You are on page 1of 5

Turning behaviour of Woodlice (Porcellio scaber) 1 Introduction We tested the hypothesis that the direction a woodlouse turns is influenced

by the direction of the previous turn. 2 Materials and methods 2.1 Experimental set up In this experiment a square of clear Perspex, measuring 18 x 18 cm, was engraved in such a way to produce a maze with five different horizontal alleys coming of a vertical central alley as shown in figure one. All of the horizontal alleys were perpendicular to the central one. Therefore in order to turn left or right off the central alley the woodlice would have to turn through 90. In addition to this, each of the alley ways was sufficiently wide, 10 mm, in order to account for slight variations in the size of the woodlice. Small square pieces of Perspex were used to block alleys; the different locations for alley blocks are shown in figure one by the shaded squares. Prior to the experiment the woodlice were contained in a large tank counting some leaf litter and other detritus. The correct number of woodlice needed for each of the three experiments Figure 1: Diagram showing the design of the maze (1, 2a and 2b) were then transferred to a petri dish, (Modified from Nuffield Foundation) with a small amount of water in. The three experiments required a minimum number of 36 woodlice. However, we deliberately exceeded this as a precautionary measure in case any woodlice were unsuitable to be used in any of the experiments. Two paintbrushes were used to transfer woodlice from both the holding tank to the petri dish and subsequently from the petri dish to the maze. 2.2 Experimental procedure 2.2.1 Experiment one In experiment one an alley block was placed in the central alley just above alley P (the diagonally shaded square just above the letter P in figure one) therefore creating a T junction. In this experiment 20 different woodlice were used once each. Firstly, the maze was then set up in order to induce a forced left turn. This was achieved by placing another alley block in alley O, the other diagonally shaded square as shown in figure one. A woodlouse was then removed from the petri dish and placed in position A (figure one) and allowed to move on its own accord through the maze. Following its forced left turn, the choice of turn the animal made at P was recorded.

Then the block in alley O was moved to the alternative location (figure one) thus producing a forced right turn. This time a new woodlouse was placed into the maze at position A (figure one). The start position alternated between A and A a total of twenty times. Therefore in total ten woodlice were tested after performing a forced left turn and another group of ten were tested after a forced right turn. The results for the direction of the second turn were recorded for the twenty woodlice. After each woodlouse was tested it was moved from the maze to another petri dish so as not to mix it up with the other untested organisms. In both parts of this experiment any woodlouse that remained static for over two minutes or turned back on itself at any point was removed from the maze and results discounted. 2.2.2 Experiment 2 This experiment was two-fold and was carried out to determine whether or not woodlice could remember the direction of the forced turn they had made. Both experiment 2a and 2b began with a forced left turn. This meant that, like the forced left turn group in experiment one, the woodlice always started at position A and the diagonally shaded square in alley O was always in place (figure one). Previously unused woodlice were used for both parts of experiment 2. 2.2.2.1 Experiment 2a The first part of the experiment involved varying the distance between performing the forced (left) turn and the choice second turn. Alley blocks were used to create a T junction at P (like in the first experiment), Q, R and S. When the T junctions at Q,R and S were produced, blocks were also placed in both (left and right) entrances to other alleys (see figure one) prior to the T junction to ensure the woodlice performed the choice turn at the desired location. Two woodlice following a forced left turn were given a choice of turn at each of the four distances, the lengths of which are shown in table one. Additionally, the time taken between making the forced turn and the choice turn was also recorded (in seconds). 2.2.2.2 Experiment 2b The second part of the experiment involved altering the time between the forced turn and the choice turn. Once the woodlouse had made the forced left turn at O and entered the vertical alley a single paintbrush was used to restrain the woodlouse for a specified delay period of time. In order for the study to proceed quickly surrogate median times were used for the delay times as the median times to travel to P, Q, R and S from experiment 2a hadnt yet been calculated. The delay times used were; 5, 10, 15 and 20 seconds. Following this delay the woodlouse was then allowed to travel to P where, using an alley block, a T junction had been created; the choice of turn here was then recorded. In this experiment two different woodlice were delayed at O by each of the four different delay times. 2.2.3 Analysis 2.2.3.1 Statistical analysis For experiment one individual groups results were pooled into a class data set to enable statistical analysis. Because the data collected in this experiment was categorical and not continuous we must Location of choice turn P Q R S Distance from O (mm) 38 76 114 152

Table 1: Distance travelled to each of the choice turn locations

use a non-parametric statistical test. Therefore we used the non-parametric test for association, the Chi square test. More specifically, as our results had two categorical variables (direction of forced turn and direction of second turn) we used a two-way contingency table with the Chi square test in order to test the association between these two categorical variables (Fowler et al., 1998). As the contingency table produced was 2 x 2 consequently, in this case the degrees of freedom would be 1. This is because to calculate the degrees of freedom in a contingency table requires the use of the following formula: Degrees of freedom = (number of columns -1) x (number of rows 1) As the contingency table only has 1 degree of freedom Yates correction must be used (Fowler et al., 1998). This correction is applied to prevent the calculated chi square test statistic from being too large in order to make the test more conservative when there are few categories. The correction involves subtracting 0.5 from the numerator prior to squaring for each of the chi square components (Fowler et al., 1998). The subtraction is made from the absolute value of the difference between observed and expected frequencies (O-E); therefore, minus signs are ignored. The new form (including Yates correction) of the Chi square formula is shown in figure two. In order to apply the 2 test the expected frequencies for each of the four observed frequencies must be calculated, the formula for this is shown in figure three in the appendix.

2 = ((|O-E| - 0.5)2) / E
Figure 2: The Yates' correction form of the Chi square formula (Fowler et al., 1998)

2.2.3.2 Non statistical analysis As with experiment one, the data was pooled for experiment 2 (a and b). For both of these experiments data was then used to present graphs displaying the percentage of woodlice turning either the same or opposite direction to the forced (left) turn. In experiment 2a these percentages were plotted for the four distances travelled. Similarly, in experiment 2b the percentage of woodlice turning the same or opposite direction to the forced turn was plotted against the four different delay times. 3 Results 3.1 Experiment one results The class results for experiment one are summarised in table 2 (appendix), these four values represent the observed frequencies. This data was then used to construct the 2-way contingency table, as shown in table 3 below.

Table 3: 2 way contingency table for experiment one

Direction of 2nd turn Left Direction of forced turn Left Right Column totals 43 164 107 Right 157 36 193

Row totals

200 200 400

Table four below shows how the yates corrected chi square statistic was calculated using the observed and expected frequencies.
Table 4: A demonstration of the necessary steps for the calculation of the Yates corrected Chi square value

Direction of second turn Direction of forced turn Left Observed (O) Expected (E) O-E |O-E|-0.5 (O-Ecorrected)2/E Right Observed (O) Expected (E) O-E |O-E|-0.5 (O-Ecorrected)2/E Column totals Forced left 43.0 103.5 -60.5 -60.0 34.8 164.0 103.5 60.5 60.0 34.8 207.0 Forced right 157.0 96.5 60.5 60.0 37.3 36.0 96.5 -60.5 -60.0 37.3 193.0 200.0 Row totals 200.0

Yates correction

Yates correction

144.2
400.0

From this table we can extract the 2 value which is 144.2 (1 d.p).

Bib http://www.nuffieldfoundation.org/practical-biology/investigating-turn-alternation-behaviourwoodlice practical stats for field biology Appendix


Table 2: Class data for experiment one

Direction of forced turn = Left Direction of 2nd turn: Left Right 43 157

Direction of forced turn = Right Direction of 2nd turn: Right Left 164 36

Expected frequencies (E) = (Column total x row total) / grand total


Figure 3: Forumla for the calculation of expected values

You might also like