USING GEOMETRIC MORPHOMETRICS TO TEST

FOR FISHING-INDUCED SELECTION IN PHILIPPINE

RABBIT FISHDepartment of Life Sciences, Texas A&M University – Corpus Christi

Micah Bachner, Abner Bucol, Jason Selwyn, Christopher E. Bird

IntroductionPopulations experience a variable landscape of

selective pressures which act on phenotypic variation,

resulting in geographic variation in phenotypes.

Selective pressures applied by fishing activities

accelerate the pace of evolution and can affect

geographic variation in morphology. In the Philippines,

91% of locally-caught fishes are consumed locally, and

an estimated 56% of protein consumption comes from

fishes. In this region, Siganus fuscescens is a staple

food item that is eaten fresh, dried, or in the case of

juveniles, turned into a delicacy fish paste and whose

fishery is showing signs of collapse. To better

understand the effects of overfishing on the evolution

S. fuscenscens, here we test for differences in body

shape among four locations in Negros Oriental,

Philippines, each experiencing different fishing

regimes.

Results• Shape differed significantly among locations

-PC2 separates sites by size of operculum,

position of anal fin (DU-AY vs BA-AM)

-PC1 separates BA-AM by body depth

• 73% of variation in fishing pressure indices

explained by PC1

• Shape related to PC1 of fishing pressure indices

-Most fishermen in BA: deeper body, smaller

operculum

-Most people in DU: shallower, elongate body

Source df SS MS Rsq F Z P

Location 3 0.015 5.1E-3 0.14 9.0 9.1 1.0E-04

Resids 165 0.094 5.7E-4 0.86

Total 168 0.11

Table 1. Values from MANOVA test for differences in shape among locations.

Figure 2. Example photograph of S. fuscescens used for geometric morphometric analysis with landmarks (red points).

AcknowledgementsMarhuma Zaman, John Whalen, Dr. Rene

Abesames, Dr. Alcala, Dr. Kent Carpenter,

SUAKREM

Discussion• Significant differences in fish shape strongly

indicate spatial variation selective pressures.

• Population size and fisherman inversely related

perhaps due to more people = more subsistence;

fewer people = more commercial

• Phenotypes in area of highest commercial fishing

pressure may reduce chances of capture via net

-deeper body, smaller operculum

Figure 1. Map of sampling sites with accentuated municipality boundaries.

Inset table contains data on the number of people and fishers per hectare of

habitat.

Mean fish shape

Fish shape for max PC1

AMAYBADU

Methodsi. Specimens collected (n=166) from 4 fish markets

along the eastern coast of Negros Oriental (Fig. 1)

ii. All individuals photographed on left side

iii. 24 landmarks identified for each fish and stored in

TPS format

iv. A PCA of several indices of fishing pressure (pop

size, # fishers, available habitat, and distance from

pop center was used to isolate PC1 for comparison

with shape data

v. Geometric morphometrics using geomorph (Adams

and Otárola-Castillo 2013) R package: a) Procrustes

superimposition, b) MANOVA correction for allometry

and sex, c) MANOVA test for effect of location and

PCA to visualize, d) multivariate regression test for

effect of fishing pressure

Figure 4. A 1-D representation of fish shape regressed against fishing

pressure index. Regression scores were calculated from the multivariate

regression of landmarks against fishing pressure (Table 2, below). Bars

represent std. error and the shaded region is the 95% CI of the model.

Wireframes show the predicted shapes at the observed extremes of fishing

pressure. Here, sites are treated as independent observations, but fish are not.

Consequently, this test is conservative relative to that reported in Table 2.

Negros Island

Ayungon (AY); n=24

Bais (BA); n=87

Amlan (AM); n=31

Dumaguete (DU); n=24 30 km

12

3°E

Source df SS MS F P

Fishing

Pressure1 0.011 0.011 18.1