2. BIOLOGY AND STARS: IS THERE A LINK?

Evidently, in our solar system life first evolved quite soon after the formation of a hospitable terrestrial environment. Suppose the typical time that it takes for life to evolve is denoted by some timescale t_bio, then from the evidence presented by the solar system, which is about 4.6 x 10⁹ yrs old, it is seems that

At first sight we might assume that the microscopic biochemical processes and local environmental conditions that combine to determine the magnitude of t_bio are independent of the nuclear astrophysical and gravitational processes that determine the typical stellar main sequence lifetime t_ms. However, this assumption leads to the striking conclusion that we should expect extraterrestrial forms of life to be exceptionally rare [10], [4], [11]. The argument, in its simplest form, is as follows. If t_bio and t_* are independent then the time that life takes to arise is random with respect to the stellar timescale t_*. Thus it is most likely that either t_bio >> t_* or that t_bio << t_*. Now if t_bio << t_* we must ask why it is that the first observed inhabited solar system (that is, us) has t_bio t_*. This would be extraordinarily unlikely. On the other hand, if t_bio >> t_* then the first observed inhabited solar system (us) is most likely to have t_bio t_* since systems with t_bio >> t_* have yet to evolve. Thus we are a rarity, one of the first living systems to arrive on the scene. Generally, we are led to a conclusion, an extremely pessimistic one for the SETI enterprise, that t_bio >> t_*.

In order to escape from this conclusion we have to undermine one of the assumptions underlying the argument that leads to it. For example, if we suppose that t_bio is no independent of t_* then things look different. If t_bio /t_* is a rising function of t_* then it is actually likely that we will find t_bio t_*. Livio [12] has given a simple model of how it could be that t_bio and t_* are related by a relation of this general form. He takes a very simple model of the evolution of a life-supporting planetary atmosphere like the Earth's to have two key phases which lead to its oxygen content:

Phase 1: Oxygen is released by the photodissociation of water vapour. On Earth this took 2.4 x 10⁹ yr and led to an atmospheric O₂ build up to about 10^-3 of its present value.

Phase 2: Oxygen and ozone levels grow to about 0.1 of their present levels. This is sufficient to shield the Earth's surface from lethal levels of ultra-violet radiation in the 2000-3000 Å band (note that nucleic acid and protein absorption of ultra-violet radiation peaks in the 2600-2700 Å and 2700-2900 Å bands, respectively). On Earth this phase took about 1.6 x 10⁹ yr.

Now the length of Phase 1 might be expected to be inversely proportional to the intensity of radiation in the wavelength interval 1000-2000 Å, where the key molecular levels for H₂O absorption lie. Studies of stellar evolution allow us to determine this time interval and provide a rough numerical estimate of the resulting link between the biological evolution time (assuming it to be determined closely by the photodissociation time) and the main sequence stellar lifetime, with [12]

where t_sun is the age of the Sun.

This model indicates a possible route to establishing a link between the biochemical timescales for the evolution of life and the astrophysical timescales that determine the time required to create an environment supported by a stable hydrogen burning star. There are obvious weak links in the argument. It provides on a necessary condition for life to evolve, not a sufficient one. We know that there are many other events that need to occur before life can evolve in a planetary system. We could imagine being able to derive an expression for the probability of planet formation around a star. This would involve many other factors which would determine the amount of material available for the formation of solid planets with atmospheres at distances which permit the presence of liquid water and stable surface conditions. Unfortunately, we know that there were many ``accidents'' of the planetary formation process in the solar system which have subsequently played a major role in the existence of long-lived stable conditions on Earth, [13]. For example, the presence of resonances between the precession rates of rotating planets and the gravitational perturbations they feel from all other bodies in their solar system can easily produce chaotic evolution of the tilt of a planet's rotation axis with respect to the orbital plane of the planets over times must shorter than the age of the system [14], [13]. The planet's surface temperature variations, insolation levels, and sea levels are sensitive to this angle of tilt. It determines the climatic differences between what we call ``the seasons''. In the case of the Earth, the modest angle of tilt (approximately 23 degrees) would have experienced this erratic evolution had it not been for the presence of the Moon [15], [13]. The Moon is large enough for its gravitational effects to dominate the resonances which occur between the Earth's precessional rotation and the frequency of external gravitational perturbations from the other planets. As a result the Earth's tilt wobbles only by a fraction of a degree around 23° over hundreds of thousands of years. Enough perhaps to cause some climatic change, but not catastrophic for the evolution of life.

This shows how the causal link between stellar lifetimes and biological evolution times may be rather a minor factor in the chain of fortuitous circumstances that must occur if habitable planets are to form and sustain viable conditions for the evolution of life over long periods of time. The problem remains to determine whether he other decisive astronomical factors in planet formation are functionally linked to the surface conditions needed for biochemical processes.