I.  Introduction

 A.  Documented examples of Evolutionary change
 
  1.  In this course we have seen several examples of evolutionary change.
  2.  Many of these examples have involved extreme selection pressures imposed by
       environmental changes caused by humans and that may not seem "natural"
       (e.g. response to pesticides, to industrial melanism).
  3.  The question may therefore arise as to whether these examples can be extrapolated
       to more natural environmental changes encountered by species living in nature.
  4.  The answer is, of course, yes, and we will, in the last lecture and a half, examine
       several more examples of evolution in a natural setting.
  5.  This is a fitting close for this course, since probably the most important lesson to
       take away from it is that evolution occurs and has been documented repeatedly.
  6.  Moreover, these examples will illustrate that substantial evolutionary change in
       morphology, behavior, and physiology oftern occurs naturally over very short
       periods of time, on the order of tens of years.
  7.  It should thus be clear that the rate of evolution observed in these examples is
       more than sufficient to account for the profound evolutionary changes that have
       occurred on this planet over a time 8 orders of magnitude greater.
 

A.  The Organism
 
1.  Small guppie that lives in rivers in South America and on islands off the continent.
2.  Study was done on fish living in rivers on Trinidad.
B.  Traits examined: several related to size and reproduction
 
1.  Size (length) of mature males and minimum size of gravid females
2.  Reproductive allocation--proportion of body weight allocated to embryos
3.  Embryo size
4.  Number of offspring produced per female per reproductive bout.
C.  Natural history
 
1.  These fish live in rivers.
2.  In the same rivers are several other species of fish that are predators on guppies:
 
a.  Crenicichla alta--predator on guppies; prefers adults
b.  Rivulus hartii--predator on guppies; prefers smaller juveniles
 
3.  Crenicichla and Rivulus seldom occur together.
 
a. In general, Rivulus occurs only in areas where Crenicichla does not occur.
b.  Crenicichla can not swim past waterfalls , whereas Rivulus can, which means that Rivulus
   tends to extend upstream farther.
 
4.  One can thus find two different kinds of habitat:
 
a.  With Rivulus present
b.  With Crenicichla present
 
5.  Guppies are found in both of these types of habitat.
 

D.  Theoretical expectations from "life-history" theory:
 

1.  If adult mortality increases in intensity, a population should evolve to mature earlier--which usually
         means at smaller size, and may often mean that females produce fewer or smaller offspring in a
         given round of breeding, because offspring size/number are limited by overall female size.
2.  On the other hand, if juvenile mortality increases in intensity, this theory predicts that populations
         should evolve to live longer as adults--which may mean that adults grow to bigger size--, that they
         should produce more offspring, and that these offspring should be larger when born so that they can
        grow more rapidly through the stages in which they are vulnerable to predation.
3.  In terms of the Guppies, theses prediction would be that
 
a.  In habitats with Crenicichla, since mortality is primarily imposed by predators on the adult stage,
     guppies should evolve to mature earlier, be smaller, devote proportionally more of their energy
     budget to reproduction, and produce fewer or smaller offspring.
b.  By contrast, in habitats with Rivulus, which impose their mortality primarily on juvenile stages,
     one would expect adult guppies to evolve to be larger, to devote proportionately less energy to
     reproduction,  and to produce more or larger offspring.

A.  The first set of experiments Reznick and Endler did was to survey a series of populations in Trinidad and
   ask whether they showed any of these predicted patterns.
 
1.  First identified a number of river localities that had guppies.
2.  Then they determined which of these had Crenicichla and which had Rivulus.
3.  They then made collections of fish from each site.
4.  They measured:
 
  a. size of mature males (males do not grow much once they reach sexual maturity).
 b. the minimum size at reproduction for females by dissecting females and determining whether
      they had embryos developing.
 c. the weight of the embryos and the weight of the rest of the body in females, from which they
      could determne average embryo weight, number of embryos, and Reproductive Allocation
      (total embryo wt/total body wt incl. embryos).
 
5.  They then compared each of these measures for the two types of site.
 
B.  Results. (See Table and Figure; Click on Table or Figure to view with Adobe Acrobat Reader.)
 

(Table and Figure from Reznick and Endler.  1982.  Evolution 36: 160-177.)

 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 

1.  Mean size of males is 14.9 at the Crenicichla sites, while it is 16.4 at the Rivulus  sites--a 10%
         increase.
2.  Mean size of females is 14.6 at the Crenicichla sites, while it is 17.4 at the Rivulus  sites--a 20%
         increase.
3.  Mean embryo wt is smaller for the Crenicichla site.
4.  Mean reproductive allocation is larger at the Crenicichla site
5.  These results are as predicted by theory, and suggest that differences in the type of selection pressures
         imposed by different predators responsible for  the differences between the two types of populations.

 
1.  Of course, to establish that the differences are evolved differences, one must establish that the
     observed phenotypic differences between the two types of populations are underlain by
     genetic differences.
2.  To address this issue, Reznnick performed a common garden experiment.
 
 a.  Fish were collected from two Crenicichla localities and two Rivulus localities.
 b.  Collected fish were brought back to laboratory and bred.
 c.  Offspring were reared individually until adulthood, then bred again (mated with individuals
       from same populations).
 d.  Second generation of offpsring were reared individually in tanks with known amount of food
       (common garden experiment).
 e.  The complete generation of rearing in the lab was done to minimize any maternal effects.
 f.  Same set of variables was measured on this second generation of lab-reared individuals as was
      measured in field.
           3.  Results ( See Figure and Table).

(Figure and Table from Reznick. 1982.  Evolution 36: 1236-1250.)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 a.  Essentially same results as in the field:   The fish from Crenicichla sites, both males and
       females, were smaller and younger at first reproduction, embryos smaller.
 b.  Thus, the differences seen in the field survey seem to be due at least in part to underlying
        genetic differences between the populations, indicating that those differences are evolved
        differences.

A.  The evidence discussed so far indicates that the observed differences between populations are evolved
      differences that can be explained by differences in the type of predators encountered.

B.  It also allows a prediction: if take guppies from a  Crenicichla site and introduce them into a Rivulus
      site that does not already contain guppies, expect evolutionary change in morphology as guppies
      adapt to new environment.

C.  Experimental test of this prediction:  a transplant experiment (Reznick, Byrga, and Endler).
 

1.  Found a stream in which there was a waterfall.
2.  This waterfall was a barrier to all fish except Rivulus.
3.  This meant that downstream from the waterfall, guppies and Crenicichla had been living together for
         a long time, whereas above the waterfall, only Rivulus was found.
4.  Because the distance separating these two parts of the stream was so short--on the order of a few
         meters--the physical and chemical differences between the two sites are minimal.
5.  Took a large, random sample of guppies from below the waterfall and introduce them above the
         waterfall.  (sample of 200 adults, including multiply-fertilized gravid females)
6.  Let the population reproduce and grow and do its own thing for 11 years (30-60 generations).
7.  At the end of this period, collected a sample of fish from each side of the waterfall.
8.  Brought them back to lab, bred them, reared a generation in common environment, bred them again,
         then reared offspring individually in common garden experiment.
9.  Measured same set of variables as in previous experiments.
10.  Results (see Table).

                    a.  divergence took place in several of the characters
b.  in particular, above the waterfall, males and females were larger at maturity, offspring were
    larger, and number of offspring (brood size) was larger.
 
11.  Thus, this population had evolved many of the same characteristics that they saw in their survey
           populations in which Rivulus were present.

      (S. P. Carroll and  C. Boyd.  1992. Evolution 46: 1052-1069)

 A.  Natural History
 
 

  2.  Seeds are covered by a fruit or seed coat, which the bug penetrates with its proboscis
       to reach the seeds.
 
  3.  Host plant species vary greatly in the thickness of the fruit, and hence in the distance
       from the fruit surface to the seed.

  (All figures in this section from S. P. Carroll and C. Boyd.  1992. Evolution 46:
   1052-1069)
 

   a.  Bugs feeding on thick fruits need long probosces to reach the seeds.
   b.  Buts feeding on thin fruits do not need long probosces; moreover, long
         probosces are thought to be less efficient than shor probosces for feeding
         on thin-walled fruits.
   c.   Thus, from functional considerations, can predict how proboscis length should
         evolve given fruit thickness.
 
  4.  Native host plants
 
   a.  Distributed in south-central U.S. (Texas, Oklahoma, Louisiana) and Florida
   b.  South-central U.S.: Sapindus saponaria; thin-walled
   c.  Florida: Cardiospermum corindum; thick-walled
   d.  Native hosts present in current locations for up to 10,000 yrs.
 

              5.  Introduced hosts

   a.  Several species in plant family Sapindaceae introduced within the last 50-100 yrs.
   b.  Florida: Koelruteria elegans (golden raintree); thin-walled fruits; introduced 30-60
        years ago.
   c.  South-central U.S.: Koelruteria paniculata and Cardiospermum halicacabum;
        thick-walled fruits; introduced 40-100 yrs ago and 20-80 years ago respectively.
   d.  The soapberry bug has colonized these introduced hosts and established populations
         on them.

 B.  Evolutionary divergence
 
    1.  Predictions
 
 

  2.  Proboscis lengths on native hosts
 
   a. proboscis 50% longer in Florida, on thick-walled host species
   b.  Indicates that proboscis length has diverged evolutionarily to match the
        thicknes of host fruits.
   c.  Because host populations have been separated for as much as 10,000 years,
        this evolutionary divergence could have occurred very slowly.
 

 
  3.  Proboscis lengths on introduced hosts.
 
   a.  In South-Central U.S., proboscis lengths are substantially larger in populations on
        introduced hosts than populations on native hosts:
 

   b.  In Florida, proboscis lengths are substantially shorter on introduced hosts than
        on native hosts:
 

 
   c.  These differences are in predicted directions and indicate rapid evolutionary
         divergence (on the order of 50 years).

   d.  Picture of soapberry bug on different hosts.

 C.  Historical records of proboscis lengths.
 
  1.  Because there are good samples of soapberry bugs in entomological collections from
       different regions over the last 80 years, it is possible to determine whether there is
       historical evidence of change in proboscis length.

  2.  Populations in Louisiana (South-central U.S.)
 

   a. Exhibit a significant increase in proboscis length, but not in other size characters,
       when individuals collected before 1965 are compared to individuals collected after
       1965:

   b.  This change in consistent with the greater proboscis lengths on the introduced
         host species in current samples.

  3.  Populations in Florida
 
 

  a.  Exhibit a strong trend of decreased in proboscis length
       over the period 1900-1980.
  b.  This change is again consistent with the observed shorter
        proboscis lengthin populations on introduced hosts in
        current samples.
 
 
 
 
 
 
 
 
 
 

 D.  Summary
 

  1.  The differences in proboscis length among current populations and the pattern of historical
       change in proboscis length in historical collections of the soapberry bug are consistent and
       provide evidence of evolutionary divergence in the direction expected based on functional
       considerations.

  2.  Because the date of introduction of alternate hosts is known, it can be concluded that this
       observed evolutionary divergence occurred very rapidly (e.g. over the course of approximately
       50 years).

 E.  Alternative explanation
 
  1. An alternative hypothesis to account for the difference in proboscis length between native and
      introduced hosts is long-distance colonization from populations with the "appropriate"
      beak lenth.
 
   a.  This hypothesis explains the longer probosces on introduced hosts in the South-
         central U.S. as being due to colonization of these hosts by long-distance dispersers
         from populations on native hosts in Florida, which have long probosces.
   b.  Similar, it explains the shorter probosces on introduced hosts in Florida as being
        due to colonization by long-distance dispersers from populations on native hosts
        in the South-central U.S., which have short probosces.
 
  2.  This hypothesis can be ruled out by showing that populations on native and introduced hosts
       from the same region are genetically more similar to each other than either is to populations
       on the native hosts in the other region.
 
   a.  One indication of such a patter is seen in body coloration.
   b.  All Florida populations, regardless of host, have extensive, deep-red markings
        distributed all over the body.
   c.  By contrast, in all South-central U.S. populations, the markings are not as
        intense, are more orange, and do not extend over seas much of the body.
   d.  Because these color differences are almost certainly genetic, this pattern
         indicates that populations on introduced species are derived from geographically
         proxomal popultations, rather than by long-distance dispersal.

 A.  Natural history
 
       
 
  1.  There has been a great evolutionary radiation of lizzards in the genus Anolis on the
       Carribbean islands.

  2.  Going along with this radiation has been extensive diversification of limb morphology.

  3.  The functional basis of this diversification is understood fairly well:
 

   a.  These lizzards spend much time perching and running on vegetation.

     i.  increased limb size generally increases running speed, but
     ii. small limbs allow more efficient maneuvering on small-diameter
         limbs
 (Figure from Losos, Irschick and Schoener.  1994.  Evolution 48: 1786-1798)

  4.  This inference can be tested by determining whether limb size evolves in the expected
       direction when a species is introduced into an environment with a drastically different
       type of vegetation from that in which it normally lives.
 
 B.  Experimental introduction of lizzards onto islands.
 
  1.  Losos and Schoener performed such an experiment by capturing lizzards from the
       island of Staniel Cay  in the Bahamas and placing some of them on each of 14 nearby
       small islands that did not at the time have lizzards.

  2.  Staniel Cay is covered with large areas of moderate to tall forests, which provide limbs
       of large diameter for lizzards.

  3.  By contrast, the vegetation on the small islands onto which the lizzards were introduced
       are covered with plants that are much shorter and have stems and branches that are
       much narrower.

  4.  The prediction, then, is that the populations on the small islands should evolve smaller
       limbs.

  5.  To test this prediction, Losos et. al. returned to the islands 14 years later and collected and
       measured lizards on both Staniel Cay (from the area the original lizards were collected) and
       the islands to wich lizards had been introduced.
 

 C.  Results
 
  1.  Morphological divergence was analyzed using a statistical procedure that is too complicated
       to explain here.  In essence, however, the following plot shows the relationship between a
       variable that is determined primarily by limb length on the x-axis, and a variable determined
       primarily by body mass on the y-axis:
 
  (Remaining figures from Losos, Warheit and Schoener.  1997.  Nature 387: 70-74)
 
 
 
 
 
 
 
 
 
 
 
 
  2.  On 11 islands, limb length was significantly smaller than the lizards on
       Staniel Cay.  On the other three islands, limbs were approximately the same size as on
       Staniel Cay.

  3.  Moreover, there is a strong negative correlation between the degree to which limb length
       decreased on an island, compared to Staniel Cay, and the height of vegetation on an
       island:
 


 
  4.  This correlation indicates that the shorter the vegetation, and hence the smaller the diameter
        of stems and limbs in the vegetation, the greater the decrease in limb size.

  5.  Finally, when all islands are compared, there is a strong correlation between mean
       limb length and the average size of perch used, indicating that on most of the islands,
       the lizards that had evolved smaller limbs were using smaller perches, presumably because
       available perches were smaller than on Staniel Cay.
 

  2.  Because introductions were replicated, and most showed change in the same direction,
       drift can not explain the changes (if drift were responsible, half should increase and
       half should decrease in limb length).  Rather, natural selection must have been the cause.

  3.  Understanding of the functional relationship between limb size, branch diameter, and
       running speed and efficiency suggests that natural selection favored reduced limb
       size in the introduced populations because individuals with smaller limbs were more
       maneuverable on the smaller-diameter vegetation and thus were more likely to escape

 General Conclusions
 

 A.  All three examples demonstrate substantial evolutionary change in
       morphology occurring over a period of just one to several decades.

 B.  In all three cases evolutionary change was clearly due to natural selection.

 C.  Moreover, in all three cases, knowledge of the natural history of the organism and of the
       functional significance of the trait(s) permitted the direction of evolutionary change to be
       PREDICTED either before it took place (guppies and Anolis) or before it was measured
       (soapberry bug)--a sure sign that in all three cases the processes responsible for change
       are well understood.

 D.  There can thus be no doubt that
 

  1.  Evolutionary change occurs by natural selection under natural conditions, or that

  2.  Evolutionary change can be sufficiently rapid within populations to account for the amount of
       morphological, physiological, and behavioral changes that have occured in producing
       extant organisms from their prokaryotic ancestors that lived some 3.5 billion years ago.