133845.pdf

100 JOSEPH P. BOTTING

some extent, origination of novel life-habits

provides additional subsidiary niches, most

obviously in the development of epifaunal

tiering. Only through understanding detailed

diversity patterns can these influences possibly

be separated, a procedure that has been advo-

cated strongly by Miller (1997a, 1998, 2000).

The approach of Sepkoski (e.g. 1993) and

most others to global diversity changes has been

that of large-scale datasets, at intermediate

taxonomic levels (usually family or genus), but

also primarily in undivided continental or

global format. Unless specifically utilized in

subsets, global databases will generally provide

only broad patterns, where the influence of each

factor is disguised by its combination with

others. Complex patterns of local and regional

migration also confuse data on precise geo-

graphic origins; it has not yet been shown

whether high-diversity, stable late Ordovician

communities occupied the same environments

as the centres of diversification, or whether the

development of long-term global diversity is

independent of areas where rapid speciation

and overturn take place. Swain (1996), for

example, described exceptionally rapid ostra-

code turnover in a bentonite-rich succession,

where diversity at the top of the section did not

greatly exceed the initial, but standing diversity

at intermediate levels was much higher. Thus, in

order to recognize meaningful diversification

patterns reliably, overall diversity changes must

be examined on regional or even intra-basinal

scales (Miller 2000). Then, to identify the

significance of any given process, a precise

ecological-evolutionary signature must be

established, where possible, and then tested.

This is best examined by testing hypotheses in

limited regions, under the specific conditions

required for the testing. If a meaningful effect is

observed, locally or regionally, then large-scale

databases (e.g. Owen & McCormick 1999) can

allow assessment of the wider significance of

established local effects. Although many factors

may have acted simultaneously in promoting

the Ordovician Radiation, specific predictions

of major regional processes should be recogniz-

able on larger scales. Indeed, the ultimate

challenge in understanding the Ordovician

Radiation will be the extrapolation of known

small-scale patterns into a global situation,

where interactions between regions and pro-

cesses will require complex palaeogeography-

based analyses.

Among the many possible influences on

Ordovician diversity, the apparent correlation

with global pyroclastic volcanism has retained a

perennial but uncertain significance. Miller

(1997a) provided a correlation with global

tectonic/volcanic intensity, in which Palaeozoic

standing diversity appeared approximately to

match levels of orogenic activity. However,

correlation of this nature may represent a

relationship between standing diversity and

instantaneous tectonism, or between rapid

speciation and tectonism plus another factor,

which was restricted to the Ordovician.

Extreme fluctuations in tectonism during the

Permian, in a relatively stable part of the diver-

sity curve, and the lack of correlation in the

Mesozoic and Tertiary, lend some support to the

latter; the correlation is much more precise for

the Ordovician than for subsequent periods.

Similarly, Miller & Mao (1995) presented a geo-

graphical correlation of diversification centres

with forearc basins. Assuming some causal

relationship, it is unclear whether the volcanism

itself or associated tectonic activity was the

more significant influence on diversity. Miller

(1997a) suggested habitat partitioning as a

primary cause of diversification, in addition to

considering substrate transitions and nutrients,

while Swain (1996) preferred an increase in

nutrient supply to explain ostracode diversifica-

tion following major ash-falls. Increased nutri-

ent supply from run-off in uplifted areas is

perhaps inevitable, but the possible link to

diversification is unclear; in modern oceans,

abundant nutrients generally encourage low

diversity and very high dominance (e.g. Boyd et

al. 2000). In particular, a regional increase in

nutrient availability would tend to homogenize

subfacies variations for which scarcity of food

supply was critical for their definition. Para-

patric or sympatric speciation based largely on

behavioural food-acquisition procedures is.

intuitively, more likely to arise in situations

where food sources are relatively rare and

competition most intense.

This chapter introduces a mechanism by

which pyroclastic volcanism could have directly

influenced the speciation rate on regional scales,

with predictions for the evolutionary patterns to

be expected. The available global data sub-

divided by taxonomy (e.g. Sepkoski 1995)

appear to be consistent with these predictions

regarding the differences in diversity changes

between ecological groups. The hypothesis

potentially allows explanation of the differential

evolutionary rates between ecological/taxo-

nomic groups, particularly the decoupled

planktic-mobile benthic-sessile benthic diver-

sity curves. A preliminary test of the hypothesis

is included, using Welsh Basin ostracodes, and

similar methods are encouraged for those

investigating other hypotheses.