Bryan Hummel

                                                                                                            Ecology 3434

                                                                                                            Ribble; 8:30 TR

                                                                                                            February 16, 1999.

 

Bird Habitat Use In Brackenridge Park

 

Introduction:

A study of birds was conducted in Brackenridge Park in order to get a better understanding of the types of habitat used by the different types of birds. It was expected that birds would be found on different trees and in different places according to their preference.  The study showed that one hypothesis was true while the other was false (now doesn’t this get your attention and push you to read further to find which one is true), and its results are discussed below taking into account the limitations of the study.

Methods:

Five different plots were simultaneously surveyed in Brackenridge Park to determine which types of trees were favored by different types of birds according to their size.  The birds were divided by size into three categories: Large (Cardinal size or larger), Medium (Sparrow size), and Small (Warbler or Vireo size).  The size of the bird landing in each of the plots was recorded as well as on what part of which species of tree they landed on.  Each tree was divided into three sections: Trunk, Inner Branches, and Outer Branches.  The data was compiled into two tables and two “null” hypotheses were made in order to compare the findings of the survey with the “null” idea that the birds would show no preference for species or placement. The first table is based on the species of trees and the number of birds landing on those particular trees.  The second graph was comparing the size of birds to the part of the tree that they landed on.

Results:

 

Tree Species

Number

Rel.Density

Observations

Expected

Chi-square

Cedar Elm

92

0.308724832

51

27.785235

19.396104

Escarpment Cherry

7

0.023489933

0

2.114094

2.114094

Ligustrum

75

0.251677852

9

22.651007

8.2270067

Live Oak

12

0.040268456

6

3.6241611

1.5574944

Percimin

3

0.010067114

0

0.9060403

0.9060403

Texas Oak

5

0.016778523

0

1.5100671

1.5100671

Shrub

104

0.348993289

24

31.409396

1.7478575

 

 

 

 

 

 

Totals

298

1

90

90

35.458664

 

            Table1.

The species of tree is listed in column 1 and the relative density, number of bird observations, expected number of bird observations, and chi-square (c2) value were calculated for each species.  The rows of the highlighted boxes should be combined into one row because the expected values are less than three.

 

 

Part Of The Tree

Small

Medium

Large

Total

Trunk

 

 

 

 

Observed

2

4

8

14

Expected

2.5452

2.7573

8.6961

13.9986

Chi-square

0.1167857

0.56007808

0.05572098

0.7325848

Inner Branches

 

 

 

 

Observed

8

5

14

27

Expected

4.90908

5.31817

16.77269

26.99994

Chi-square

1.946146

0.019035147

0.45835282

2.423534

Outer Branches

 

 

 

 

Observed

2

4

19

25

Expected

4.545444

4.924231

15.530267

24.999942

Chi-square

1.425446

0.173469307

0.77519898

2.3741143

 

 

 

 

 

Total Birds

12

13

41

66

Total chi-squares

3.4883778

0.752582535

1.28927278

5.5302331

 

            Table 2.

The trees are divided into three sections: trunk, outer branches, and inner branches.  For each of these three sections, a row was constructed for the number observed, expected number of bird observations, and chi-square (c2) value.  There are three columns according to size: large, medium and small.

 

 

            The results of the two tables were essentially that birds did show a preference for different species of trees, but not necessarily for their placement in those trees.  The null hypotheses were accepted or rejected by comparing the calculated chi-squared values with a table of c2 values in relation to the degrees of freedom.  The null hypothesis can be rejected if the P value (taken off the c2 chart) is less than 0.05. 

            Table 1 compares the numbers of trees by species in the observation area and the number of birds using each species of tree.  The chi-square values for each tree added up to 35.46 and the table had 4 degrees of freedom (# of rows – 1).  When finding a P value using the table of c2 values, it was found that P << .005, thus the birds were not selecting different tree species at random.  Table 2 compares the part of the tree used with the size of the bird using the tree.  The c2 values were totaled for the small, medium, and large birds in each area of the tree.  These totals for trunk, inner branches, and outer branches were then totaled to get an overall total c2 value.  The final value was not large enough to reject the null hypothesis, so it cannot be said that the birds are not choosing the area of the trees at random (c2 = 5.53; P<0.5).

 

Discussion:

The results may at first seem puzzling.  Why would the birds associate with specific species of tree, but not with the parts of that species?  I will try and explain the results from two different approaches.  First, assuming that birds are choosing the species of tree because it contains a resource they like (namely food items) it seems feasible that the birds would show preference for the trees containing the resource, which is what Table 1 concludes.  It would also seem logical at first that the resource (seeds or fruit) would be unevenly distributed throughout the tree and thus the birds would show preference for certain parts of the tree.  Table 2 does not conclude that there is an association, so an alternate logic must be used to support the data.  Lets assume that the resource the birds are using is not produced by the tree, but rather it is also on the tree and it has no association with the parts of the tree (ex. a worm that lives on the bark of the tree).  This would allow the bird access to the resource on any part of the tree, and thus the data would show no definitive association between the birds and different parts of the tree. 

            One also might expect predation to play a role in the areas used by the various birds even though it counters the data.  If predation were prevalent, one would expect the birds to land on the part of the tree that afforded them the most protection (i.e. the inner branches) yet this was not supported by the data.  To support the data, I would conclude that predation is not a major factor on where the birds land in the trees.  The lack of predators, namely hawks and other birds of prey, is one way to explain a low level of predation.  This assumed lack of predators might be caused by the location of the observational area.  Brackenridge Park is located near downtown San Antonio, which is not a very natural place for a hawk to live.  If the number of predators were lowered because of location or proximity to an undesirable predator habitat, one would expect the effect on the birds being studied would be lessened.  If there were so few predators that they did not pose a real threat, then the birds would not select a part of the tree because of protection from the predators.  This logic would also support the data of Table 2.  I am sure that there are several (even hundreds) of alternate explanations for the results that were obtained, but I just outlined two that could possibly explain the results.  A much larger study or the same area over a much longer time would take out a significant amount of possible error and may also change the outcome of the data, but there was not enough space or time allotted for further study.

 

                                               

Bird Habitat Use In Brackenridge Park

 

Bryan Hummel  Ecology 3434

February 16, 1999.

 

  Back to Biology Index