4.1, 4.2 - Biodiversity, endemism and diversity indices

Variety of life and human threats

The present variety of organisms is the outcome of evolution over many generations, including the formation of new species and the extinction of others.

A woodland illustrates these different scales. Its plants, animals, fungi and microorganisms belong to many species. Individuals within one of those species can carry different alleles. Woodland, grassland and freshwater habitats also support different communities. Counting tree species alone therefore measures one aspect of biodiversity, rather than all of it. This broad meaning follows the Convention on Biological Diversity.

Human activities can reduce this variety through several connected pressures.

These are established categories of human pressure identified by IPBES. They can act together: a population with much of its habitat removed may also have fewer places to move as climate changes. A population decline is not automatically a global extinction. Local extinction means disappearance from a particular area; global extinction means no living individuals of that species remain anywhere.

Losing a species removes a distinct part of biodiversity. Losing individuals within a species can also remove alleles if those alleles occur only in the lost individuals. A species count may initially stay unchanged even while genetic variation is being lost. This is why measuring diversity requires more than counting all organisms together.

Endemism and geographical scale

A species' distribution describes where it occurs, whereas its abundance describes how many individuals there are.

Always name the area. A species could be endemic to an island, a mountain range or a whole country. Being native to a place does not by itself make a species endemic: a native species may also occur naturally elsewhere. The Convention on Biological Diversity's account of islands uses this geographical meaning.

Consider two invented species. Species P has 500,000 individuals, all naturally confined to Island L. Species Q has only 30 individuals in a survey on Island L but also has natural populations across several continents. P is endemic to Island L; Q is not. P's large population does not prevent endemism, and Q's low local abundance does not establish it.

Rare describes low abundance or limited occurrence in a stated context. Endangered concerns extinction risk; a formal assessment uses evidence such as population decline and geographical range, rather than the word endemic alone. The IUCN Red List assesses extinction risk. Neither an endemic label nor one small sample is enough to assign an actual conservation category.

A restricted natural range can nevertheless make a species vulnerable to a pressure affecting that whole area. If Island L contains all of P's natural population, destruction throughout the island could affect the entire species. The same local event would affect only part of Q's wider distribution. The outcome still depends on the severity of the pressure and the species' biology; endemism does not mean inevitable extinction.

Species richness and evenness

Counting individuals and counting species answer different biological questions.

Suppose the same plant-counting method and sampled area produce these original practice observations. Species A–D are four distinguishable, hypothetical non-clonal plant species. Each counted plant is one individual; the data do not represent all life in either habitat.

Both samples have species richness 4, although the distributions of individuals differ. Evenness describes how evenly individuals are distributed among species. U has equal counts; V is dominated by A, which makes up $28/40=0.70=70\%$ of its counted individuals. Richness alone cannot describe this contrast. An index incorporating each species' abundance can.

Richness also says nothing directly about allele differences within one species. Forty individuals of one species contribute richness 1, whether they have many different genotypes or the same genotype. Genetic diversity is variation in genetic information, including alleles, within a species or population. We need genotype data to investigate it.

A recorded richness is usually an estimate of the habitat's richness. A larger sampled area or longer search can reveal more species simply because there are more opportunities to encounter them. An apparent difference between habitats may therefore reflect unequal sampling effort. Use comparable effort and methods before interpreting the difference as biological.

Heterozygosity from genotype counts

Heterozygosity measures the proportion of individuals carrying two different alleles at a specified gene locus.

Recall that a diploid heterozygote has two different alleles at that locus, such as Aa; AA and aa are homozygous genotypes. Counting Aa individuals is not the same as counting copies of allele A or allele a.

The original practice table gives genotypes at the same autosomal locus in every individual of two small, defined populations of one hypothetical diploid species.

For J, $H=24/(10+24+6)=24/40=0.60$: 60% of individuals are heterozygous at this locus. For K, $H=8/(20+8+12)=8/40=0.20$: 20% are heterozygous. J has the higher observed heterozygosity at this locus. Both populations still contain the two alleles A and a. The result does not mean J has three times as many alleles or three times the species richness.

There are 80 allele copies at this autosomal locus in J, because each of its 40 diploid individuals contributes two. Its A-allele count is $2(10)+24=44$, giving A frequency $44/80=0.55$. That differs from $H=0.60$, because allele frequency counts allele copies, whereas this heterozygosity index counts heterozygous individuals. Do not double the denominator when calculating $H$.

Rearranging the relationship

From $H=h/M$, multiply both sides by $M$ to obtain $h=HM$. If 40 individuals have $H=0.45$, then $h=0.45\times40=18$. To find an unknown total, rearrange again to $M=h/H$, provided $H>0$. Thus 18 heterozygotes and $H=0.45$ imply $M=18/0.45=40$. If $H=0$, this division cannot recover the total; zero heterozygotes could occur in populations of many different sizes. With no individuals recorded, $H$ itself cannot be calculated.

If only a sample from a larger population is genotyped, divide by the number of sampled individuals with usable genotypes at that locus and describe the result as observed sample heterozygosity. It estimates population heterozygosity only to the extent that the sample represents the population. A convenient sample of close relatives may be misleading. Compare the same loci using comparable sampling and genotyping methods; several representative loci give stronger evidence about wider genetic diversity than one locus. No Hardy–Weinberg assumption or predicted genotype frequency is needed to calculate this observed ratio.