs on Santa Catalina Island off the Californian coast. It is restricted to a single population and is hybridizing extensively with its close relative Cercocarpus betuloides. Eventually, this may lead to total introgression of the mahogany such that no pure-bred plants remain. There are several other botanical examples, but hybrid extinction is also a possibility in animal populations. We encountered one example from the haplochromine cichlids of Lake Victoria (Chapter 1) where mating barriers in sympatry are breaking down due to water turbidity. Another animal example comes from the endangered European White-headed duck, which has been hybridizing extensively with Ruddy ducks, introduced from North America and now considerably more abundant. Just as hybridization can create species, so can it destroy them!
Inbreeding depression is another genetic process that may cause extinction. This occurs when declining populations experience a decline in fitness through increasing homozygosity of deleterious recessive alleles. It is normally expressed through reduced fecundity and offspring viability. Inbreeding depression is readily created experimentally in captive populations. In Drosophila melanogaster, about half the loss in fitness is due to recessive lethal/semilethal mutations at about 5000 loci. The remainder is due to many slightly deleterious alleles that are mildly recessive (Lande 1988). When mating occurs among relatives, the chances of an offspring being homozygous for these recessive alleles are increased, and their deleterious effects are expressed. Inbreeding depression is not an automatic consequence of small population size: gradual inbreeding tends to create little depression of fitness because there is more opportunity for selection to purge the population of homozygotes, thus removing the deleterious alleles. Many insect and plant species consequently inbreed with little sign of fitness depression. In contrast, when a population suddenly declines the effects of inbreeding depression are more marked. As with Allee effects, the loss in fitness can create a positive feedback 'extinction vortex', whereby in a small population inbreeding occurs, fitness is reduced, which reduces population size, which in turn increases inbreeding which reduces fitness and so on.
Ralls etal. (1979) looked for signs of inbreeding depression by measuring juvenile survival in ungulate populations in captivity kept in either inbred or outbred conditions. In 41 of 44 species examined,juvenile survival was lower in the inbred populations. Inbreeding depression has also been inferred from genetic and fitness studies of wild populations. The Chihuahua spruce (Figure 13.7) was once widespread in Mexico but its range has contracted into a few isolated populations of between 15 and 2400 trees. There is very low gene flow between populations, and parents are more closely related than half-sibs would be in a typical outbred population. Forty-five per cent of seeds do not contain embryos, which is likely due to inbreeding depression (Ledig etal. 1997). In some cases population extinction can be attributed to inbreeding. In 42 subpopulations of the Glanville Fritillary (Figure 13.5) in Finland, the risk of extinction was negatively correlated with the proportion of heterozygosity in the populations (Saccheri et al. 1998), and one generation of full sib-mating in the laboratory led to substantial inbreeding depression.
In several wild populations the effects of inbreeding depression have been reversed by transfer of individuals between populations (Hedrick and Kalinowski 2000; Hedrick 2001).The Florida panther (Felis concolor coryi),a subspecies of mountain lion, has been isolated in southern Florida since the 1900s and the current effective population size in recent years has been 25 or less. The males showed a high frequency of undescended testes and deformed sperm. After the introduction of females from Texas in 1995,14 offspring were produced from crosses with the Florida population, with a reduced frequency of sperm defects.
How big must a population be to be safe from inbreeding depression? Franklin (1980) suggested the simple rule that an effective population size of 50 or less is at risk. This was based on very simple logic: in the absence of selection the inbreeding coefficient increases by 1/2Ne per generation due to random genetic drift, where Ne is the effective population size. Therefore, to limit inbreeding to 1% per generation, a level tolerated in domesticated animals, the effective population size should be 50. It is, however, only a rough guide, and as Hedrick (2001) has pointed out, there are several cases of endangered species, such as the Californian Condor, that have had founder numbers of much less than 50 and that have rebounded from the brink of extinction without apparent inbreeding depression.
Two other genetic changes can lower the survival prospects of small populations: loss of genetic variation (and hence evolvability) through drift, and accumulation of deleterious mutations of small effect, also by drift. In small populations, random fluctuations in gene frequency (drift) tend to reduce genetic variation, leading to increased homozygosity and consequent loss of adaptability. The process countering this loss of variation is mutation, such that at any population size an equilibrium level of genetic variation is achieved, which is lower for small populations. Franklin (1980) suggested that an effective population size of 500 was sufficient to maintain genetic variation that balanced the variation in the environment that necessitates future evolutionary potential. This was based on data from Drosophila bristle numbers suggesting that the rate of production of genetic variation in this trait is about one thousandth of the environmental variance. However, Lande (1995) has suggested that this rule of thumb may considerably overestimate the amount of genetic variation maintained. Recent data on Drosophila suggest that only about 10% of the mutational variance is in fact neutral or quasineutral (thus, replacing that lost through drift), 10 times less than assumed above. The population size required to maintain sufficient variation would then be 10 times as great (5,000).
In small populations, slightly deleterious mutations can also become fixed by chance, a source of genetic stochasticity. Individually, these mutations do not affect the risk of extinction, but collectively they can. Lande (1995) found that if no selection is assumed (just mutation and drift),mean time to extinction from such new mutations is an exponential function of population size (Figure 13.4). Thus, as for demographic stochasticity, at reasonably large population sizes (Ne > 100), there is little risk of extinction from genetic stochasticity. However, also assuming selection decreases the mean time to extinction because it becomes a power of (asymptotically related to) population size (Figure 13.4). As a result, genetic stochasticity might be much more important at large population sizes, comparable to environmental stochasticity, making very large populations necessary for long-term viability.
Thus, a range of processes, both evolutionary and ecological, may contribute towards extinction. These work both by making species rare and by causing extinction once rare. Evolution can determine whether a species becomes rare indirectly via selection on the ecological niche, and due to co-evolutionary interactions with other species. Among small populations, a range of evolutionary processes, such as drift, inbreeding depression, and introgression, may finally cause extinction. Many of these may only be important in very small populations, but others, such as loss of fitness due to drift, may be active in comparatively large populations. In the next chapter, we examine in more detail the characteristics of taxa that might be associated with extinction, speciation, and species richness.
Was this article helpful?