NATIONAL UNIVERSITY OF SINGAPORE

Stream to Stream
A statistical analysis of the differences in biodiversity of two freshwater streams in Singapore.

U090023J / Group 3 / Youguang 10/4/2011

Introduction
A stream is a lotic (flowing) system. The diversity of aquatic communities in streams, are dependent not only on its geographical features, but also on its chemical features, such as substrate composition, pH and amount of 1 dissolved oxygen . In Singapore, freshwater habitats are plentiful; they range from harsher ones, such as monsoon canals, to reservoirs and streams. Forest streams in Singapore, are home to about 30 species of local fishes and other flora 2 and fauna . This specific combination of communities is not random, but determined by a number of factors, conditions, processes and activities, from historical to species interactions to human disturbances. All these factors interact in a complex network of relationships to create the observed species diversity, abundance and 3 composition in stream communities . In this study, freshwater ecology sampling was conducted at 2 locations, Bukit Batok Nature Park and Venus Drive. Since both streams differ in characteristics, it would be highly probable that there would be a significant difference in biodiversity between them.

Methods
The two sites chosen were Bukit Batok Nature Park and Venus Drive. These are 2 freshwater streams in Singapore that are easily accessible, with water flow that is known to be slow enough for safe sampling. Bukit Batok Nature Park is in the western part of Singapore, and is close to Bukit Timah Nature Reserve, while Venus Drive is located around the Central Catchment Area. At each location, 4 independent sites were chosen along the stream for sampling. Each site was 20 meters in length, divided equally into 4 transects. At each site, numerous measurements were taken to determine the general characteristics of that particular portion of the stream. To measure pH, temperature, conductivity and dissolved oxygen, a multimeter probe was inserted into the water. As far as possible, the entire probe was submerged into the flowing water. An observation of the bank vegetation and substrate composition was done prior to calculating the canopy cover using a densitometer. Measurement of the wetted depth was done with a ruler, while wetted width was done using the measuring tape. To determine the surface flow rate, a distance of 30cm was marked out and a stopwatch was used to time how long it took a leaf to cover that distance. Sampling of the species in the lake was done via two methods, namely, the surber and the tray net, where each method was used 4 times, once in each site. For the surber samples, each collection was stored separate (in ethanol), alternating the location between the side and the middle of the stream to ensure that each microhabitat was covered. In the tray net collection procedure, all samples were taken from the side of the stream, stored together, and done only once the surber samples were taken. All samples were taken in an upstream manner. All samples were subsequently examined using microscopes to determine and identify the species present. Microinvertebrates were identified using a dichotomous key.

To determine whether the results obtained were significantly different amongst the two methods used, the MannWhitney test statistic was used. Similar Mann-Whitney U tests were also used to check for constancy of species diversity and water quality of the two streams.

Results
Qualitative: *Environmental parameters: Location pH Temperature (C) Dissolved Oxygen (mg/L) Conductivity (microSeimens) Canopy cover (%) Leaf litter (% cover) Wetted width (m) Wetted depth (m) Surface flow rate (m/s) *All Values are the means. BBNP 6.75 28.05 6.17 0.11 67.63 41.64 0.93 0.09 0.1 Venus Dr 7.24 27.06 9.13 0.02 34.53 20.08 1.37 0.08 0.23

The table above shows the comparisons between the environmental parameters of the two streams. Both streams have relatively slow flow rate and wetted depth. This would likely favor smaller species as they would have no need to fight the current flow or deep waters. From the data collected (See Appendix), the number of different species collected at Venus Drive is 25, while BBNP has only 20. However the number of samples collected from BBNP is 114 as compared to 80 from Venus Drive. This suggests that while the biodiversity of the stream at BBNP is more diverse, the stream at Venus Drive seems to support more individuals. Quantitative: Overall richness was calculated using the Mann-Whitney U Test. For the surber samples, the U value was 136.5, giving an approximate z value of 0.32036. The test shows that the samples from the 2 locations for biodiversity are not significantly different (P>0.05, two-tailed test, as shown in table 1) n1 16 n2 16 U 136.5 P (2 tailed) 0.751168 0.748698 P (1 tailed) 0.375584 0.374349

normal approx z = 0.320356 Table 1: Mann-Whitney U Test for Surber Samples

A similar test was done for the tray net samples from both locations. In this case, the U value was 9.5, giving an approximate z value of 0.43301. This test confirms the above result, showing that there is no significant difference. n1 4 n2 4 normal approx z = 0.433013 U 9.5 P (2 tailed) 0.685714 0.665006 P (1 tailed) 0.342857 0.332503

Overall abundance (rank abundance curve)

Overall diversity (Shannon index) A comparison of both locations using the Shannon Index, followed by the Mann-Whitney U test was done, the results, like the above tests, also show that there are no significant differences between the 2 locations for overall diversity. This is likely caused by the irregularities in number of species and types of species recovered from either site.

Stream Quality (SingScore)

For BBNP; SingScore = 82.105. For Venus Drive; SingScore = 96.190 In both streams, the SingScores correspond to fair water quality. A Mann-Whitney U test of the 2 SingScores shows that they are not significantly different.

Discussion
The test statistic scores seem to suggest that both streams are similar in their ecologies, which contradict the initial hypothesis. A key factor that was not considered for this experiment was the effect of human activity on the ecology of the streams. In a built up nation such as Singapore, it is inevitable that freshwater habitats, such as the above mentioned streams would be affected in some way or another. Both streams are located in areas where construction has taken place in the past. Also, the number of local native species in Singapore is relatively small compared to other countries, hence in each individual stream; the differences between these might not be significant, as shown by the data. The 2 methods used for sampling were the surber and the tray net. The surber was used to do a qualitative analysis of the stream, counting each individual species collected, it also gave a more accurate overview of which species were found at which parts of the stream, since each of the 4 samples were transect were stored

separately. The tray net on the other hand, only takes into account the presence of any given species, without concern for its numbers. The advantage of the tray net is that it gives a quick way of estimating diversity of species without the need for statistical analysis. The disadvantage is that there is no way to tell whether a particular species collected is commonly found in the stream. This is covered by the use of the surber, which

1.

How might location of the transect and samples influence the sampling results?

A stream can be divided into 3 sub-sections (using the River Continuum Hypothesis), low order streams, intermediate order streams and high order streams. If the transect was located in a low order stream, it would mean that shredders would dominate the samples, due to the heavy shading and allochthonous input of coarse particulate organic matter (CPOM). In intermediate order streams, there is less shading, and almost all the CPOM has been broken down into fine particulate organic matter (FPOM), meaning that grazers and filter feeders dominate. In high order streams, where there is almost no shading, turbidity of waters inhibit photosynthesis, allowing collectors to dominate. Hence the location of the transect would determine the amount and type of species recovered. 2. How might such water quality parameters influence the macro-invertebrates and fishes obtained?

Each species of macro-invertebrates or fishes have their own niche, and adaptations to their environment. Often, they require specific environmental conditions to survive. Varying the pH, temperature and amount of dissolved oxygen would favour certain species over others. Also, certain species might require differing conditions throughout their life cycles. The water quality parameters from different locations, transects and sections could and should contain a variety of organisms. 3. How might such physical parameters influence the water quality and the macro-invertebrates and fishes?

In the absence of canopy cover, more sunlight is allowed to reach the stream; this would directly increase the rate of photosynthesis, or the number of photosynthesizing producers. This in turn would lead to an increase in the number of subsequent species of higher trophic levels. The vegetation on the banks for the stream would probably draw certain herbivores or insects; this in turn could result in more insect eating fishes, to that particular area of the stream.

4.

What is the rationale for working in a downstream to upstream direction?

If samples were taken in the opposite direction (i.e. upstream to downstream), the use of surber sampling and tray nets would mean that a lot of substrate (such as mud, or soil) and even smaller organisms will be washed down with the current. This will affect subsequent samplings, and give an inaccurate result when comparing locations using the Mann-Whitney U test. In conclusion, the results seem to suggest that there is no significant difference in biodiversity between the Bukit Batok Nature Park and Venus Drive streams, which does not correspond to the null hypothesis that there would be a difference. It is highly likely that due to the limited land area in Singapore, that many of these streams could have shared similar histories, possibly originating from the same hill (eg Bukit Timah Hill), with modern day construction works separating them. It is also likely that while there are differences in the physical and chemical components of each stream, the differences are not significant enough to warrant any vast differences in the biodiversity.

References
1. 2. 3. 4. 5. 6. 7. 8. 9. Cole, Gerald R. Textbook of Limnology. Prospect Heights, IL: Waveland Press, 1994. Lim, K. K. P. and Ng, P. K. L. (1990). A Guide to the Freshwater Fishes of Singapore. Singapore Science Centre. Stream and River Community Structure. Alexander L.C. and Palmer M.A (2002) Ng, H.H. and Ng, P.K.L. (1995). Fishes of the Forest. Nature Watch Vol. 3 No. 2:14-17. Freshwater Biology. Hildrew A.G. and Townsend C.R. (2010)
Angelier, E. 2003. Ecology of Streams and Rivers. Science Publishers, Inc., Enfield. Pp. 215. Allan, J.D. 1995. Stream Ecology: structure and function of running waters. Chapman and Hall, London. Pp. 388. Giller, S. and B. Malmqvist. 1998. The Biology of Streams and Rivers. Oxford University Press, Oxford. Pp. 296.

Hauer, F. Richard, and Gary Anthony Lamberti. Methods in Stream Ecology. San Diego:Academic, 1996.

Appendix