This experiment was set up at Egerton University, Njoro campus, about 20 km from Nakuru town in Kenya. The University main campus lies at coordinates 0˚22′11.0ʺS and 35˚55′58.0ʺE (Longitude: 35.932779; Latitude: 0.369734). The University is 2,238 meters above sea level (7,324 feet). This makes it a high-altitude area. It is in ecological 3 and the environmental data obtained from Kenya Agricultural and Livestock Research Organization Njoro Centre indicates the temperatures have been very low for the last ten years.

The ponds were housed within a polythene covered pond (PCP) and in the open field ponds (OFP). Polythene covered ponds were being tried as a way to grow fish in a controlled environment while also benefiting from the natural light and warmth. Eight (four polythene covered and four open field ponds) identical concrete ponds per species, measuring 12 × 9 m were established for Nile and blue tilapia. Two brooders and one sire of a specific species were placed in a hapa of 1 m³ for three months. The PCP had eight hapas and OFP had eight hapas. This made a total of eight brooders of aureus in the PCP and eight brooders of niloticus in PCP. This was repeated in the OFP. Every hapa had a sire of the same species both in PCP and OFP as shown below in Figure 1 .

The linear mixed model below was used to study the fecundity

Data analysis:

$\begin{array}{c}Y=\alpha +\left\{{\beta }_1\left(\Delta pH\right)\right\}+\left\{{\gamma }_2\left(\Delta Cond\right)\right\}+\left\{{\delta }_3\left(\Delta Temp\right)\right\}+\left\{{\epsilon }_4\left(\Delta DO\right)\right\}\\ +\left\{{\zeta }_5\left(AP\right)\right\}+\left\{{\eta }_6\left(WP\right)\right\}+\left\{{\theta }_7\left(MT\right)\right\}+\left\{{\iota }_8\left(TR\right)\right\}+\left\{{\kappa }_9\left(Cover\right)\right\}\\ +\left\{{\lambda }_{10}\left(Species\right)\right\}+\left\{{\mu }_{11}\left(\Delta pH\left(Cover\right)\right)\right\}+\left\{{\nu }_{12}\left(\Delta pH\left(Species\right)\right)\right\}\\ +\left\{{\xi }_{13}\left(\Delta Cond\left(Cover\right)\right)\right\}+\left\{o_{14}\left(\Delta Cond\left(Species\right)\right)\right\}\\ +\left\{{\pi }_{15}\left(\Delta Temp\left(Cover\right)\right)\right\}+\left\{{\rho }_{16}\left(\Delta Temp\left(Species\right)\right)\right\}\\ +\left\{{\sigma }_{17}\left(\Delta DO\left(Cover\right)\right)\right\}+\left\{{\tau }_{18}\left(\Delta DO\left(Species\right)\right)\right\}\\ +\left\{{\upsilon }_{19}\left(AP\left(Cover\right)\right)\right\}+\left\{{\varphi }_{20}\left(AP\left(Species\right)\right)\right\}\\ +\left\{{\chi }_{21}\left(WP\left(Cover\right)\right)\right\}+\left\{{\psi }_{22}\left(WP\left(Species\right)\right)\right\}\\ +\left\{{\omega }_{23}\left(MT\left(Cover\right)\right)\right\}+\left\{{\vartheta }_{24}\left(MT\left(Species\right)\right)\right\}\\ +\left\{{\aleph }_{25}\left(TR\left(Cover\right)\right)\right\}+\left\{{\Upsilon }_{26}\left(TR\left(Species\right)\right)\right\}\end{array}$

where: $Y$ is the dependent variable, $\alpha$ is the intercept (baseline fecundity), ${\beta }_1\left(\text{Δ}pH\right)$ $\Delta pH$ is the effect of the change in pH, ${\gamma }_2\left(\text{Δ}Cond\right)$ is the effect of the change in conductivity, ${\delta }_3\left(\text{Δ}Temp\right)$ is the effect of the change in temperature, ${\epsilon }_4\left(\text{Δ}DO\right)$ is the effect of the change in dissolved oxygen, ${\zeta }_5\left(AP\right)$ is the effect of atmospheric pressure, ${\eta }_6\left(WP\right)$ is the effect of pond water pressure, ${\theta }_7\left(MT\right)$ is the effect of midnight temperature, ${\iota }_8\left(TR\right)$ is the effect of temperature range, ${\kappa }_9\left(Cover\right)$ is the effect of pond cover type, ${\lambda }_{10}\left(Species\right)$ is the effect of fish species, ${\mu }_{11}\left(\text{Δ}pH\left(Cover\right)\right)$ is the interaction between change in pH and cover type, ${\upsilon }_{12}\left(\text{Δ}pH\left(Species\right)\right)$ is the interaction between change in pH and species, ${\xi }_{13}\left(\text{Δ}Cond\left(Cover\right)\right)$ is the interaction between change in conductivity and cover type, $o_{14}\left(\text{Δ}Cond\left(Species\right)\right)$ is the interaction between change in conductivity and species, ${\pi }_{15}\left(\text{Δ}Temp\left(Cover\right)\right)$ is the interaction between change in temperature and cover type, ${\rho }_{16}\left(\text{Δ}Temp\left(Species\right)\right)$ is the interaction between change in temperature and species, ${\sigma }_{17}\left(\text{Δ}DO\left(Cover\right)\right)$ is the interaction between change in dissolved oxygen and cover type, ${\tau }_{18}\left(\text{Δ}DO\left(Species\right)\right)$ is the interaction between change in dissolved oxygen and species, ${\upsilon }_{19}\left(AP\left(Cover\right)\right)$ is the interaction between atmospheric pressure and cover type, ${\varphi }_{20}\left(AP\left(Species\right)\right)$ is the interaction between atmospheric pressure and species, ${\chi }_{21}\left(WP\left(Cover\right)\right)$ is the interaction between pond water pressure and cover type, ${\text{Ψ}}_{22}\left(WP\left(Species\right)\right)$ is the interaction between pond water pressure and species, ${\omega }_{23}\left(MT\left(Cover\right)\right)$ is the interaction between midnight temperature and cover type, ${\vartheta }_{24}\left(MT\left(Species\right)\right)$ is the interaction between midnight temperature and species, ${\varkappa }_{25}\left(TR\left(Cover\right)\right)$ is the interaction between temperature range and cover type, ${\text{Υ}}_{26}\left(TR\left(Species\right)\right)$ is the interaction between temperature range and species, $\epsilon$ is the error term.

Multiple Linear Regression Model was used

$\begin{array}{c}Y={\beta }_0+\left\{{\beta }_1\left(\text{Δ}pH\right)\right\}+\left\{{\beta }_2\left(\text{Δ}Temp\right)\right\}+\left\{{\beta }_3\left(\text{Δ}Cond\right)\right\}+\left\{{\beta }_4\left(\text{Δ}DO\right)\right\}\\ +\left\{{\beta }_5\left(AP\right)\right\}+\left\{{\beta }_6\left(WP\right)\right\}+\left\{{\beta }_7\left(MT\right)\right\}+\left\{{\beta }_8\left(TR\right)\right\}\\ +\left\{{\beta }_{9}Cover\right\}+\left\{{\beta }_{10}Species\right\}+\epsilon \end{array}$

where:

Y is the dependent variable, representing the reproductive potential or egg production of the fish. It's what we are trying to predict or understand, ${\beta }_0$ is the intercept. It represents the baseline fecundity when all other variables are zero or at their reference levels. It’s a constant in the model, ${\beta }_1\left(\text{Δ}pH\right)$ is the effect of the change in pH (Morning pH - Evening pH) on fecundity. A positive β₁ means that a larger difference between morning and evening pH is associated with higher fecundity, while a negative β₁ suggests the opposite, ${\beta }_2\left(\text{Δ}Temp\right)$ is the effect of the change in conductivity (Morning Conductivity - Evening Conductivity) on fecundity. A positive β₂ means that a larger difference between morning and evening conductivity is associated with higher fecundity, while a negative β₂ suggests the opposite, ${\beta }_3\left(\text{Δ}Cond\right)$ is the effect of the change in temperature (Pond Temperature Morning - Pond Temperature Evening) on fecundity. A positive β₃ means that a larger difference between morning and evening pond temperature is associated with higher fecundity, and vice-versa, ${\beta }_4\left(\text{Δ}DO\right)$ is the effect of the change in dissolved oxygen (Dissolved Oxygen Morning - Dissolved Oxygen Evening) on fecundity. A positive β₄ means that a larger difference between morning and evening dissolved oxygen is associated with higher fecundity, and vice-versa, ${\beta }_5\left(AP\right)$ is the effect of atmospheric pressure on fecundity, ${\beta }_6\left(WP\right)$ is the effect of pond water pressure on fecundity, ${\beta }_7\left(MT\right)$ is the effect of midnight temperature on fecundity, ${\beta }_8\left(TR\right)$ is the effect of temperature range on fecundity, ${\beta }_{9}Cover$ is the pond cover type (Open Pond or Polythene Cover). One category would be the reference (e.g., Open Pond). β₉ would then represent the difference in fecundity between the Polythene Cover and the Open Pond, ${\beta }_{10}Species$ is the fish species (Aureus or Niloticus). One species would be the reference. β₁₀ would then represent the difference in fecundity between the other species and the reference species and $\epsilon$ is the error term. It represents the unexplained variation in fecundity, the part not accounted for by the other variables in the model [14] . It captures random fluctuations and other factors not included in the model [15] .