Properties of Populations

Chapter 26: Properties of Populations

26.2 Does isolation increase the likelihood of extinction?

· Context: Most species vary in local densities and have spatially distinct populations.

· Major themes: Biological systems exceed the sum of their parts, and randomness within a biological system provides flexibility of response.

· Bottom line: The isolation of small populations from other populations of the same species can lead to local extinction.

Biology Learning Objectives · Describe and explain the various aspects of spatial structure of populations. · Be able to explain the concept of metapopulation. · Analyze the emergent properties of populations, and explain how spatial structure arises from characteristics and behaviors of individuals.

 

You are very familiar with the spatial distribution of the human population. At an early age, you learned your address and that you lived in a particular neighborhood in a particular city in a particular country. The population of humans in one country is usually quite distinct from populations in other countries, and they are spatially separated and often kept apart with borders and barriers. Yet you are probably also familiar with the movement of humans from country to country, either temporarily, as with vacationers or migrant workers, or permanently, with immigration and acquisition of new citizenship. Although the structure and rules around movement from country to country are codified for humans, this is not so for other species. Yet the separation of populations in space and the movement of individuals between populations are common features of most populations.

A population is a group of individuals of one species, living in one place at one time. {Connections: The population concept is introduced in Chapter 17.} Not all the individuals of one species live in one place. You learned in Section 19.3 how population structure affects gene flow, but you also learned about the basic idea of population structure. Recall the fungi distributed on logs in a forest, bladder campion in southwestern Virginia, and starlings across North America. {Connections: The relationship between gene flow and population structure is explored in Section 19.3.} Populations of individuals may be separated in space, which could affect their growth, but there are other demographic characteristics of populations that could affect the survival and growth of populations. You will learn in this section how a population is an emergent property that arises from the actions and interdependencies of the individuals in the population, and how the structure of a population may affect the long-term survival of a population.

 

A single population living in a single location
Black-cheek lizard (Calotes nigrilabris) in Sri Lanka     <div class=”player-unavailable”><h1 class=”message”>An error occurred.</h1><div class=”submessage”><a href=”http://www.youtube.com/watch?v=D1izmWBCrOw” target=”_blank”>Try watching this video on www.youtube.com</a>, or enable JavaScript if it is disabled in your browser.</div></div>
The black-lipped lizard at Horton Plains National Park, Sri Lanka

 

Consider first a species that exists only in one location. The black-lipped lizard, Calotes nigrilabris , is an endemic species found in Sri Lanka, an island off the southern tip of India about the size of West Virginia (Figure 26.5).

Walter Erdelen studied the distribution of lizards on Sri Lanka for several years. He sampled for lizards in numerous locations across the island. He also had data on locations where museum specimens had been found. The past and present data were used to estimate the extent of the geographic range for several lizard species prior to habitat fragmentation by humans. If a species was found in many locations within a particular habitat type, Erdelen assumed that it lived wherever this habitat type occurred. For C. nigrilabris, Erdelen accumulated 15 total records of this species and its location. This lizard was found in a forest in the Central Hills (see Figure 26.5B), inhabiting tree trunks, hedges, and shrubs, where it hunts for insects and worms. The other species were found in wider geographic distributions.

Figure 26.5 The black-lipped lizard, C. nigrilabris (A) and its distribution on Sri Lanka in hatched areas (B). Altitude and distribution of habitats on Sri Lanka (C); 1, monsoon scrub jungle; 2, semi-evergreen monsoon forest; 3, forest between dry and wet zones; 4, lowland wet zone rainforest; hatched areas, mountain forests (915 to 1525 meters in elevation); black areas, cloud forests (> 1525 m). A, http://commons.wikimedia.org/wiki/File:Horton_Plains_National_Park_126.JPG, Author: Cherubino. 9 May 2013. This file is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license. From http://www.slendemics.net. B and C, Figures 1 and 4d from Erdelen, 1984, Journal of Biogeography, © 1984 Wiley.

Erdelen went on to perform a mark-recapture study of C. nigrilabris to study population demographics of lizards in the cloud forest, which was in the middle of the range of C. nigrilabris. For the mark-recapture study, Erdelen carefully searched for lizards in the study area every 3 to 4 weeks. The biologist used a small noose to capture as many lizards as possible, which he then uniquely marked, and then released them at the place where they were captured (Table 26.3).

Erdelen used three methods to mark the lizards. The first marking was a small piece of tape with a unique identifying number affixed to the base of the tail, which was reliable for short periods of time and useful for quick identification. The second method was to place a unique pattern of color markings on both sides of the neck region using correction fluid. This was useful for identifying lizards using binoculars up to the time that a lizard shed its skin. Clipping scales of the dorsal crest on large lizards or clipping toes on small lizards was a third, permanent mark. Each individual had unique patterns of clipping. Lizards that were marked and could be identified positively were not recaptured. If the markings could not be seen clearly or if a lizard was unmarked, they were captured during a sampling period.

Table 26.3 Sampling and marking of C. nigrilabris lizards in the Sri Lanka cloud forest. From Erdelen, 1988, Table 2, Journal of Herpetology 22(1):42-52. Reprinted with permission from SSAR.

Table 26.4 Population demographics of C. nigrilabris lizards in the Sri Lanka cloud forest. Only 17 samplings could be used in the model to estimate density. All 19 samplings used to calculate percentages. ha, hectare; SD, standard deviation. From Erdelen, 1988, Table 3, Journal of Herpetology. Reprinted with permission from SSAR..

Figure 26.6 Population dynamics of the black-lipped lizard at one site within the Sri Lankan cloud forest. The purple area represents the density of males, orange represents the density of females, and teal represents the density of juveniles, all extrapolated from the proportions of these classes from the sampling (see Table 26.4). From Erdelen, 1988, Figure 3a, Journal of Herpetology 22(1):42-52. Reprinted with permission SSAR.

Erdelen calculated the percentage of females, males, and juveniles in each sampling and used a mark-recapture model to estimate the population density of the lizards in the cloud forest (Table 26.4; Figure 26.6).

This model took into account birth, death, and migration out of the study area—even though the entire species was in one large population, individual lizards move around within the forest. The model also took into account differences in mortality rates between juveniles and adults. Bio-Math Exploration 26.1 explains the model that Erdelen used to estimate the population densities given in Table 26.4.

From the distribution map you determined that the black-lipped lizard is found in mountain and cloud forests at elevations of about 1,000 meters and above. Erdelen assumed that this lizard would be found in all such forests in the Central Hills. There is one small area of cloud forest outside of that area where the lizard is not found. The two areas are separated by 10 kilometers or more of unsuitable lowland habitat, so moving to the new forest would be a difficult journey for these lizards. A lizard in a suitable habitat would be unlikely to travel through the unsuitable habitat unless it was highly motivated to leave where it was, perhaps by high density of other lizards or low food availability. {Connections: Density-dependent factors are discussed in Section 29.2.}

The density estimated by the researcher was higher than that determined for other lizard species on Sri Lanka, although apparently not high enough to cause migration from the cloud forest, because it was never found at other study sites below the mountain forest. Erdelen concluded that population density and other demographic traits were stable in the black-lipped lizard. He also concluded that lizards tended to remain in one area for long periods of time, although they were capable of moving about within the forest. Within just 1 hour of searching the surrounding forest outside of the study site, Erdelen found eight individuals that were marked between 7 and 17 months prior. At the same time, new unmarked individuals were found in censuses, and it was unlikely that this frequency of lizards had been overlooked in previous samplings. Juveniles were recaptured less frequently than adults; they could have been migrating as they matured or they were suffering higher mortality. It is difficult to determine based on the available data. Females were less common than males, and females may have a greater tendency to disperse. Although the reason is not completely known, you may have also speculated that they are more difficult to find if they live higher up in the trees or are more secretive while preparing to lay eggs. Females may be more susceptible to predation or parasitism because of greater energy requirements to provision or lay eggs.

You have learned that this species of lizard is endemic to one forest. The long-term data indicate that it has a relatively stable population, and it moves about the mountain and cloud forests, probably in search of food and mates. The entire forest habitat contains just one population of this lizard, making it the only population of this species. The population is a dynamic emergent property of the individuals within it. As long as the habitat is protected and is not altered by climate change, the black-lipped lizard population should survive. Given that there is only one population of this species, it could be in danger of extinction. If the forest gets cut down or encroached upon by humans, the black-lipped lizard could lose its very specific habitat. If the forest was fragmented by logging, the population could be split into two or more smaller populations, and this could affect the ability of the isolated populations to survive.

 

Populations living in temporary habitats

 

Consider a species that has many small populations, called subpopulations, all living in different areas and yet connected to each other by gene flow. {Connection: Gene flow is discussed in Section 19.3.} We’ll call that set of connected subpopulations a metapopulation. Each subpopulation may exist at different densities in different places. Variation in the habitat combined with limited dispersal abilities can result in this spatial structure. For instance, the black-lipped lizard does not live in both cloud forests in Sri Lanka because of the lack of suitable lizard habitat between the forests making movement from one forest to the other difficult. But in many populations, there are distinct subpopulations within a larger metapopulation. How does the isolation of subpopulations affect population size fluctuations and the likelihood of local extinction?

You can imagine that plant metapopulations would be very amenable to study demographics, because plants are immobile; and subpopulations, once their locations are known, can be tracked through time. Spencer Barrett and his graduate student, Brian Husband, did just that for the aquatic plant Brazilian water hyacinth (Eichhornia paniculata) , an annual plant that goes through its life cycle each year (Figure 26.7, inset). This plant occurs in temporary pools and roadside ditches in the arid scrub habitat of northeastern Brazil. Temporary pools are aquatic habitats that dry for at least part of the year and fill with rain water or snow melt.

Figure 26.7 Distribution in Brazil of sampled Brazilian water hyacinth populations. Abbreviations indicate different Brazilian states. Brazilian water hyacinth (inset). From Barrett and Husband, 1997, Figure 1, J Heredity 88:277-284, Copyright © 1997, by permission of Oxford University Press.. Inset, From http://www.jardineiro.net/images/banco/ eichhornia_paniculata.jpg.

This plant also occurs in Cuba, Jamaica, and Nicaragua. In the range studied by Barrett and Husband, the Brazilian water hyacinth has subpopulations concentrated in regions where there is a high abundance of pools and ditches. The aquatic habitats occupied by this plant are ephemeral; that is, they do not persist year-round but only after they fill up during the rainy season. During the dry season, when habitats are dried out, Brazilian water hyacinth seeds remain dormant, germinating when habitats refill. Germinated seedlings must then grow, mature, and reproduce again before the pool or ditch dries up.

Barrett and Husband surveyed a large number of pools, ditches, and flooded pastures in 1982, 1987, 1988, and 1989 within 100 meters of roads (Figure 26.7). The plant is known to be associated with disturbed areas, which led to the researchers surveying near roads. Barrett and Husband sampled transections along stretches of road, from 29 to 300 km in length and recorded the location of every one of the 167 subpopulations for which abundance was determined. Most sites with subpopulations found in the first three surveys were examined during subsequent surveys for an existing subpopulation. The researchers were interested in studying the variability in the demography of subpopulations (Figure 26.8), occupation of suitable habitats, and persistence of subpopulations (Table 26.5).

Figure 26.8 Changes in representative Brazilian water hyacinth subpopulation sizes in northeastern Brazil. From Barrett and Husband, 1997, Figure 2, J Heredity 88:277-284, Copyright © 1997, by permission of Oxford University Press.

Table 26.5 Persistence of Brazilian water hyacinth subpopulations in northeastern Brazil. The number of sites differed among years as the researchers sampled different geographic areas. From Husband and Barrett, 1998, Table 1.

The researchers estimated the size of each population during each sampling. For each subpopulation, they counted the number of plants if the subpopulation contained less than 250 individuals. If the subpopulation was larger than that, they sampled several areas, determined the average density, and multiplied that by an estimate of the habitat area. The researchers examined the changes in size for each subpopulation over time (Figure 26.8).

Figure 26.9 Size distributions and density of Brazilian water hyacinth subpopulations surveyed in northeastern Brazil. For size distributions (A and B), subpopulations were categorized as either present (teal bars) or absent (purple bars) 1 year later. Densities in C and D were determined as number of subpopulations or patches per kilometer of road surveyed. From Husband and Barrett, 1998, Figures 4 and 6.

Barrett and Husband also determined the annual probability of persistence, measured as the proportion of subpopulations present at one time that were still present the next year for surveys conducted 1 year apart (Table 26.5). For the 5 year span between 1982 and 1987, the probability of persistence was estimated and multiplied by five to get an annual rate, assuming that subpopulations did not become absent and then reappear between 1983 and 1986. Metapopulation demographics also consisted of determining the proportion of habitat patches occupied by the plant out of the total number of patches surveyed in 1988 and 1989 (Table 26.5). If the plant was absent from a habitat, the researchers carefully checked to make sure it was suitable habitat by searching for other species of plants known to live in the same habitat.

Barrett and Husband examined the relationship between population size and whether it was present the following year or had gone locally extinct for 1987 and 1988 (Figure 26.9). They determined the density of subpopulations and the density of available patches for each transect of road surveyed, occupied or not (Figure 26.9).

 

About half of the water hyacinth subpopulations contained less than 100 individuals, and subpopulations did not all decrease or increase in any particular year. Barrett and Husband also found that subpopulations present in 1 year had a 64% probability of reappearing the next year but did not persist indefinitely. Subpopulations of all size classes had a relatively equal probability of persistence. This lack of pattern over time among subpopulations suggests that environmental factors within a habitat patch may have a large impact on subpopulation size. Isolation in combination with low migration and rapid environmental changes makes Brazilian water hyacinth subpopulations vulnerable to dramatic size fluctuations and possible local extinction.

Although many subpopulations did go locally extinct during the study, the metapopulation as a whole persisted regionally with a small fraction of habitat patches occupied by water hyacinth at any time. Metapopulations consist of subpopulations connected via immigration and emigration. A spatial pattern observed by Barrett and Husband was that there was a positive relationship between the density of subpopulations and the total patch density. Patches can be recolonized through migration, and recolonization rates should increase as patch density rises. Increasing patch density results in a higher proportion of patch occupancy, assuming that local extinction rates are constant. A low density of patches causes more subpopulation isolation and that causes a lower density of patch occupancy. No subpopulations were found in areas with patch densities below 0.2 patches per kilometer, suggesting a threshold below which recolonization rate of unoccupied patches is lower than the extinction rate of subpopulations.

Most organisms are found within suitable habitat that is distributed patchily across a landscape. You discovered that the Brazilian water hyacinth is found in finite local subpopulations that are vulnerable to size fluctuations and local extinction events but that subpopulation size did not affect the probability of local extinction. The persistence of this metapopulation occurs in disjunctive ephemeral habitats with subpopulations existing for short periods of time and then going locally extinct. Many other species exist as metapopulations but in habitats that are more permanent and stable than temporary pools. How does population structure affect the likelihood of extinction in such species?

 

Colonization and extinction in metapopulations

 

Mauricio Lima and his colleagues studied five species of small mammal in Chile for 6 years. The scientists were interested in the factors that affected colonization and extinction rates in metapopulations. The metapopulations were monitored in a mountain range in Chile. In this region, slopes that face north receive more sun and have sandy soils and scattered rocky outcrops. Grasses are scarce, but cacti are abundant. Slopes that face south receive less sun, soils appear to contain more organic matter, grasses are more abundant, and cacti are nonexistent.

The biologists monitored metapopulations of five small mammal species using mark-recapture procedures on two opposite-facing slopes in two creeks. Lima and colleagues trapped mammals in each creek every other month for 5 consecutive nights. During trapping sessions, traps were checked every morning, and all individuals captured were marked with a metal ear tag. The species, body mass, and sex were determined for each individual. The five species were the leaf-eared mouse (Phyllotis darwini) , the fat-tailed opossum (Thylamys elegans) , the degu (Octodon degus) , the South American field mouse (Akodon olivaceus) , and the pygmy rice rat (Oligoryzomys longicaudatus) .

Figure 26.10 Relationship between extinction (A) and colonization rates (B) and population density for five small mammal species in Chile. Rates were determined for each species in each subpopulation, and regressions were calculated for north and south slopes separately. Closed dots represent values for north exposures, and open dots represent values for south exposures. Figure 1 from Lima et al., 1996, with kind permission from the Springer Science and Business Media.

To estimate extinction and colonization rates, the researchers considered each slope a separate subpopulation and assumed that they were all connected in a larger metapopulation. Lima and colleagues determined extinction and colonization rates for each species in each subpopulation (Table 26.6). They defined extinction rate as the probability that a subpopulation went extinct on a slope between successive census periods. Similarly, the probability that a subpopulation became present after being absent in a preceding census defined colonization rate. The relationship of extinction and colonization rates with average population density was determined by linear regression (Figure 26.10). {Connections: Linear regression is illustrated in Bio-Math Exploration 16.1.} The researchers calculated average population density only for trapping periods where a species was found and then took the logarithm of the mean.

Table 26.6 Extinction (E) and colonization (C) rates of small mammals in Chile, by species, creeks, and solar exposures. Table 1 from Lima et al., 1996, Oecologia 107(2):197-203. With kind permission from the Springer Science and Business Media.

 

The leaf-eared mouse had the lowest mean extinction and colonization rates. It also was found in the highest densities across all subpopulations. Colonization rates were low for the degu and the pygmy rice rat. The highest extinction rates were observed for the opossum, the degu, and the pygmy rice rat. The latter two species thus had low colonization rates and high extinction rates. However, you observed in Figure 26.10 that the degu had colonization and extinction rates that were habitat specific. On north-facing slopes, this mammal had lower extinction and higher colonization than for subpopulations of the same species on south-facing slopes. The other four mammal species had rates that clustered together, indicating less of an effect of local conditions and suggesting that patches facing in different directions were equally suitable. However, as you calculated, and as Lima and his colleagues concluded, extinction rates for all species were lower on north-facing slopes.

You determined that there was a negative linear relationship between extinction rate and population density. The colonization rate results were mixed. Lima and his colleagues determined that there was a negative relationship between colonization rate and the logarithm of population density for north-facing slope subpopulations. Lima and his colleagues concluded that this was partly an artifact of very low colonization rates of the leaf-eared mouse, which had low colonization rates because it rarely went locally extinct. It may actually have higher colonization rates, which prevented extinctions from occurring, but the researchers could only determine colonization events according to their census schedule; they could have occurred more frequently. Although there was a slightly positive slope for the relationship between colonization rate and population density for south-facing slopes, the researchers determined that the slope was not significantly greater than a slope of 0, suggesting no statistical relationship. {Connections: Interpreting slopes of regressions is discussed in Bio-Math Exploration 16.1.} Lima and his colleagues concluded that population size was an important factor determining extinction rates and most species had higher extinction than colonization rates. Local extinction occurs often and persistence of subpopulations over long periods of time is driven by repeated colonization.

Extinction rates in these mammalian metapopulations differ in habitat patches of differing quality. Higher extinction rates in south-facing slope habitats leads to smaller populations. Smaller populations are more vulnerable to catastrophic or random disturbances that could cause local extinction. Although population size was not related to subpopulation extinction for the Brazilian water hyacinth, the trend for these five species is that extinction rate is inversely related to population size or density. Empirical observations of other species reveal that extinction rates are negatively correlated with subpopulation size. Isolation of small subpopulations (such that colonization is largely prevented) would increase the likelihood of local extinction.

This section began with the observation that humans live in distinct populations, separated by real or imaginary boundaries. You then constructed knowledge about the demographics of populations of other species within both populations and metapopulations. The spatial structure of a population is an emergent property of individuals and affects the growth and survival of the population. The metapopulation is a property that emerges from the set of subpopulations connected through migration. Think of the metapopulation as a string of blinking Christmas tree lights. Each light represents a habitat patch; and when it is on, the patch is occupied by a subpopulation. A light blinking off represents a local extinction, and a light blinking on represents establishment of a new population. Of course, subpopulations of a metapopulation do not have a blinking pattern like Christmas tree lights do, but the constant change of occupation of habitats can be visualized using this metaphor. In the next section, you will examine a different type of population structure, the age structure, and learn how that affects growth of a population.