The Effects of Habitat Fragmentation on Fish Developmental Stability

Matthew W. Gosses and Dr. Robert. U. Fischer

Department of Biological Sciences, Eastern Illinois University

INTRODUCTION

What is stream habitat fragmentation?

A lack of connectivity between upstream and downstream populations or sites which occurs when the longitudinal continuum is disrupted or when lateral connections are severed between the stream channel and adjacent wetlands or riparian zones.

How are streams fragmented?

By land-use changes such as deforestation of riparian and floodplain areas, urban development, livestock grazing and conversion of floodplain and riparian areas into cropland for agricultural use resulting in a 70% loss of natural landscape in Illinois

Outcome of habitat fragmentation:

Land use practices can cause stream habitat fragmentation along both a lateral and longitudinal gradient which can lead to the creation of distinct good and poor patches within a stream system.

Within poor patches the possible abiotic changes that occur due to habitat fragmentation include increased sedimentation, increased water temperatures, decreased dissolved oxygen levels and decreased habitat diversity.  These abiotic changes may cause increased habitat stress within poor patches which leads to a disruption in an organism’s development processes or decreased developmental stability.

How is developmental stability measured?

Fluctuating Asymmetry (FA) is the most widely utilized index of developmental stability.  The pattern of bilateral variation in a sample is where the asymmetry values are distributed around a mean of zero.  The further the value is away from zero, the more fluctuating asymmetry there is in an individual organism.

OBJECTIVES

To determine if:

1.  Abiotic factors differ between sites with different habitat quality (good/bad).

2.  Developmental stability (Fluctuating Asymmetry) of fish differ between sites of varying habitat quality.

3.  Fluctuating Asymmetry values could be used as an indicator of stream integrity.

METHODS

Seven sites along Polecat Creek were analyzed for habitat quality using the Stream Habitat Assessment Procedure (SHAP).  SHAP determines habitat quality based on fifteen metrics in the three categories; channel hydrology and morphology, instream habitat and riparian and bank use.  The sites were classified into four good and four bad sites as determined by SHAP.

Methods for abiotic analysis:

Substrate composition was determined by use of an Eckman dredge.  Two samples were taken per site and percent composition was determined for six substrate size classes.

Water temperatures were measured with temperature probes (Hobo Loggers) for a period of four months (June – September) and the mean, minimum and maximum were determined for both habitat classes.

Substrate composition results:

Figure 1.  Substrate composition (* indicates a significant difference between habitat classes).

Temperature results:

Figure 2.  Comparison of temperature ranges a good and bad patches.

Habitat conclusions:  poor patches set up a stressful environment with increased sedimentation and large temperature fluctuations which may lead to the loss of developmental stability.

Methods for determining developmental stability.

Fish sampling technique:

Fish were sampled from a 300-500 ft. length of stream in each site.  Both ends were blocked off to prevent fish from entering or exiting sample area. The area was sampled with a 30 ft. electric seine powered by a single phase, 110 V A.C., 3000 W generator for a minimum of 30 minutes.  Fish (Stonerollers and Striped shiners) were collected with dip nets and placed in buckets for preservation for later analysis.

The Procrustes method of shape analysis uses fish landmarks including the anterior- and posterior-most part of eye, dorsal and ventral tip of the operculum, anterior and posterior insertion points of pectoral and pelvic fins, dorsal fin ray four, anal fin ray three and ray nine of the caudal fin.  In addition to these eleven landmarks, Striped Shiners (Figure 4) were landmarked at the ends of the lateral line.

Procrustes method:

1.  Reflect the landmark configuration of one body side to its mirror image.

2.  Scale the configurations to unit centroid size.

3.  Superimpose the left and right configurations so they have the same centroid size.

4.  Rotate the configurations to achieve optimal fit.

This provides a consensus configuration as the mean coordinates of landmarks, the new set of variables that contain the complete shape information = Fluctuating Asymmetry.

RESULTS

Figure 5.  FA values for Striped shiner and Stoneroller at good and bad patches.

Levens Test indicated that there was no significant difference in shape between sites for both fish.

What is a B.L.U.P.?

Best Linear Unbiased Predictor.  It is used to remove observer error.

Random effect analog to a fixed effect mean.

Each individual value is regressed to the group mean as a function of the distance from the group mean and the variance components of the mixed model effects.

Results (Using BLUPs)

Common Stonerollers: the F-max Test indicated a significant difference in FA among sites (p < 0.05).

The Bonferroni Test illustrated that sites 7 and 2 (both good sites) had significantly lower FA values as compared to the other sites.

Striped Shiner: the F-max Test: indicated a significant difference in FA among sites (p < 0.05).  The Bonferroni Test illustrated that sites 1 (good site) and 5 (bad sites) had significantly higher FA values as compared to the other sites.

CONCLUSIONS

1.  Abiotic factors differed between good and bad sites with bad sites having increased sedimentation and temperature fluctuation and decreased habitat variability.

2.  No significant difference for FA values between good and bad sites for either Striped shiners and common Stonerollers.

3.  Significant differences in FA values were observed for both species among sites. Common Stonerollers from good sites (7, 2) had lower FA values.  Striped shiners showed no trend related to habitat type.

Thus, Fluctuating Asymmetry values showed only moderate success as an indicator of habitat quality in this study.