Goose Point site growth analysis from 20250522 sampling

cgigas
goose-point
hardening
oyster
growth
Goose Point site growth analysis from 20250522 sampling
Author

Ariana Huffmyer

Published

August 18, 2026

Goose Point growth data update

Overview

Noah O. has completed photo analysis for oyster growth at the Goose Point site from the 05/22/2026 assessment. This brings analysis of Goose Point growth up to date with the most recent assessment.

In this project, we conducted thermal and salinity hardening in 2024 and haev outplanted and monitored oysters for the past 2 years. This analyses looks at growth data collected in spring 2026.

Data and code

Data from this assessment are available on GitHub here. The code used to analyse the data is also on GitHub here.

Approach

I analyzed the growth data as I have done in past analyses detailed here.

Noah used our image analysis protocol to collect data from field images.

The analyses is conducted as follows:

  • Length and width are collected from images using the protocol above
  • Volume is calculated from model projections using manual length, width, and depth measurements to predict volume from length and width from photos
  • Volume is analyzed as the metric of growth over time
  • Models are evaluated for assumptions and outliers
  • Significance of hardening effects and time are included in models with post hoc tests performed as needed

tl;dr

There are no strong differences in growth or survival due to hardening.

Results

Here is a summary of the results including the most recent time point.

Survival

There was no difference in survival for either salinity or temperature hardening, although there is a trend for higher mortality in treated oysters in both experiments.

Temperature hardening:

    Kruskal-Wallis rank sum test

data:  total_dead by treatment
Kruskal-Wallis chi-squared = 1.1905, df = 1, p-value = 0.2752

Salinity hardening:

    Kruskal-Wallis rank sum test

data:  total_dead by treatment
Kruskal-Wallis chi-squared = 3.2386, df = 1, p-value = 0.07192

Growth

I used a polynomial model to predict growth from length and width data. The model had a high R2 value of 0.96.

The model showed a strong linear relationship between predicted and actual volumes.

This model was then used to plot growth for each treatment.

Temperature experiment

There was a significant interaction of treatment and time in growth in oysters that went through the temperature hardening experiment.

Type III Analysis of Variance Table with Satterthwaite's method
                 Sum Sq Mean Sq NumDF  DenDF   F value    Pr(>F)    
treatment            43      43     1    4.0    0.0281     0.875    
date           38393614 6398936     6 3690.1 4182.7838 < 2.2e-16 ***
treatment:date   101260   16877     6 3690.1   11.0317 3.194e-12 ***

This was driven by significantly smaller oysters in the temperature treated group in August 2025, but there were no other differences and there is no difference at the most recent time point.

date = 20240624:
 contrast          estimate   SE  df z.ratio p.value
 control - treated    -2.62 7.83 Inf  -0.334  0.7382

date = 20240909:
 contrast          estimate   SE  df z.ratio p.value
 control - treated     3.15 7.82 Inf   0.403  0.6873

date = 20241211:
 contrast          estimate   SE  df z.ratio p.value
 control - treated    -4.16 7.89 Inf  -0.527  0.5984

date = 20250520:
 contrast          estimate   SE  df z.ratio p.value
 control - treated     6.34 7.85 Inf   0.807  0.4194

date = 20250825:
 contrast          estimate   SE  df z.ratio p.value
 control - treated    23.59 8.00 Inf   2.949  0.0032

date = 20251008:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   -14.92 8.03 Inf  -1.857  0.0633

date = 20260522:
 contrast          estimate   SE  df z.ratio p.value
 control - treated    -2.86 8.05 Inf  -0.355  0.7224

Salinity experiment

There was a significant interaction of treatment and time in growth in oysters that went through the salinity hardening experiment.

Type III Analysis of Variance Table with Satterthwaite's method
                 Sum Sq Mean Sq NumDF  DenDF   F value  Pr(>F)    
treatment          6004    6004     1    6.0    3.8519 0.09711 .  
date           44799146 7466524     6 5011.5 4789.9626 < 2e-16 ***
treatment:date   397599   66267     6 5011.5   42.5117 < 2e-16 ***

This was driven by significantly larger oysters in the salinity treated groupfrom May 2025-Oct 2025, but there were no other differences and there is no difference at the most recent time point.

date = 20240624:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   -0.746 4.60 Inf  -0.162  0.8711

date = 20240909:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   -7.058 4.59 Inf  -1.539  0.1238

date = 20241211:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   -4.944 4.68 Inf  -1.057  0.2906

date = 20250520:
 contrast          estimate   SE  df z.ratio p.value
 control - treated  -12.102 4.58 Inf  -2.645  0.0082

date = 20250825:
 contrast          estimate   SE  df z.ratio p.value
 control - treated  -46.072 4.86 Inf  -9.478 <0.0001

date = 20251008:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   27.092 4.89 Inf   5.542 <0.0001

date = 20260522:
 contrast          estimate   SE  df z.ratio p.value
 control - treated   -7.947 4.92 Inf  -1.615  0.1064