communities and ecosystems - university of california, irvinedarwin.bio.uci.edu/~mueller/pdf/rose...

15

Communities and Ecosystems

When scientists first began studying bio-logical communities, they were so fasci-nated with the interactions and

dependencies between species that they saw the bi-ological community as a superorganism. Wholespecies were viewed as organs that performed spe-cific functions for the complete ecological superor-ganism. The integration and communicationbetween these “organs” was thought to be deliber-ate and well tuned. One way to think of this idea isto imagine a stitched-together Frankenstein, eachsewn-on body part a distinct species.

Today biologists find the analogy between bio-logical communities and organisms superficial. Tobe sure, there are populations within communitiesthat are highly dependent on each other. And it isalso true that biological communities and theirphysical environments support all life on Earth bysuch processes as recycling nutrients. The impactof this recycling can be profound. For instance, at-mospheric levels of carbon dioxide depend onplant photosynthesis and the respiration of all aer-obic organisms. Global temperatures and weatherare in turn dependent on atmospheric carbon

dioxide levels, which are covered in Chapter 16(The Biosphere and the Physical Environment).The coordination and integration of biologicalcommunities has vast implications for the Earth.

For this reason, there are few biological topics asimportant for the future of life on Earth as the func-tioning of ecosystems. In this chapter, we surveyhow ecosystems function, from the flow of energy inModule 15.1 (Energy Flow) and the recycling of nu-trients in Module 15.15 (Ecosystems) to the porten-tous problem of the fragility of ecosystems. InModules 15.8 (Community Organization) and 15.4(Equilibrium and Nonequilibrium Communities),we consider the factors that determine the numberof species in a community. Surprisingly, in somecommunities predation and environmental distur-bance may promote increased species diversity. Is-lands represent interesting communities, becausevirtually all species on an island must travel therefrom some larger mainland. Species diversity on is-lands is a consequence of dynamic processes. Under-standing these forces has important practicalapplications for the design of ecological preserves, tobe covered in Chapter 17 (Conservation). ❖

The feeding relationships between species canoften be complicated.

437

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 437

Laurence

Typewritten Text

Copyright © Michael R. Rose and Laurence D. Mueller, used by permission, for academic purposes only

438 Chapter 15 Communities and Ecosystems

ENERGY FLOW

15.1 The flow of energy is a central organizing theme in community ecology

We have all strolled through forests or walked along theseashore or lakeside. Even the untrained person will notice avariety of plants in a forest or the many insects and birds nearlakes and oceans. These interacting plant and animal popula-tions are part of a biological community. The members ofsuch a community will be apparent from their associations ortheir geographic location. As we have seen in the previouschapter, some plants and animals may interact very closelyand affect each other’s evolution. While the details of process-es such as coevolution were unknown to early ecologists,there was a strong sense that there was a mutual interdepend-ence among the members of a community.

Communities Early in the twentieth century, F. E.Clements developed some of the first ideas about communi-ties. If a tract of land is cleared but then left undisturbed, itwill be recolonized by plants over time. This recolonization,or succession, may follow a predictable pattern, with somespecies appearing early in the sequence of recolonization, butlater giving way to different species. (Figure 15.1A shows oneexample of succession.) In studying ecological succession,Clements thought that the species that appeared during suc-cession made up a superorganism, with strong interdepen-dencies much like the organs of a single plant or animal. Wewill cover succession in more detail in Module 15.7.

In the 1920s, Charles Elton developed a more sophisticat-ed view of communities, one that still persists today. He stud-ied a tundra community on Bear Island in the North Atlantic.Elton’s focus was on feeding. Which species feeds on which isone of the most important interactions in an ecosystem.Figure 15.1B shows some of the results from Elton’s study.These feeding relationships also reveal a directional flow ofenergy. Moss captures energy from the sun. Energy in themosses is then consumed by herbivorous rotifers that are ul-timately eaten by ducks. Diagrams that show energy flows arecalled food chains.

The nature of an ecological community is not solely afunction of the organisms that make up the community. Thephysical environment also influences the numbers and typesof organisms in a community. Likewise, photosynthesis, res-piration, and decomposition affect the physical environment.In 1935 the English plant ecologist A. G. Tansley coined theterm ecosystem to describe ecological communities and theirassociated physical environment. In Module 15.15, we willdiscuss the interactions between biological communities andthe physical environment.

Trophic Levels In 1925 A. J. Lotka published his bookThe Elements of Physical Biology. Influenced by his training inchemistry, Lotka advocated the study of communities from a

thermodynamic perspective, emphasizing the transfer of en-ergy. This thermodynamic perspective and the importance offood chains were both embraced by Raymond Lindeman in1942. For his Ph.D. thesis, Lindeman studied the feeding rela-tionships in a bog community in Minnesota. He simplifiedthe analysis of energy flow in this community by focusing onorganisms that were at a similar position in the food chain.Such positions are referred to as trophic levels.

In most ecosystems, the lowermost or first trophic level ismade up of the primary producers or plants, such as themosses in Figure 15.1B. These organisms depend on sunlightfor their energy. The next trophic level up consists of the her-bivores, organisms that consume plants, such as the herbivo-rous rotifers in Figure 15.1B. The biomass or energy availableto herbivores comes directly from the primary producers. It isalso affected by the efficiency of conversion of energy. Con-sumers of herbivores—such as the ducks in Figure 15.1B—are at the next trophic level, and so on.

Lindeman noted that the dependence of each trophic levelon the one below it suggests that the amount of energy con-tained in each level (for example, as plant or animal biomass)should decline as one moves from the lower to the highertrophic levels. Lindeman called this natural progression the

Eltonian pyramid. InModules 15.2 and 15.3,we will study in moredetail what is knowntoday about the energyrelationships withincommunities and thefactors that affect ener-gy transfer from onetrophic level to thenext. ❖

FIGURE 15.1A The InitialStages of Succession in aTemperate Forest .

Long-Tailed Duck

Rotifers

Moss

FIGURE 15.1B Some Feeding Relationships from Elton’s study ofBear Island

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 438

Energy Flow 439

During his short life and even shorter academic career, Lindemanmanaged to write six scientific papers. One of these appeared in thejournal Ecology after his death, with the title, “The Trophic-Dynam-ic Aspect of Ecology.” This paper is credited with influencing manyecologists to look at the energy and feeding relationships amongorganisms as an important aspect of community structure. Linde-man received his Ph.D. from the University of Minnesota in 1941.Shortly afterward he moved to Yale University, where he beganpostdoctoral work with G. Evelyn Huchinson. The original versionof Lindeman’s trophic-dynamic paper was rejected by the journalEcology. It was only after an appeal by Hutchinson that the editor ofEcology, Thomas Park, agreed to publish the paper.

Raymond Lindeman (1915–1942)

FIGURE 15.1C Raymond Lindeman

RoseCh15_0104043_437-474_2p 11/18/04 3:07 PM Page 439


3

2

1

0

Log chlorophyll a (μg/L)

Log

zo

op

lan

kto

n b

lom

ass

(μg

/L)

�1 0 1 2 3

FIGURE 15.2A Biomass of Herbivores (zooplankton) versusBiomass of Plants (chlorophyll a) from Several DifferentFreshwater Lakes

Plant biomass

Zoo

pla

nkt

on

bio

mas

s

Low fishdensity

High fishdensity

FIGURE 15.2B Changes in Zooplankton and Plant BiomassFollowing Manipulation of Fish Numbers The arrows show howzooplankton and plant biomass changes in going from low tohigh fish numbers.

15.2 In most biological communities, all energy comes from the sunBiological systems are complicated, but they must follow thesame laws of thermodynamics that physical systems obey.The first law of thermodynamics tells us that energy can beneither created nor destroyed. Energy can be changed, how-ever, from one form into another. In biological communities,almost all energy originates from the sun. Green plants cap-ture solar energy and turn it into chemical energy. Because ofthis special function, green plants are called primary produc-ers. The chemical energy is stored by plants as bonds holdingorganic molecules together. Not all the captured energy fromthe sun is stored as chemical energy. Plants use some energyfor metabolic maintenance, and some is lost as heat.

All trophic levels above plants gain energy by feeding onmembers of other trophic levels. Herbivores feed on plants andderive their energy from the energy stored in plant tissue. Theflow of energy goes in one direction, from plants to herbivores—not the other way. Consequently, the energy content of all theherbivores in a community cannot exceed the energy containedin the primary producers. In fact, it will often be much less, for atleast two reasons. (1) The herbivores cannot consume all theplants, or there would be no future source of energy for the herbi-vores. (2) The conversion of chemical energy in the plants tochemical energy in herbivores is not perfect. Energy is lost as heator is unused due to incomplete digestion. Thus the total amountof energy in an ecosystem is determined by the primary produc-ers, though much will be lost to the biological community.

A rough indication of the amount of energy in the pri-mary producer level is their biomass. Figure 15.2C shows theEarth’s biomes (major communities classified according totheir predominant vegetation) and gives an indication oftheir typical biomass. As we can see, the amount of energy lo-cated at the level of the primary producers varies substantial-ly from one type of community to the next.

What causes this variation? The variation seen in Figure 15.2A is a function of environmental factors. Tun-dra has low primary productivity due to the short growingseason near the Earth’s North Pole. On the other hand, se-vere water shortage keeps the productivity of deserts low.The open ocean has plenty of light near the surface andmostly benign temperatures, but nutrients such as phos-phorus are in short supply, limiting plant growth. Plantsand animals on the ocean surface die and settle to the bot-tom of the ocean, where their decomposition does not im-mediately return nutrients to the surface. In some parts ofthe ocean, currents carry water from great depths up to thesurface. These upwelling currents (see Module 11.4) areimportant for supplying the surface waters of the oceanswith nutrients.

Given the dependence of higher trophic levels on lowertrophic levels, we expect that the biomass of herbivoreswould be positively correlated with the plant biomass. Forfreshwater lakes, this predicted relationship is generallyobeyed (Figure 15.2A). When the biomass or the number ofspecies in a community is controlled by the amount of pri-mary production, the community is bottom-up regulated.Conversely, if species biomass at most trophic levels is con-trolled by predation, the community is regulated top-down.There is nothing that prevents a single community from ex-periencing both bottom-up and top-down effects.

In lakes, the effects of top-down regulation by predationcan be studied by artificially increasing the numbers of fishthat feed on lake zooplankton. In Figure 15.2B, we see thatincreases in fish numbers lead to a decrease in zooplanktonbiomass and an increase in plant biomass. Because manyzooplankton species feed on plants, reductions in theirnumbers benefit plants. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM 40

Energy Flow 441

Tropical Rainforest Swamp and Marsh

Boreal Forest Temperate Grassland

Tundra and Alpine Desert

Open Ocean Lakes and Streams

FIGURE 15.2C Primary Productivity in Major Biomes (grams per square meter per year)

RoseCh15_0104043_437-474_2p 11/19/04 8:54 AM Page 441

15.3 The efficiency of energy transfer from one trophic level to the nextvaries among communities

What happens to the energy that is captured by primary pro-ducers? Some of the energy that is present in the plants istransferred to the next higher trophic level by herbivory andthus the conversion of plant biomass into herbivore biomass.This process continues up the energetic pyramid, as herbi-vores are eaten by carnivores, and so on.

Insights into community processes can be gained by ana-lyzing such energy flows from one trophic level to the next.No chemical or physical process of energy conversion can be100 percent efficient. Energy is lost in a variety of ways. InFigure 15.3A, we show how energy is lost as it flows fromlower trophic levels to higher trophic levels.

The green arrows in Figure 15.3A represent the energy thatsuccessfully makes it from one stage to the next. The red ar-rows represent the energy that is lost in these transfers. Ateach of these steps, we can compute the efficiency of energytransfer if we know how much energy from one stage makes itinto the next stage. In our example, a fox feeds on birds. Notall birds will be captured and eaten by the foxes, so not all theenergy present in the bird trophic level can be converted tofox biomass, for this reason alone. The efficiency of this partof the energy transfer is called the exploitation efficiency.

Once a bird is eaten, the fox must convert the energy of itsprey to energy it can use. Plants and animals consist in part ofmaterials, such as cellulose and bone, that contain energy but

Predators and herbivores will not find and consume all prey.

Undigestible parts of plants and animals are not assimilated.

Some assimilated energy is used for work, and respiration, some individuals die.

Prey = P

Ingestedprey = I

Assimilatedenergy = A

Growth &reproduction = G

Eaten

Remain in community

ExcretedAssimilated

Convertedto morepredators

Respiration

Exploitation = I/P

Assimilated = A/I

Net production = G/A

Ecological efficiency = G/P

FIGURE 15.3A This figure shows the flow of energy from one trophic level (prey) to asecond (predator). The amount of energy is represented by symbols in the leftmost column.The fraction of energy that passes through each step in this process is called efficiency andis shown in the middle column.

cannot be digested and assimilated by most of the consumersthat eat them. Thus, only a portion of the total energy devouredis chemically assimilated. The fraction of consumed energy thatis assimilated is referred to as the assimilation efficiency.

Some of this assimilated energy will be used for work andmaintenance. The rest will be used for growth and reproduc-tion, adding to the energy level of the foxes. The fraction ofthe assimilated energy that is made into new biomass by foxesdetermines the net production efficiency.

The efficiency of the entire process of energy transfer be-tween trophic levels is called the ecological efficiency. Eco-logical efficiency is the energy content of the higher trophiclevel divided by the energy content of the lower trophic level,as Figure 15.3A shows.

Energetic efficiencies vary across communities, dependingon the lifestyles of the organisms that make up these commu-nities. Figure 15.3B shows the net production and assimila-tion efficiencies of three categories of organisms. There is nogeneral pattern of assimilation efficiency, with ectothermsand endotherms showing a mixture of high and low values.But the net production efficiency of ectotherms is consistent-ly higher than that of the endotherms. This makes sense, be-cause endotherms must spend a larger fraction of theirenergy budget maintaining their body temperature and thushave less energy to devote to growth. ❖


RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 442

Energy Flow 443

Assimilation efficiency (%)

Net

pro

du

ctio

n e

ffic

ien

cy (%

)

Aquatic ectothermsTerrestrial extothermsTerrestrial endotherms

100

50

00 50 100

FIGURE 15.3B Net Productivity Efficiency versus AssimilationEfficiency of Several Groups of Animals

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 443


FIGURE 15.4B Area Surrounding Mount St. Helens ImmediatelyAfter the Volcano Erupted All signs of life are gone, and theground is covered with ash.

FIGURE 15.4A Mount St. Helens, Washington, during Its MostRecent Eruption

EQUILIBRIUM AND NONEQUILIBRIUM COMMUNITIES

15.4 Community stability can be disrupted by sudden changes in thephysical environment

What Is an Equilibrium? Populations in ecologicalequilibrium maintain relatively constant numbers of indi-viduals, and if displaced from these numbers, will return backto their equilibrium levels. Density-dependent populationgrowth is a mechanism that can help maintain equilibrium,as we saw in Chapter 10. Communities of competing speciesmay also achieve equilibrium.

One concept of a community in equilibrium is much likethe concept of an evolutionary equilibrium of species num-bers, maintained by extinction and speciation, introduced inChapter 6. The metaphor used there was of water dripping intoa plugged sink from a faucet, counterbalanced by water drip-ping over the side of the sink. At an equilibrium populationlevel, some factors tend to increase population size, while oth-ers tend to decrease it. If an equilibrium is stable, any change inthe numbers of one or more species will be followed by a re-turn to population sizes from before the perturbation. Whenthe community ecology maintains equilibrium, the number ofspecies in a community depends on the strength of competi-tion between close competitors and the food-chain structure.Populations of all species reach a stable equilibrium size in astable community.

A different view of community structure is that it does notproduce equilibrium. Nonequilibrium theory proposes thatnatural disturbances prevent populations from reaching anequilibrium. Unfortunately, it is hard to choose betweenthese alternative theories. Just to determine if natural popula-tions are at a stable equilibrium or not is time consuming.One must get accurate census data from populations overmany generations. Even then, a community may maintainroughly constant population densities only because of a lackof perturbations. This would be like the “stability” of a tallboulder that would roll or fall over if it were pushed, but hasnot yet been pushed.

Population Disturbance Populations can be prevent-ed from reaching an equilibrium due to environmental dis-turbances. These disturbances can come in many forms:storms, fires, drought, floods, even volcanoes. The distur-bance may cause relatively small changes in the populationsize or resource levels or, as in the case of the Mount St. He-lens volcano (Figures 15.4A through 15.4C), it may wipe outentire communities over large areas. Large disturbances, likethe Mount St. Helens volcano, effectively sterilize an area andthen open it to the process of succession that we will reviewin Module 15.7. The community structure during successionis in constant change. Depending on the community, it cantake many years for an equilibrium to be reached after distur-bance. In some instances, an equilibrium may never bereached before the next major disturbance, or before thehabitat and its resources are consumed.

RoseCh15_0104043_437-474_2p 11/18/04 2:55 PM Page 444

Equilibrium and Nonequilibrium Communities 445

FIGURE 15.4C Destruction of Vegetation Following the Eruptionof Mount St. Helens

FIGURE 15.4D Destruction of Forest from Fire Although fires canbe man-made, they are a regular disturbance in manyecosystems.

There are also natural processes that rapidly create unoccu-pied habitat for colonization and succession. The volcaniceruptions already mentioned are one such example. Glaciersmay retreat and expose bare soil. Fires and tree falls may createconditions for succession to begin. These processes are referredto as autogenic succession. Unlike degradative succession, au-togenic succession usually culminates in a mature community,referred to as the climax community. Oftentimes the speciesthat appear early in the successional process may alter environ-mental conditions in a way that will permit the later succes-sional species to invade and survive. For instance, when theglaciers in Alaska retreat, the exposed soil is first colonized bymosses and shallow-rooted herbs. A little later, alder plants in-vade. Both the herbs and alder have the ability to fix nitrogen.In time they will increase the nitrogen content of the soil sub-stantially. Alder also acidifies the soil. The result is that the soilthen becomes suitable for larger trees, like Sitka spruce.

In some places there may be a gradual change in thespecies composition as a result of externally changing physi-cal or chemical conditions. These types of changes are some-times referred to as allogenic succession. An example wouldbe the changes that occur as silt accumulates at the mouth ofa river system. This is a gradual process that transformsbrackish water to soil. As a result, terrestrial species may grad-ually colonize this new land and displace species adapted tothe brackish water conditions. ❖

Other disturbances are less dramatic and may be impor-tant determinants of the species diversity in the community.For instance, forest fires have been a regular disturbance insome ecosystems well before humans arrived (Figure 15.4D).As humans began to manage forests, they naturally tried toavoid the destruction of forests by fires, including fires ofhuman origin and those due to natural factors, such as light-ning. It was soon realized that the health and composition ofsome forests actually required occasional fires and thathuman intervention to stop all fires was ill advised. We willsee in Module 15.5 that, for some communities, moderatelevels of disturbance can actually increase species diversitycompared to communities with no disturbance.

Island Biogeography Islands are very special commu-nities that rely on colonization from distant locations to pop-ulate communities with plants and animals. Some of thebasic forces determining species diversity on islands were firstdescribed more than 30 years ago in a book by RobertMacArthur and Edward Wilson. Their keen insights haveproven to be generally accurate and form the core of present-day theory in island biogeography. We will review some ofthese important ideas in Module 15.6.

Succession We have already introduced the concept of bi-ological succession in Module 15.1. There are several circum-stances that create opportunities for biological succession.Organisms die, snakes shed their skin, and large mammalsleave substantial amounts of organic matter in their feces. In allthese examples, the sudden appearance of organic matter cre-ates a new habitat that can be colonized through a serial re-placement of organisms that we call degradative succession.Of course this succession process ends once the organic mate-rial has been consumed. We will consider this process in detailin this module and examine how this process of succession isused to produce forensic evidence in criminal cases.

RoseCh15_0104043_437-474_2p 11/18/04 2:55 PM Page 445

In 1961 the ideas of competitiveexclusion (see Chapter 12) pre-sented a problem for G. E.Hutchinson. His work with ma-rine and freshwater planktondemonstrated that manyspecies of phytoplankton cancoexist in the same top layers ofwater. However, in these topwaters, several resources, suchas nitrogen and phosphorus,were often in limited supply.Why didn’t competition driveall but one species of phyto-plankton to extinction?

In a paper entitled “TheParadox of the Plankton,”Hutchinson suggested an an-swer. He argued that the com-petitive relationships amongplankton species were specificto a particular set of environ-mental conditions. If these en-vironmental conditionschanged before the superiorcompetitor could achieve nu-merical dominance, thenmany species might persist in alocality for some time.

The role of environmentalvariation in determining speciesdiversity has been important inecological thought. In the inter-tidal zone, for example, manyplants and animals compete for space. In an undisturbed envi-ronment, species diversity will decrease as the competitivelydominant species eliminates other species. However, the inter-tidal zone is subject to disturbance. Waves or storms may turnover rocks. When this happens, some of the plants or animalson these rocks may be displaced or killed. As the graph inFigure 15.5A shows, when disturbance is very low, we have lit-tle displacement of species (green line) and high levels of com-

petition (red line). Theresult is low species di-versity, with the com-petitively dominantspecies being most nu-merous. On the otherhand, when disturbancelevels are high, there is


FIGURE 15.5C A gap in a forest is occupied by new growth. Gapslike this may be caused by mature trees being blown down duringstorms or falling down after being attacked by insect pests.

Disturbance

Displacement

Competition

FIGURE 15.5A As environmental disturbance increases,displacement of organisms from rocks increases. However,competition is greatest when there is little disturbance.

15.5 The diversity of species in a community may depend on environmental disturbance

little competition (red line), butspecies are constantly displaced(green line), so only a few goodcolonizing species are found onthe rocks. Therefore, ecologistsexpect the highest levels ofspecies diversity on intertidalsurfaces with intermediate levelsof disturbance. This idea isknown as the intermediate dis-turbance hypothesis of speciesdiversity.

The intermediate distur-bance hypothesis was tested inthe intertidal zone by WayneSousa. Sousa studied the speciesdiversity on intertidal rocks ofdifferent sizes. He reasoned thatlarge rocks would be movedonly by greater forces, so plantsor animals on these rocks wouldbe displaced less often thanthose on small rocks. The num-ber of species of plants and ani-mals on rocks in three sizecategories were measured. AsFigure 15.5B shows, the mostspecies were found on rocks ofintermediate size. Thus, speciesdiversity on intertidal rocks isconsistent with the intermediatedisturbance hypothesis.

Similar forms of disturbancecan be found in other commu-

nities. For instance, in forests strong winds or attacks by in-sects may result in large trees falling (Figure 15.5C). Thefallen tree no longer shades the soil surface, and many newplants may find opportunities to grow in these types of openareas. ❖

Nu

mb

er o

f Sp

ecie

s

Small Medium Large

2

4

3

0

1

FIGURE 15.5B Small rocks tend to be dominated by thegreen alga Ulva. Large rocks are dominated by the red algaGigartina. The medium-sized rocks have a much larger arrayof species including Ulva, barnacles (Cthamalus), seaanemones (Anthopleura), and additional species of algae(Gelidium and Rhodoglossum).

RoseCh15_0104043_437-474_2p 11/18/04 2:58 PM Page 446


The number of species on islands represents a balance between 15.6extinction and immigration

Small

Large

Rat

e

Number of species

FIGURE 15.6C Extinctions will be less likely in large populationsthan in small populations. Small islands will generally supportsmaller populations and thus have higher extinction rates thanlarge islands will.

Biogeographers study the distribution of species. Their pointof view is that “islands” may be small parcels of land sur-rounded by water, or they may be mountain peaks surround-ed by valleys. The important requirement is that suitablehabitat is surrounded by inhospitable habitat. Plants and an-imals will occupy the “island” only if they travel from some“mainland” location to the island, or from another island.When a new plant or animal species successfully establishesitself on the “island,” we say it has successfully immigrated.The persistence of species on an island requires its successfulreproduction and the growth of its population to substantialnumbers. In general, the larger the population, the longer itspersistence. Eventually all populations go extinct. On an is-land, the only way for an extinct species to become reestab-lished is by immigration. Ultimately the number of specieson an island will represent a balance between immigrationsand extinctions (Figure 15.6A).

What factors affect immigration and extinction? Thenumber of species already on an island is one factor. Rates ofsuccessful immigration are high when species numbers on anisland are low, for several reasons. When species numbers arelow, most newly arrived plants and animals represent speciesnot currently on the island, so there will be fewer competitors.Extinction rates will be low when species numbers are low.The reasons for these low extinction rates are basically the op-posite of the reasons for high immigration rates: Few speciesare at risk of going extinct, because there is little competition.

Rates of extinction and immigration also depend on fac-tors other than species numbers. Immigration rates are affect-ed by proximity to mainland sources of plants and animals.Islands that are near the mainland have higher immigration

rates than do more distant islands (Figure 15.6B). Extinctionrates should be sensitive to island size. All other things beingequal, small islands should support smaller populations ofany particular species, making them more vulnerable to ex-tinction (Figure 15.6C).

These theoretical expectations can be put together tomake predictions about the relative number of species on is-lands as a function of their size and distance from mainland.For instance, if the useful habitat of a particular island weresubstantially reduced, extinction rates should increase. How-ever, immigration rates should remain the same, so the num-ber of species on the island would be expected to fall.

Daniel Simberloff conducted an experiment to test this pre-diction on mangrove islands. Over a period of three years,species counts were made on these small islands and then por-tions of the habitat were cleared and removed from the island(Figure 15.6D). In all cases, reductions in habitat area resultedin observable reductions in the numbers of species. ❖

Immigration

ExtinctionRat

e

Number of species

FIGURE 15.6A The number of species on an island will representthe balance between immigration and extinction. The arrowshows this equilibrium.

Near

Imm

igra

tio

n r

ate

Number of speciesFar

FIGURE 15.6B Immigrants from the mainland are more likely toreach nearby islands than distant islands.

90

80

70

60

50

Area

Spec

ies

nu

mb

er

50 100 250 500 1000

First year Second year Third year

FIGURE 15.6D Changes in Species Numbers on Small IslandsEach island was followed for three years. The occupied part ofthe island was reduced, usually in years two and three. Thereduction in island area should be accompanied by increasedextinction rates and lower numbers of species on the island. Thisgeneral expectation was observed in each case.

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 447


15.7 Habitats go through predictable changes in species compositionover time

Most of us have observed the biological changes that occurwhen a new piece of habitat becomes available for species tooccupy. This can happen when a wooded area is cleared forconstruction or farming, or when fruit falls to the groundand is then left undisturbed. In each case, we see certain or-ganisms making use of the new resources in these habitats.For instance, cleared land might first be occupied by smallweedy plants. If the land is left undisturbed for a sufficientlylong time, these weeds are displaced by larger bushes. Andafter very long periods of time, the species of tree that origi-nally occupied the land takes over as the dominant species.

These transitions in community structure over time arecalled ecological succession. Geologic events such as volcanoesand glaciers may leave areas with no existing vegetation. Thedevelopment of communities on these sites is called primarysuccession. Other places may already have established vegeta-tion and leave seeds and well-developed soil after a distur-bance. The changes in these habitats are called secondarysuccession. The end of succession often yields a community ofstable species composition. This community is called a climaxcommunity. As we mentioned earlier, degradative successiondoes not produce a climax community, because it ends withthe exhaustion of some resource. We consider an application ofthis successional process next.

The sequences of species that characterize degradativesuccession depend on the habitat available. These species se-quences are sufficiently reliable that they can be used in crim-inal investigations to determine the time of death of corpses.The field of forensic entomology uses the principle of ecolog-ical succession to determine when a dead body began to de-compose. As a body decays, the moisture content and pHchange in such a way that the types of insect species living onthe corpse go through a predictable sequence of changes orsuccession (Figure 15.7A). Bodies that are found within thefirst day or so of death can have the time of death estimatedby the change in the temperature of the body. However, oncea body has reached ambient temperature, other techniquesmust be used. The identification of the insects on the decay-ing body is one means of making these estimates. The appli-cation of these techniques to forensics was developed byPierre Mégnin in 1894. There have since been many usefulapplications of these techniques.

Dr. Buck Ruxton, a London physician, had a stormy re-lationship with his wife, Isabella. On September 15, 1935,Ruxton’s wife and their housekeeper, Mary Rogerson, wereseen for the last time. Dr. Ruxton said they had gone onholiday to Scotland. On September 29, the parts of two fe-male bodies were found floating in the River Annon. Fin-gerprints identified one body as Mary Rogerson. Thepresence of third instar larvae of the blowfly Calliphora vic-ina (Figure 15.7A) was used to determine that the two

FIGURE 15.7A Forensic Entomology

Calliphora vicina

Sarcophaga carnaria

Piophilia casei

Ophyra leucostyoma

Derestes lardarius

Tineola biselliella

Initially fresh the corpse is subject to bacteria decay and the bloating from gasses. Larval forms of these insects feed on blood and tissues, first focusing on the heart and digestive system

After fermentation of fats corpse continues to dry

State of CorpseTime Insect FaunaFirst 3 months

3–6 months

4–8 months

1–3 years

Remaining body fluids are now absorbed

The corpse is completely dry

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 448


FIGURE 15.7B The Body Farm at the University of Tennessee

women had been dead for 12–14 days. Based on this andother evidence, Dr. Ruxton was convicted of murder andexecuted on May 12, 1936.

Much of the scientific study of insect succession in decay-ing corpses has been done not with humans, but with othervertebrates such as pigs and dogs. Because it is not clear if theresults on these smaller animals would be similar to those ona human corpse, Dr. William Bass has created an experimen-tal site to study this problem (Figure 15.7B). On a three-acrewooded lot near the University of Tennessee campus, he hasestablished an outdoor laboratory known informally as the“Body Farm.” At this location, human cadavers have beenplaced under different conditions to monitor their progres-sive change over time. Some bodies are embalmed, othersnot. Some are buried under a carpet of leaves, while others lieon the surface, exposed to the elements. This research willhelp law enforcement officials determine date of death, oftenthe most important piece of information in determining guiltor innocence in murder cases. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 449


COMMUNITY ORGANIZATION

15.8 The diversity of a community may be affected by competition,predation, or primary productivity

We have already seen how the members of a community de-pend on each other for energy. This dependence structurestheir ecological relationships. The most obvious pattern is re-duced biomass at higher trophic levels.

Communities can also be understood based on the diver-sity of the species of which they are composed. Species diver-sity is the number and relative abundance of species in acommunity. Among the factors that shape species diversityare competition and predation, which we reviewed inChapters 12 and 13. These processes can affect the number ofspecies coexisting on the same trophic level or on differenttrophic levels.

Interspecific Competition In Chapter 12, we exam-ined the conditions for the coexistence of competing species.We saw that if intraspecific competition is stronger than in-terspecific competition, two species can coexist. The most detailed examination of the Lotka-Volterra competitionequations (see Module 12.7) was carried out by Vandemeer.

In his experiment, three species of Paramecium and a speciesof Blepharisma were raised separately to estimate their carry-ing capacity and intrinsic rates of increase, when growing ontheir own (see Module 10.6). Then pairs of species wereraised together to estimate their competition coefficients (seeModule 12.7). Finally, all four species were placed togetherand allowed to grow. Figure 15.8A shows the observed num-bers of P. aurelia and P. bursaria. The population sizes pre-dicted from the Lotka-Volterra equations are also shown as asolid line. The Lotka-Volterra theory not only correctly pre-dicted the extinction of P. bursaria, but it also did reasonablywell at predicting the actual numbers of Paramecium.

Interspecific competitors usually occupy the sametrophic level. In some communities, however, an importantlimiting resource may affect species at several trophic lev-els. For example, space in the intertidal zone is a limitingresource for species at many different trophic levels, as de-scribed in Chapter 7.

Predation Predation and herbivory involve feeding rela-tionships between species on different trophic levels. However,the ecological factor controlling the numbers of predators orherbivores is not necessarily their food source, as we will see inModule 15.9. For instance, most terrestrial herbivores live inenvironments with an enormous amount of available plantmaterial. This suggests that the numbers of herbivores in a bio-logical community is not related to the amount of food in anysimple way. In 1960 Hairston, Smith, and Slobodkin suggestedthat it is more likely that the numbers of herbivores are regulat-ed by predators that reduce herbivore numbers well belowwhat can be supported by the primary production. This hy-pothesis is sometimes called “why the earth is green.” It is anexample of top-down regulation. However, not all herbivoresand predators have this type of top-down regulation. Later inthe module, we consider additional possibilities.

Food Webs Food webs show the feeding relationships be-tween organisms. A species or group of species is represented bya node (point) in the food web. A link (line) connects twonodes, indicating a predator-prey or plant-herbivore relation-ship. Links may be either undirected or directed, as Figure 15.8Bshows. A directed link shows the flow of energy between twospecies, while an undirected link only indicates that a feedingrelationship exists (Figure 15.8B). A cycle exists when twonodes feed on each other. When two or more species make aclosed circuit, this is called a loop. A chain is a series of directedlinks starting from a species that feeds on no other species andending in a species that is not fed on by any other species. Thenumber of links in a chain is called its length.

Log

(po

pu

lati

on

siz

e)

Time (days)

4 16 32

Paramecium bursaria

4 16 32

Paramecium aurelia

FIGURE 15.8A The numbers of P. aurelia and P. bursaria whenthese species are in competition with themselves and P. caudatum and Blepharisma. Solid lines are the predictions fromthe Lotka-Volterra equations; circles are the observed numbers.

RoseCh15_0104043_437-474_2p 11/18/04 2:51 PM Page 450

Community Organization 451

Directed link Undirected link

Cycle

Loop

FIGURE 15.8B Components of Food Chains See text for details.

There are several types of food webs. A source web ariseswhen a group of species derive all their energy from a singlefood source. A sink web arises when the feeding relationshipsdirect all energy ultimately to a common top carnivore. Acommunity web shows the feeding relationships of all mem-bers of a community. The mean chain length of a web is thearithmetic average of the lengths of all its component chains.

Understanding the factors that affect the mean chainlength of food webs has been an active area of research inecology. Several hypotheses have been proposed, and we willreview evidence for some of them in this module. The ener-getic or productivity hypothesis suggests that the meanchain length will be proportional to the amount of energy atthe primary producer level. Because energy will be lost with

each link in the chain, the total number of links should de-pend on the amount of energy at the base level.

The dynamical stability hypothesis suggests that longchains will be inherently less stable. Thus, longer chains aremore likely to be found in benign environments where popu-lations are not subject to large fluctuations in population size.

The ecosystem-size hypothesis suggests that chainlengths will be greater in ecosystems with greater physicalvolume, all other things being equal. This is a complicatedtheory, but basically a greater number of different speciescan be supported in larger areas. It happens that averagechain length increases with increasing species numbers andthus average chain length also increases with increasingecosystem size. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 451


15.9 The number of species in a community may depend on predation

Some of the most spectacular examples of the effects ofpredators involve biological control. In Chapter 14 welearned about the explosive growth of rabbits introduced toAustralia at the end of the nineteenth century. The intro-duction of a viral parasite was effective at reducing thenumbers of rabbits. A similar example is the aquatic fern,Salvinia molesta, that is native to waterways in Brazil. Thisfern has become a pest in much of the tropics. Its densepopulations block waterways and prevent fishing. Effective

control has been achieved using a small weevil, Cyrtobagussingularis. The adult weevil feeds on the buds of the fern,while the larvae feed on the plant’s roots and rhizomes. Thespecificity of this interaction was demonstrated by the fail-ure of the first control efforts. A weevil from a closely relat-ed fern, Salvinia auriculata, was tried but failed to control S.molesta, even though these two species were thought to bethe same for some time.

The effects of Pisaster on th e species below is indicated by a “ +” wh en the affected species shows an increase in numbers, by a “–” wh en their numbers decline, and by a “ 0” wh en there

Chitons (–)two species

Porphyra (0)Endocladia (–) Rhodomela (–)

Algae

Carolina (–)

Limpets (–)two species

Mytilus (+)Acorn barnacles (–)

3 species

Mitella (+)

Thais (+)

Pisaster ochraceous

FIGURE 15.9A Pisaster, a Keystone Predator in the Intertidal Zone

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 452


D. putrida

D. falleni

D. tripunctata

40

60

0

20

�20

�20

Perc

ent

incr

ease

du

e to

pre

dat

ion

FIGURE 15.9B Effects of a Predatory Beetle on the Composition of a DrosophilaCommunity In the absence of the beetle, D. tripunctata is competitively superiorto D. putrida and D. falleni, often eliminating them. When the beetle—whichfeeds on all three species—is added to the community, the numbers of D.tripunctata fall and the numbers of the other two species increase. All threespecies coexist at similar population sizes in the presence of the predator.

These examples demonstrate the potential for predationand herbivory to affect the numbers of a prey species, but notthe numbers of species. One of the earliest and most influen-tial demonstrations of the importance of predation on speciescomposition was a study by Robert Paine on an intertidal ma-rine invertebrate community. In this study, the ecology of acarnivorous starfish (Pisaster ochraceous) was observed at twodifferent research sites. In the experimental site, the starfishwas removed and kept out of the area. In the control site, thestarfish were allowed to forage as usual. A total of 15 differentspecies were monitored. In the experimental site, the speciesrichness, or number of species, dropped from 15 to 8, while inthe control area it remained unchanged.

What explains these results? The decline in species num-bers in the experimental area was due to increased numbersof the competitively dominant mussel Mytilus. Other speciesof animals and plants were eliminated because the removalof the starfish led to the explosive growth of Mytilus. InFigure 15.9B, the species that were negatively affected by the

removal of starfish have a negative sign next to their names.A positive sign indicates a benefit to the species due tostarfish removal, whereas a 0 indicates no effect. Paine calledthe starfish a keystone predator, because of its central rolein maintaining species diversity in this community.

Species numbers in terrestrial communities may also beaffected by keystone predators. Wade Worthen studied threespecies of mushroom-feeding Drosophila. These fruit fliesserve as food for the predatory rove beetle Ontholestes cin-gulatus, which eats adult Drosophila. In the absence of thebeetle, there is strong interspecific competition between thethree species of Drosophila, with D. tripunctata often elimi-nating the other two species. However, when the beetle wasadded to experimental cultures, the level of competitionamong Drosophila larvae was reduced, and all three speciescoexisted. As Figure 15.9A shows, beetle predation affectedthe competitively dominant species D. tripunctata most.With fewer D. tripunctata, the other two species were able toincrease their numbers and so stably coexist. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 453


15.10 River communities show a top-down structureThe starfish study by Paine demonstrates the potential forpredation to shape the diversity of species in the trophiclevel directly below it. The theory of food webs also pre-dicts that in long food chains the qualitative effects of re-moving top predators should alternate as one goes down tolower trophic levels. Thus, this theory would predict that ina community with four trophic levels (producers-herbi-vores-small carnivores-top carnivores) removal of the toppredator would have a beneficial effect on the small carni-vores. However, once the number of small carnivores in-creased, that would cause a negative effect on theherbivores. The reduction in herbivore numbers would

then have a positive effect on the primary producers. Thispredicted series of positive and negative effects is some-times called a trophic cascade.

To investigate whether communities showed such cas-cades, Mary Power undertook a study of a river community(Figure 15.10A). This community also has four trophic lev-els, as does the example used in the previous paragraph. Atthe bottom are green algae, Cladophora and Nostoc. Theseare consumed by herbivorous insect larvae called chirono-mids. Chironomids are related to mosquitoes. The chirono-mids are eaten by predatory insects and juvenile fish, like thestickleback. The rivers that Power studied in northern Cali-

fornia also have large fish, steelhead androach, that feed on the small predators(Figure 15.10A).

Power introduced replicate cages cov-ered with a small mesh that permittedsmall insects to pass through freely, but notfish. At random, some cages were designat-ed as enclosures and others as exclosures.Each enclosure was stocked with 20 steel-head and 40 roach fish. The exclosureswere kept free of large fish. After five weeks,samples were made of all the enclosures todetermine the numbers of predators andthe biomass of algae. The results were con-sistent with the predictions of a trophiccascade. The small predators all showed in-creased numbers in the exclosures relativeto the enclosures. However, the next lowerlevel in the food web, chironomids, showeda dramatic decline in the exclosures. Thisdecline of chironomids was accompaniedby an increase in algal biomass. Thus, thechange in numbers of top carnivores wasfollowed by a series of changes that were al-ternatively positive and negative at differ-ent trophic levels. ❖

The two examples we have just considered show how strong the ef-fects of top predators in a marine and freshwater community canbe. The evidence for such effects in terrestrial communities is lesscompelling, however. This finding has led some to suggest thatthere may be a difference between aquatic and terrestrial commu-nities that will naturally give rise to aquatic communities experi-encing more top-down regulation. Possible reasons for suchdifferences include the following: (1) Terrestrial food webs aremore complex than aquatic food webs. This complexity mightmake it less likely that the effects of top predators would work theirway down to primary producers. Terrestrial plants are also morelikely to protect themselves from herbivores by producing toxic sec-ondary chemical compounds. This would also make such plants in-

sensitive to changes in numbers of herbivores. (2) Timescales andturnover rates are faster in aquatic systems. Because the plankton atthe top layer of the water column contain many small algae, theirgrowth rates are rapid. This may lead to fundamental differencesbetween terrestrial and aquatic systems, or it may simply make iteasier to detect strong food-web interactions in aquatic ecosystems.

However, other scientists have argued that an examination oflarge numbers of studies, not just a few prominent ones, suggeststhat terrestrial interactions are not substantially different fromaquatic communities. There is a lot of heterogeneity within aquaticand terrestrial communities, making it difficult to reach generalconclusions about the average behavior. At this time, more work isneeded before these questions can be settled.

Are Aquatic Food Webs Different from Terrestrial Food Webs?

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 454

1000

500

0

40

20

0

exclosureenclosure

AlgaeChironomids

Predatory insects

Fish fry

1.2

0.6

0

40

20

0

Roach fryPredatory insects Stickleback fry

Chironomids

Algae

SteelheadLarge Roach

Cages were set up in the reiver to prevent large fish from entering the study site. In cages called enclosures both Roach and Steelhead were put inside. In the cages called exclosures no large fish were added.

After five weeks the numbers of each food web level were censused. The cages that excluded large fish (exclosures)showed an increase in predatory insects and fish fry, a large decline in chironomids, and an increase in algae compared to the enclosures.

FIGURE 15.10A Effects of Predation in River Food Webs


RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 455


A number 1 in the matrix indicates a feeding relationship that

1. Algae2. Chironomids3. Predatory insects4. Roach fry

1 2 3 4 5 6 7

5. Stickleback fry6. Large roach7. Steelhead

Predators

Prey

1 0 1 0 0 0 1 02 0 0 1 1 1 0 03 0 0 0 0 0 1 14 0 0 0 0 0 0 15 0 0 0 0 0 0 16 0 0 0 0 0 0 07 0 0 0 0 0 0 0

FIGURE 15.11A A Food-Web Matrix

15.11 Many features of food webs can be described by the cascade modelSeveral theories have been proposed concerning the proper-ties of food webs, such as the mean chain length, maximumchain length, and total number of links. One simple theory,called the cascade model, provides numerical estimates of allthese quantities. Before describing this model, let’s look athow food web data are summarized.

In Figure 15.11A, the food web for the river communityconsidered earlier is shown. A community food web matrixlists the trophic species that are prey as row entries (boldnumbers) and those that act as predators in the column en-tries (bold numbers). If a 1 appears in the matrix, it indicatesthat the species in that column feeds on the species of thatrow. A zero means no feeding relationship exists. Quantita-tively this is an easier way to summarize a food web than thepictures that show connections between species. A basalspecies is one that is prey for one or morespecies and does not eat any other species. Inthe food web matrix, a basal species will have acolumn of zeros under its number. In Figure15.11A, the algae are the only trophic speciesmeeting this definition. A top species is onethat feeds on one or more species but is not it-self prey for any other species. A top speciesshould have a row of zeros next to its numberin the food web matrix. Both the steelhead andlarge roach are top species in Figure 15.11A.The diagonal elements in this matrix are thecells where the same-numbered row and col-umn intersect. A 1 at these positions wouldimply that the species can eat itself. In this foodweb none of the species are cannibals, becauseall diagonal elements are zero.

Joel Cohen and Charles Newman have de-veloped a simple model to study the propertiesof food webs. We review the assumptions of themodel in Figure 15.11B. It is assumed that thereare no loops or cycles in the food web. So nospecies are cannibals. This assumption meansthat all the diagonal elements will be zero andall the elements below the diagonal will also bezero. This also means that the species in thecommunity can be organized as a cascade (seeFigure 15.11B). Trophic species may feed onspecies with lower numbers, but are never fedon by species with lower numbers.

The model then assumes that feeding rela-tionships are created at random and that thechance of any allowable relationship formingis determined by a common probability, p.From this simple formulation, the modelpredicts with some accuracy the mean chainlength in many different communities (seeFigure 15.11B). This may seem surprising,because community food webs are not put

The cascade model can be used to predict the mean chain length of a food web. The figure on the right shows the predicted mean length versus the observed length for 113 different community food webs. The solid line indicates prec ise agreement between the predictions and the observations.

The trophic species within a community are assumed to form as a cascade. That is, species are ordered so that those with the highest number can feed on any species below them; however, a trophic species cannot feed on a species above them.

In the figure on the left the arrows indicate the direction of energy flow. Trophic species 3 can potentially feed on species 2 and 1, but not on species 4.

The properties of food webs are studied by

assuming that they are constructed at random. Thus, using the rules in the paragraphs above, the chance that any pair of trophic species will form a feeding relationship is a constant, p. This

Because each of these food webs has two links, the probability of each forming would be p2.

Consider a community with the four trophic species illustrated above. All possible food chains having two links with species 1 as the basal species and species 4 as thetop species are shown below.

4

3

2

1

4 3 1

4 2 1

"Pre

dic

ted

mea

n"

chai

n le

ng

th

7

5

3

1

Observed mean chain length3 5

FIGURE 15.11B The Cascade Model of Food Webs

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 456


together at random. The ability of the cascade model to dosuch a good job of predicting mean chain length suggeststhat while food webs are not created randomly, the detailsof how they are created are not needed to make accuratetheoretical predictions.

This is not unusual in science. It is possible to create a de-tailed theory of the kinetics of a coin placed in motion by theflip of our finger. We could take into account the forces thatour finger generates and the atmospheric conditions—par-ticularly the presence of wind—as well as the position atwhich the coin is typically caught, and then predict how oftenwe should get a head or tails. If we do this calculation correct-ly, we would see that the coin will come up heads about 50percent of the time. Or we can, and often do, use a simple sta-tistical model to predict the chances of getting heads without

worrying about all the details of the forces that affect the tra-jectory of coins. One such model is simply heads half thetime, tails the rest.

One interesting result from the theoretical cascade modelis that the mean chain length increases slowly with total num-ber of species in the food web. This result can be combinedwith the theory of island biogeography to arrive at a predic-tion about the mean chain length and size of the habitat.From the theory of island biogeography, we know that therate of extinction decreases with island size. If all other vari-ables remain constant, then the equilibrium number ofspecies should increase on larger islands. This result, in com-bination with the conclusions that mean chain length in-creases with species numbers, suggests that mean chainlength increases as community size or volume increases. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:32 PM Page 457


15.12 Food-web chain length is proportional to ecosystem size in lakes

In the last module we noted that the cascade model and the the-ory of island biogeography predict increasing chain length inlarger communities. To test this idea requires samples frommany communities that vary in size. There would also have to besome way to rapidly collect information on the mean chainlength of each community. David Post and his colleagues (2000)have been able to collect this sort of information for lake ecosys-tems. They also collected information on the productivity sothat they could simultaneously compare mean food-chainlength to productivity.

If mean chain length depends only on ecosystem size, wewould expect the data collected by Post to look something likeFigure 15.12A. Mean chain length would increase with the size ofthe ecosystem independently of the productivity. Likewise,ecosystems that varied in productivity would show no consistenttrend in mean chain length. So what did the results look like?

In Figure 15.12B, we summarize the results of Post’s study.Ecosystem size was estimated from the volume of the lake, whichis fairly easy to do. The lakes studied were in the northeasternUnited States. In these lakes, primary productivity is limited bythe amount of available phosphorus. Levels of total phosphorus(TP) are highly correlated with the primary productivity andmay be used as an estimate of primary productivity. A detailedstudy of the feeding relationships of the entire community ofeach lake would take an enormous amount of time. To estimatethe mean chain length of each lake, an ingenious method thatutilizes radioisotopes of nitrogen was used (Figure 15.12C). Thistechnique is objective and allows information to be collected ona large number of communities.

When the data are analyzed, we can see that there is a verystrong positive relationship between ecosystem size and meanchain length (see Figure 15.12B). This relationship holds for

All systems

Large system

Small system

Foo

d-c

hai

n le

ng

th

Ecosystem size Productivity (per unit size)

Foo

d-c

hai

n le

ng

th

If productivity was unimportant, then there should be no change in food-chain length with increasing productivity, no mater how large the ecosystem.

The ecosystem size hypothesis predicts that the food-chain length of a community will increase with increasing ecosystem size, no matter what the productivity.

FIGURE 15.12A Ecosystem Size and Food-Chain Length

5.5

5.0

4.5

4.0

3.5

5.5

5.0

4.5

4.0

3.5

Mea

n c

hai

n le

ng

th

Mea

n c

hai

n le

ng

th

High productivityModerate productivityLow productivity

Productivity (total phosphorous, μg/L�1)103102101100

Ecosystem size (volume m3)1013109 1011107105

Large lakes

Medium lakes

Small lakes

The mean chain length was estimated by the stable isotope technique (Figure 15.12C) in several lakes that varied in size and productivity.The mean chain length increases in direct porportion to size, no matter what the productivity of the lake is.

For these same lakes examined above, there was no consistent relationship between mean chain length and productivity.

FIGURE 15.12B Tests of the Ecosystem Size Hypothesis

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 458


14N and 15N are naturally occurring isotopes of nitrogen. Their chemical and physical properties are similar but not identical.

Diffusion14N

14N

15N

15N

The rate of some chemical reactions will differ between isotopes. For instance, we expect the lighter 14N molecule to diffuse faster due to its smaller mass.

In reversible chemical reactions, those compounds with stronger bonds tend to have more of the heavy isotope.

These differences in the behavior of isotopes can lead to changes in the relative amounts of isotope as molecules work their way through food webs. In animals nitrogen waste products, like ammonia and urea, tend to have more 14N than the animals’ food. The tissues of animals then tend to have less 14N than their food and more 15N. This decrease in relative amounts of 14N and increase in 15N continues as protein is passed up the food chain.

Protein Protein Protein

Plant Herbivore Predator

Nitrogen wastes Nitrogen wastes

FIGURE 15.12C Radioisotope Measurement of Food-Chain Length

lakes that varied in productivity. On the other hand, there is ei-ther no relationship or a very weak relationship between pro-ductivity and mean chain length. These observations providestrong empirical support for the importance of ecosystem size inthe structure of food webs. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 459

15.13 Increased productivity can increase food-chain length but decrease stability

In this chapter, we have already seen evidence that increasesin primary productivity lead to increases in the biomass ofherbivores. This observation leads naturally to the hypothesisthat increases in primary productivity might also lead to in-creases in the biomass of predators that feed on herbivores,and so on. Furthermore, it is conceivable that the total num-ber of trophic levels—that is, food-chain length—might re-spond to changes in primary productivity.

This idea has been investigated experimentally by Jenk-ins and his colleagues (1992). They studied communities ofbacteria and insects that live in treeholes in the Australiantropics. These communities can be found in a variety ofplants that collect water, like bromeliads (Figure 15.13A).Decaying leaves and animals provide the primary energysupply to these communities. Jenkins was able to replicatethese communities artificially by placing plastic containersunder trees. Each container initially started with some waterand decaying leaves. The amount of decaying leaves wasvaried over three levels: high, medium, and low. The num-bers of species and their feeding relationships were deter-mined at regular intervalsover a 48-week period.Figure 15.13B shows theresults. As the levels of pri-mary energy to the communi-ty were increased, the food-webstructure became more complex.That is, there were more species in thecommunity and more trophic links. Onaverage, the maximum food-chain lengthwas greatest at the highest productivity. Thus,this study supports the idea that food-chainlength is positively correlated with primaryproductivity.

Are there other factors that might limitthe length of food chains? Is it rea-sonable to suppose thatfood chains will be-come longer andlonger as pri-mary pro-ductivity isincreased?One problemthat must be

FIGURE 15.13A The leaves and central tank of the bromeliadsupport a community of bacteria and insects.


RoseCh15_0104043_437-474_2p 11/18/04 2:48 PM Page 460


Weeks0 10 20 30 40 50

Weeks0 10 20 30 40 50

Weeks0 10 20 30 40 50

Nu

mb

er o

f sp

ecie

s

5

4

3

2

1

0

Nu

mb

er o

f tr

op

hic

lin

ks

7

6

5

4

3

2

1

0 Max

imu

m f

oo

d-c

hai

n le

ng

th 4

3

2

1

0

LowMediumHigh

Over the first 24 weeks, the number of species increases in all treatments. However, the number of species is always greater in the treatments with higher productivity.

The pattern for the number of trophiclinks is essentially the same as for the numbers of species.

There is essentially no difference in the maximum food-chain length between the low- and medium- productivity treatments. However, the high-productivity treatment shows a greater food-chain length at all sample points.

FIGURE 15.13B Community Structure as a Function of Primary Productivity

Log

(N +

1)

4

3

2

1

0

0 5 10 15 20 25 30 35 40 45

4

3

2

1

0

4

3

2

1

0

4

3

2

1

0

Day0 5 10 15 20 25 30 35 40 45

II

I

IV

III

Bacteria Bacteria

Bacteria

Colpidium

Colpidium

Colpidium

Colpidium

Actinosphaerium Blepharisma

BlepharismaBacteria

FIGURE 15.13C The Numbers of Colpidium (squares) Over Time In each panel, the foodchain of the introduced species is shown. Panel I—a two-species community. II—threespecies, two links. III and IV—the same three species are shown in both these panels but indifferent replicate cultures. In III, Colpidium persists but fluctuates wildly. In IV, Colpidiumgoes extinct by day 12.

considered is the stability of long food chains. If smallchanges in the numbers of primary producers cause largechanges in the numbers of top carnivores, then even modestvariation may cause top predators to go extinct.

Lawler and Morin (1993) examined the stability of smallcommunities of protists (Figure 15.13C). In the simplestcommunity there were two species, a bacteria and a bacteri-vore (bacteria-eating) protist, Colpidium. In this simplecommunity, Colpidium rapidly reaches its carrying capacityand then changes little in population size over time (panel I).When an additional Actinosphaerium carnivore is added,

Colpidium persists, but its numbers fluctuate (panel II). In asecond community with bacteria, Colpidium, and the omni-vore Blepharisma, the numbers of Colpidium fluctuate wild-ly (panel III), and in one replicate the Colpidium populationgoes extinct (panel IV).

Together these two studies show that food-chain lengthmay depend on both primary productivity and the ecologicalstability of top carnivores. Increasing primary productivitymay increase food-chain length up to a point. Very high levelsof primary productivity may not result in longer food chainsdue to the extinction of top carnivores. ❖

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 461

Rel

ativ

e ab

un

dan

ce

amygdalina amygdalina Backcross

to

amygdalina

Leaf phenotypes from each type of cross

Backcross

to

risdoniiF1

risdonii

risdonii

10

0

40

50

10

20

30

Spec

ies

rich

nes

s

30

15

20

25

Hybrid zone

Two species of Eucalyptus (E. amygdalina, E.risdonii) intermate and form hybrids in parts of Australia. In these hybrid zones the characteristics of the plants range from mostly amygdalina to mostly risdonii, and the F1 hybrids have intermediate traits.The species richness and relative abundance of 40 different insect and fungal taxa in-crease dramatically in comparing the paren-tal strands of Eucalyptus to those consisting of mostly hybrids (F1s). A backcross is a cross between a hybrid and one of the pa-rental species.The bars of the same color have statistically the same values of either species richness or relative abundance.

E.


15.14 The structure of communities is also affected by the genetic structure of its members

We have already seen that many predator-prey and host-para-site relationships can be species specific. It is not unreasonableto think that some interactions may even depend on the geno-types within a population. Certainly, for some parasites wehave already considered the idea that some genotypes may besusceptible to parasites while others are resistant. Althoughmost species consist of many genotypes, plants possess theability to generate large levels of genetic variability in a smallarea by hybridization with other species. It is not unusual tofind closely related species of plants that can form viable hy-brids. In fact, 30 to 80 percent of all plant species may arisethrough hybridization.

The areas where different species meet and form hybrids iscalled a hybrid zone. The progeny that result from the cross oftwo different species are called the generation. The indi-viduals may then mate with one of the parental species. Theprogeny from this type of cross are called back-cross progeny.

F1F1

In Australia, two different species of eucalyptus trees (E. amyg-dalina and E. risdonii) hybridize. These hybrids have very differ-ent leaf morphologies than those seen in either parent (Figure15.14A), and different physiology. These trees harbor a numberof insect and fungal species. Thomas Whitham and his colleagueshave examined the insect and fungal communities of eachparental species of eucalyptus and of trees found in the hybridzone. They measured the species richness of these insect and fun-gal communities, which is a measure of the number of differentspecies and their relative abundance. In general for a community,the more species there are and the more even their distribution,the greater that community’s species richness will be.

This study revealed that the species richness and relativeabundance of insects and fungi was much greater in the hybrid

community than in either of the parental eucalyptus popula-tions (see Figure 15.14A). This effect is seen in natural pop-

FIGURE 15.14A Biodiversity in a Eucalyptus Hybrid Zone

ulations as well as in controlled conditions. These ob-servations show that the genetic structure of an impor-tant member of a community can have a profoundeffect on the species composition.

It is not entirely clear what is causing this effect among theeucalyptus populations. However, these plants produce a largenumber of oils that deter attacks by insects. The hybrids ap-pear to make intermediate levels of these oils. Thus, they maynot have sufficient levels of chemical protection from insects.

These findings are not peculiar to eucalyptus. Cottonwoodcommunities in Utah also form hybrids between Populusfreemontii and P. angustifolia. The bud gall mite appears to spe-cialize on the hybrids of these two cottonwoods [Figure15.14B, part (i)]. It is almost never found on the parentalspecies or even on backcrosses.

The effects of these hybrids are not limited to insects andfungi. The hybrid cottonwoods also have a very different mor-phology than either of these parents [see Figure 15.14B, part(i)]. These hybrids are apparently more attractive to a varietyof bird species and, as a result, many more bird nests are foundin the hybrid zone than in either of the parental populations[Figure 15.14B, part (ii)]. ❖

F1

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 462


FIGURE 15.14B Populus Hybrid Zones

Nu

mb

er o

f n

ests

/100

tre

es

P. fremontii

zone

P. angustifolia

zone

P. fremontii

Populus

fremontii

zone

Populus

angustifolia

zone

F1 types

The hybrid zone suports a much larger number of nesting birds.

The bud gall mite is found almost exclusively on F1 h

ybrids.

Backcrosses

& a few

P. angustifolia

35

0

25

20

30

5

10

15

Hybrid zone

Nu

mb

er o

f g

alls

/tre

e

P. fremontii

P. angustifolia

zone

P. fremontii

Populus

fremontii

zone

Populus

angustifolia

zone

F1 type

and hybrid

Phenotype

Backcrosses

& a few

P. angustifolia

0

900

100200300400500600700800

Hybrid and overlap zoneBud gall mite (Aceria parapopuli)


ECOSYSTEMS

15.15 An important feature of ecosystems and their biologicalcommunities is their interaction with the physical environment

Life on Earth is almost entirely sustained by energy from thesun. Plants form the essential link between the sun’s energyand the energy that is used by virtually all biological life. TheEarth is an open system with respect to energy. Energy comesfrom outside the boundaries of the Earth and its immedi-ate atmosphere (Figure 15.15A).

One consequence of this dependenceon the sun is that certain thermody-namic laws that apply specificallyto closed systems are violated,because the Earth is an opensystem for energy. Inclosed thermodynamicsystems, entropy (ordisorder) should al-ways increase. How-ever, on Earth theorganization ofchemical elementsinto living organ-isms represents a de-crease in entropy.This decrease in en-tropy is possible onlybecause of the flow ofenergy into the Earthfrom the sun.

Nutrient Cycles Life de-pends on other components inaddition to energy. All organismsare composed of various essen-tial elements, such as carbon,nitrogen, sulfur, phosphorus,and others. In addition to theseessential elements, all life re-quires water. We will collectively refer to these required sub-stances as essential nutrients. Because the Earth gets nosignificant amount of these nutrients from outside its bound-aries, the Earth is a closed system with respect to these essen-tial nutrients.

Does life use up these nutrients? Is it possible that we willrun out of these essential nutrients just like we may one dayrun out of fossil fuels? The basic answer is no, because life onEarth recycles these essential nutrients between living organ-isms and nonliving components of the Earth such as the at-mosphere and oceans. These cycles of nutrients are calledbiogeochemical cycles, since they depend on both living andnonliving components. We will review three important bio-geochemical cycles in Module 15.16.

Large pools of nutrients reside in reservoirs on Earth. Fornutrients that have a gaseous stage in their cycle, such aswater, nitrogen, and oxygen, the atmosphere and oceans serveas important reservoirs. For nutrients with sedimentary cy-

cles, such as phosphorus, rocks and soil serve as themain reservoirs.

The movement of nutrients fromone component of the biogeo-

chemical cycle to another iscalled flux. The flux is

measured in the amountof nutrient per unit of

time. The parts of thecycle with large flux-es are key to under-standing thedynamics of thenutrient. Theseparts of the cycle,if perturbed,would be expect-ed to have thegreatest impact

on the availabilityof the nutrient.

Ecosystem Function

Understanding ecosystemsrequires that we understand

the interaction be-tween biologi-cal communities and the physicalenvironment. Indeed, it is im-possible to understand someprocesses completely withoutlooking at the interaction of the

environment and organisms. Many biogeochemical cycles haveimportant biological components. The cycling of nutrientssuch as nitrogen depends critically on the ability of microor-ganisms to carry out important chemical reactions. Physicalprocesses such as weather also depend on the activities of livingorganisms. Plants and animals have played an important rolein the cycling of in the atmosphere. However, over the lastcentury humans have significantly increased atmospheric through the burning of fossil fuels. This rise in is continu-ing to this day (Figure 15.15B).

As we will see in Chapter 16, plays an important role inthe global climate, along with water vapor and nitrous oxide. Asignificant amount of heat energy that would otherwise radiatefrom the Earth back into space is captured by these molecules,

CO2

CO2

CO2

CO2

FIGURE 15.15AThe Earth is an open system for energy, but it is a closedsystem for nutrients essential to life.

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 464

Ecosystems 465

19971993198919851981197719731969196519611957

CO

2 co

nce

ntr

atio

n (p

pm

)370

365

360

355

350

345

340

335

330

325

320

315

310

keeping the Earth warm through the greenhouse effect.This effect refers to the action of our atmosphere that letsvisible and ultraviolet light through to the Earth’s surfacebut absorbs much of the heat energy that the Earth radi-ates back to the atmosphere. There is currently concernthat human production of may be leading to increas-es in global temperature. There may also be a positivefeedback loop between temperature and atmospheric lev-els of In the remaining modules we will consider ev-idence that elevated temperatures may accelerate thedecomposition of soil carbon and its release into the at-mosphere. If these changes were to continue, local climatescould change in a significant fashion.

Atmospheric levels also change as Earth’secosystems change. In Module 15.18, we will see thatchanges in the diversity of biological communities mayalso affect a community’s ability to take up Thus,preservation of species diversity may ultimately be criti-cal to preserving the physical environment on which alllife depends. ❖

CO2.

CO2

CO2.

CO2

FIGURE 15.15B Change in Atmospheric ConcentrationThese measurements, made in Hawaii, show a steady increase inthe levels of over a 20-year period. Within a single yearthere are also small rises in concentrations during the wintermonths. This change corresponds to the decline in uptakeby plants during the dormant winter months.

CO2

CO2

CO2

CO2

RoseCh15_0104043_437-474_2p 11/18/04 2:33 PM Page 465


15.16 Essential nutrients are recycled through biological systemsHere we review the cycles of three important nutrients: water(Figure 15.16A), carbon (Figure 15.16B) and nitrogen(Figure 15.16C). In the hydrologic cycle, the largest reservoirof water is the Earth’s oceans. The oceans are also the compo-nents with greatest flux of water, through evaporation andrainfall, as the numbers in Figure 15.16A show. On land,water is lost by evaporation from the Earth and from plantsand animals. This flux is called evapotranspiration.

The Earth’s atmosphere is composed of 78 percent nitro-gen, 21 percent oxygen, and only 0.03 percent carbon diox-ide along with other trace gases. Nevertheless all plantsdepend on atmospheric for the carbon used in photo-synthesis (Figure 15.16B). In the oceans, carbon dioxide isdissolved in water, where it exists as carbonate ion The carbon fixed by plants either stays in the plant or is con-sumed by animals. Ultimately, both plants and animals die,and then they decompose through the action of microor-ganisms. Much of the carbon content then returns to the at-mosphere after decomposition.

Some of the carbon in dead organisms, however, is lost to thecycle in sediments. Most of these sediments have very low car-bon concentrations, but occasionally there are high concentra-tions of carbon in fossil fuel deposits. Before the advent ofhuman civilization, these reservoirs contributed little to atmos-pheric concentrations. However, over the last 100 yearsthese reservoirs have been returned to the carbon cycle throughhuman burning of fossil fuels. Although this burning also con-

CO2

1HCO3-2.

CO2

sumes atmospheric oxygen, the net change in atmospheric oxy-gen levels has been very small, while the levels in the at-mosphere have significantly increased over the last 100 years.

The nitrogen cycle is more complicated than the other two(Figure 15.16C). This cycle depends critically on the action ofnumerous microorganisms that convert nitrogen from oneform to another. Composed of 78 percent nitrogen, the at-mosphere represents a tremendous reservoir of nitrogen; butrelatively few organisms are able to take atmospheric nitro-gen and convert it to a biologically useful form. Thisconversion is accomplished for plants by nitrogen-fixing bac-teria that live in the soil or in close association with the rootsof certain plants. Both aerobic bacteria such as Azotobacterand anaerobic bacteria such as Clostridium fix nitrogen. Ani-mals get their nitrogen from the consumption of plant or an-imal proteins. As plants or animals die, their proteins andamino acids are converted to ammonium ions by microor-ganisms that derive energy from this process. Ammoniumcan be taken up by plants for their nitrogen needs, or it can beoxidized to nitrate by a process known as nitrification. Thefirst step in nitrification is the conversion of ammonia to ni-trite by the bacteria Nitrosomonas. Nitrite is then ox-idized to nitrate by another group of bacteria, theNitrobacter. Nitrate can then be taken up by plants or re-turned to the atmosphere as nitrogen by the process ofdenitrification. This last process is carried out by bacteria ofthe genus Pseudomonas. ❖

1N22

1NO3

-21NO2

-2

1N22

CO2

FIGURE 15.16A The Hydrologic Cycle All numbers are in units 1000 km3/yr.

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 466

Ecosystems 467

Decomposition

Dead organicmatter

SedimentationShallow ocean

storage (39,000)

FIGURE 15.16B The CarbonCycle Storage units are inbillions of metric tons ofcarbon.

N2 fixationcyanobacteria

FIGURE 15.16CThe Nitrogen Cycle

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 467


15.17 Soil carbon levels are affected by temperatureThe levels of atmospheric carbon dioxide are important fortheir greenhouse effects. The amount of carbon that is storedin soils is two to three times greater than the amounts in theatmosphere. Much of this carbon is in the form of soil organ-ic matter. Because soil organic matter can be decomposedand its carbon released as it is a potentially importantdynamic source of atmospheric carbon dioxide.

Factors that affect the decomposition of organic matter mayalso affect the levels of atmospheric carbon. One obvious fac-tor is temperature. Because the metabolic rates of decompos-ing organisms will increase with increasing temperature, therates at which soil organic matter is recycled into the atmos-phere may also depend on temperature.

In an odd twist of fate, nuclear testing in the western Unit-ed States has provided an opportunity to explore the impor-tant ecological relationship between temperature and soilcarbon levels. Nuclear testing between 1958 and 1963 rough-ly doubled the levels of carbon-14 near the Sierra Nevada

CO2,

mountain range. About the same time, soil samples from theSierra Nevada were taken and stored as archive samples forlater testing.

After 1963, the organic matter in the soils of the SierriaNeveda would be expected to start showing elevated levels ofcarbon-14 as this isotope became incorporated into plants,and these plants died or shed leaves into the soil (Figure15.17A). Of course the relative amounts of carbon-14 in theorganic matter of the soil would depend on how fast the oldorganic matter was decomposing and moving out of the soilcarbon reservoir.

The Sierra Nevada range is also an interesting study site be-cause the mean annual temperature declines steadily and sub-stantially as one moves from the base of the range to thesummits. Using soil samples taken in 1992, Susan Trumboreand her colleagues were able to compare these soils to thearchive samples. From these comparisons, the rates of turnoverof soil carbon could be estimated at many different elevations

CO2

14C

CO2 CO2

Soil carbonSoilsamplestaken

1958–63 1963–1992

Time

Mean annual temperature (°C)

Turn

ove

r ti

me

(yea

rs)

80

60

40

20

00 10 20 30

HawaiiBrazilSierra

After nuclear testing, the levels of 14C from these tests will increase in the soil carbon at a rate that depends on the rate of carbon turnover in the soil. The red indicates the level of soil carbon-14 at different times. This increase can be documented by coparing soil samples in the 1990s with those taken in the period of 1958–1963.

FIGURE 15.17A Soil Carbon Levels and Temperature

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 468

Ecosystems 469

Years since 0.5˚C increase

Tropical

BorealTemperate

0 20 40 60 80 100

8

6

4

2

0

Car

bo

n r

elea

sed

(1015

g)

FIGURE 15.17B Global Warming and Carbon Release

(temperatures). Their study showed dramatically that soil car-bon turnover is much higher at higher temperatures (seeFigure 15.17A). These findings suggest that global warmingwill increase atmospheric carbon dioxide levels still further.

Based on their studies, Trumbore and colleagues estimat-ed the effects of a 0.5°C increase in temperature on carbonlevels in various ecosystems (Figure 15.17B). The effects aregreatest in the tropics. In just a single year, all forests wouldrelease nearly grams of carbon, which is nearly 25percent of the amount released by all fossil fuel consumptionin a year. These findings show the complicated dependence ofglobal nutrient cycles on many factors. ❖

1.4 * 1015

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 469


15.18 Species diversity affects ecosystem performance

FIGURE 15.18A Replicate communities witheither 9 (low diversity) or 31 (high diversity)species were followed over time. In the morediverse communities, there was greater plantcoverage of the soil surface.

More diverse communities also reduce available nitrates tolower levels (Figure 15.18D). This means that less nitrogen isleached from the soil. Ultimately this would have a positiveeffect on sustaining nutrient cycling and soil fertility.

An important implication of these studies is that reduc-tions in species diversity might have a negative impact onecosystem function. Consequently, we have additional rea-sons for being concerned about the loss of species due tohuman activity. The ecosystems that humans rely on for re-cycling the nutrients that we need may be impaired by theloss of species. ❖

We have just seen how ecosystems function to recycle essen-tial nutrients. Does the ability to recycle nutrients depend onproperties of the community? Many have suggested thatspecies diversity may have multiple effects on ecosystems, in-cluding increased productivity and lower loss of nutrients.

This problem has been examined by Shahid Naeem andcolleagues with replicated artificial communities. A large en-vironmental chamber called the ecotron was used to createreplicate communities with four trophic levels and differentnumbers of species (Figure 15.18A). The high-diversity com-munity had 31 species, while the low-diversity communityhad 9 species. These communities were followed for a total of206 days.

As the plants in each community grew, researchersrecorded the fraction of surface area covered by them. Thelargest changes in percentage of cover were observed in themore diverse community (see Figure 15.18A). This result isnot an unavoidable consequence of having more plantspecies. However, in the more diverse communities, theavailable space was filled more densely than in the low-diver-sity communities.

The net consumption of was used as a measure ofoverall photosynthetic rates. These rates were also higherin diverse communities (Figure 15.18B). Given these re-sults, it is perhaps not surprising that the primary produc-tivity was also greater in the high-diversity communities(Figure 15.18C). One possible explanation for these resultsis that the diverse communities of plants include speciesthat vary in height and leaf shape. This variety may resultin the community more effectively making use of theavailable energy from sunlight.

CO2 Ch

ang

e in

co

ver

16

12

8

4

0

photos

photos

Time (days)1701105020

High diversity

Low diversity

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 470

Ecosystems 471

0.40

0.30

0.20

0.10

Number of species

Nit

rate

in r

oo

tin

g z

on

e (m

g p

er k

g)

2520151050

FIGURE 15.18D Diversity vs. Nitrate Concentration

Lowerphotosynthesis

Higherphotosynthesis

200160120

CO

2 flu

x

Time (days)

Ecotron

FIGURE 15.18B Overall rates of photosynthesis were alsogreater in the more diverse communities.

195151107Time (days)

Lowproductivity

Highproductivity

FIGURE 15.18C The amount of plant material(productivity) was greatest in the more diversecommunities

RoseCh15_0104043_437-474_2p 11/18/04 3:04 PM Page 471


SUMMARY1. Ecological communities are bound to follow the other laws of

science, including the law of thermodynamics.

a. Because energy cannot be created or destroyed, importantinsights about communities can be gained by following thetransfer of energy from one trophic level to the next.

b. This transfer of energy is not very efficient, although somecommunities do better than others.

2. Communities in change are quite common.

a. Some of the earliest ecological work was motivated by theprocess of ecological succession.

b. There is enough regularity in the succession process that ithas been used to date the time of death of decaying humanbodies.

3. The number of species and the number of trophic levels in acommunity depend on many factors.

a. Predators may prevent a superior competitor from elimi-nating less-effective competitors.

b. Occasional environmental disturbances can also have asimilar effect.

c. The numbers of species in a community may also have pos-itive effects on the functioning of the entire ecosystem.

4. The cascade model predicts many features of real food webswhile making few assumptions.

a. This model can also be used in conjunction with the theoryof island biogeography to predict that the average chainlength in food webs will increase with ecosystem size.

b. Lake communities show increases in food-chain lengthwith ecosystem size.

5. Ecosystems perform many functions that are essential for alllife on Earth.

a. The recycling of important nutrients is one of these functions.

b. Atmospheric gases like carbon dioxide are produced byecosystems, and these in turn have effects on global climates.

c. Ecosystem function appears to be improved by increasingspecies diversity.

REVIEW QUESTIONS1. Why does Figure 15.1B support the conclusion that lake com-

munities are top-down regulated?

2. For each of the following terms, explain how energy is lost,making the measure of efficiency less than 100 percent: (i) ex-ploitation efficiency, (ii) assimilation efficiency, (iii) net pro-duction efficiency.

3. What did G. E. Hutchinson find paradoxical about planktoncommunities? How did he explain this paradox?

4. Would you expect the equilibrium number of species on an is-land far from the mainland to be greater, smaller, or equal tothe number on a nearby island? Draw a graph to show howyou reached your conclusion.

5. What is a trophic cascade? Are the results in Figure 15.1B con-sistent with a trophic cascade?

6. In this food-web matrix, indicate which species are primaryproducers, herbivores, and top carnivores:

1 2 3 4

1 0 0 1 0

2 0 0 1 0

3 0 0 0 1

4 0 0 0 0

7. Review the roles of different species of bacteria in the nitrogencycle.

KEY TERMSallogenic successionassimilation efficiencyautogenic successionbiogeochemical cyclesbiomassbiomebottom-up regulationclimax communitycommunitycommunity webdegradative successiondenitrificationdirected link

dynamical stability hypothesisecological efficiencyecological equilibriumecosystemecosystem-size hypothesisessential nutrientsevapotranspirationexploitation efficiencyfluxfood chainsfood webfood-web chainfood-web cycle

food-web lengthfood-web linkfood-web loopgreenhouse effectintermediate disturbance

hypothesiskeystone predatornet production efficiencynitrificationnodenonequilibriumprimary producersprimary succession

productivity hypothesissecondary successionsink websource webspecies diversityspecies richnesssuccessiontop-down regulationtrophic cascadetrophic levelsundirected link

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 472

Further Readings 473

FURTHER READINGSCohen, J. E., 1978. Food Webs and Niche Space. Princeton, NJ:Princeton University Press.

Jenkins, B., R. L. Kitching, and S. L. Pimm. 1992. “Productivity, Dis-turbance and Food Web Structure at a Local Spatial Scale in Experi-mental Container Habitats.” Oikos 65:249–55.

Lawler, S. P., and P. J. Morin. 1993. “Food Web Architecture andPopulation Dynamics in Laboratory Microcosms of Protists.”American Naturalist 141:675–86.

Morin, P. J. 1999. Community Ecology. Malden, MA: BlackwellScience.

Naeem, S., L. J. Thompson, S. P. Lawler, J. H. Lawton, and R. M.Woodfin. 1995. “Empirical Evidence that Declining Species Diversi-ty May Alter the Performance of Terrestrial Ecosystems.”Philosophical Transactions of the Royal Society LondonB 347:249–62.

Paine, R. T. 1966. “Food Web Complexity and Species Diversity.”American Naturalist 100:65–75.

Post, D. M., M. L. Pace, and N. G. Hairston Jr. 2000. “Ecosystem SizeDetermines Food-Chain Length in Lakes.” Nature 405:1047–49.

Power, M. E. 1990. “Effects of Fish in River Food Webs.” Science250:811–14.

Smith, K. G. V. 1986. A Manual of Forensic Entomology. London:Trustees of the British Museum of Natural History.

Sousa, W. P. 1979. “Disturbance in Marine Intertidal Boulder Fields:The Nonequilibrium Maintenance of Species Diversity.” Ecology60:1225–39.

Trunbore, S. E., O. A. Chadwick, and R. Amundson. 1996. “RapidExchange between Soil Carbon and Atmospheric Carbon DioxideDriven by Temperature Change.” Science 272:393–96.

Worthern, W. B. 1989. “Predator-Mediated Coexistence in Labora-tory Communities of Mycophagous Drosophila (Diptera:Drosophilidae).” Ecological Entomology 14:117–26.

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 473

474

RoseCh15_0104043_437-474_2p 11/18/04 2:34 PM Page 474

communities and ecosystems - university of california, irvinedarwin.bio.uci.edu/~mueller/pdf/rose...

Documents