Two Galaxiid Populations Above and Below a Waterfall
Two populations of a non-migratory galaxiid fish live in neighbouring tributaries of a South Island river, separated by a 12 m waterfall that has been impassable for at least 6000 years. The lower tributary is warm (summer 16 °C, winter 7 °C) and silty; the upper tributary is spring-fed and stays between 8 and 10 °C year-round with stable cobble substrate.
Fish from the two streams differ consistently: upper fish are stouter with shorter fins and larger pectoral muscles; lower fish are slender with longer fins. Upper fish spawn in August at 8 °C; lower fish spawn in October at 12 °C. In laboratory crosses, eggs fertilise and hatch at normal rates, but hybrid juveniles survive to one year at only 30% of the rate of either pure cross, and hybrid males produce very few viable sperm.
A landslide 40 years ago briefly connected the two streams; a survey that year found no fish above the waterfall. A recent eDNA survey of the main river below the confluence detected DNA matching both forms within 2 km of each other, but no specimens have been collected there.
1. Evaluate whether the two populations should be classified as separate species, using the biological species concept and at least one other species concept.
Read the worked answer
Biological species concept (BSC): species are groups of actually or potentially interbreeding natural populations that are reproductively isolated from other such groups. These populations never meet, so the test is potential interbreeding, which the lab crosses supply. The F1 hybrids are viable and hatch normally, so prezygotic isolation is incomplete in the lab. But hybrid juveniles survive at only 30%, and hybrid males are effectively sterile - that is strong postzygotic isolation, and it means the hybrid class would not persist in a mixed population. On the BSC, that supports separate species.
Limitation of the BSC here: it is awkward for allopatric populations, because 'potential' is an inference, not an observation, and one set of crosses cannot exclude that the two forms would assortatively mate if they met. The BSC also says nothing about the ecological and morphological divergence, which is substantial.
Morphological and ecological species concepts: the two forms are diagnosably different in body shape and fin proportions, and they occupy distinct thermal niches with non-overlapping spawning times. Under a morphospecies or ecological species concept they are clearly separate units.
Judgement: the balance of evidence supports recognising them as distinct species, but the decisive test is contact. The timing difference (August versus October) is a prezygotic barrier that the lab crosses bypassed by artificial fertilisation, so real-world isolation may be stronger than the lab suggests.
2. Explain how reproductive isolation could have evolved in this situation, and identify the type of speciation, justifying your classification.
Read the worked answer
This is allopatric speciation by vicariance. The waterfall is a physical barrier that has prevented gene flow for at least 6000 years; the brief landslide connection 40 years ago did not create gene flow, because no fish were present above the falls at that time, so the populations were never reconnected reproductively.
With gene flow stopped, three things drive divergence. Divergent selection differs sharply between the habitats: 8–10 °C spring water versus a 7–16 °C silty stream. Thermal regime sets development rate and spawning cue, substrate sets body and fin shape, and flow regime sets pectoral muscle mass. Genetic drift is strong because headwater populations are small and isolated. Mutation supplies the raw variation.
Reproductive isolation then arises mainly as a by-product of that divergence, not because selection favoured isolation itself. The August/October spawning difference is a prezygotic temporal barrier. The 30% hybrid survival and poor hybrid sperm are postzygotic barriers, consistent with Dobzhansky–Muller incompatibilities: alleles that are fine in their own genetic background interact badly when brought together.
Six thousand years is short in geological terms, but with a generation time of two to three years that is 2000–3000 generations - ample for drift plus strong divergent selection in small populations to produce the observed barriers.
3. Explain what the eDNA result does and does not establish, and describe the evidence you would need before concluding the two forms are interbreeding where they now meet.
Read the worked answer
What eDNA does establish: DNA from both forms is present in the water within 2 km of each other below the confluence. That is a detection of genetic material, not of organisms. It is consistent with both forms occurring in the main river, but also with material washing downstream from each tributary, or with DNA bound to sediment and persisting after the fish have gone.
What it does not establish:
- That live fish of both forms occupy the same stretch of river at the same time.
- That they meet during the spawning season, when temporal isolation could still separate them.
- That any interbreeding occurs. eDNA cannot distinguish a hybrid from a mixture of two pure forms, and standard eDNA markers are usually short mitochondrial or single-locus fragments that carry no information about admixture.
Evidence needed: capture fish at the confluence and genotype them at multiple diagnostic loci; calculate each individual's hybrid index and look for F1s, F2s and backcrosses; check whether any admixed individuals survive to adulthood; and survey spawning timing and habitat use in the zone of contact. If admixed adults are absent or rare, the barrier is holding in nature and the case for separate species is strengthened, regardless of what the lab crosses showed.