Learn from trials for optimum performance
Let’s say you are happy with your current hatchery results. But what if there is still room for improvement and you are not aware of it? How will you know if your hatchery can do better?
By Guy Whetherly, Incubation Researcher, Hatchery Development Department
Using comparative trials as a measure of performance is a simple way to indicate how certain changes to your incubation profile or hatchery practices can affect your hatchery results. In essence, you use one group of eggs as a standard and a similar group of eggs as a trial. By changing one hatchery parameter at a time, you can adapt your process to produce better results or check whether your current results are already the best they can be. Let’s dig a little deeper and see how we set up these trials.
Clear and reachable goals
Before you start any trial, you first need clear objectives. Without well-defined objectives, setting up and executing any trial could result in waste of both time and resources. To make sure your goals are clear and reachable, each one should be SMART: Specific, Measurable, Achievable, Realistic and Time bound. Setting SMART goals will make it easier for you to evaluate the trial data.
Let’s see how this works for a trial on layer breeder culls. The main goal of this trial was to reduce the number of culls produced in a layer breeder hatchery due to the vulnerability of layers to excessive or insufficient cooling. This vulnerability occurs before and during the phase in the incubation cycle where the yolk sac and vascular system retract (between internal and external pipping). For this trial, we defined the following SMART goals:
- Specific: reduce the number of culls (mainly due to button navels)
- Measurable: collect accurate hatch debris analysis including button navel data
- Achievable: try to reduce the cull level gradually and consistently
- Realistic: set a realistic cull goal as culls are always high from layer breeder eggs
- Time bound: achieve the cull target within 6 months
Setting up these goals made it easy to compare and follow up. We achieved the reduction of the number of layer breeder culls by changing the hatcher transfer temperature so it matched the end temperatures of the setter. This resulted in less button navel and string navel culls. All the SMART criteria were met during this trial.
Ideal conditions to identify trends
When conducting comparative trials, repetition, sample size and identical conditions of the samples are essential elements. Repetition of representative trials under the same conditions is the only way to identify trends correctly.
Unlike scientific research, the aim of commercial trials is to repeat patterns in order to spot trends. If necessary, these trends can then be further investigated at a scientific level.
Sample size in commercial incubators is defined by tray size as a minimum. However, we prefer to define the sample size by machine size so thousands of eggs are measured as a sample for both trial and control.
All conditions for the eggs, such as source, flock age, breed, storage conditions, fumigation, pre-heating etc. have to be equal. Only the trial parameter that you are testing should be different in each sample.
Data collection for a clear overview
Gather as much data as possible and look for unexpected effects and trends. From hatchery measurements such as breakout analysis, machine history files, HOF and HOS data along with chick weights, chick quality and hatch debris analysis, you can identify any positive or negative trends. With this information, you can change programs or settings to optimise your results.
* HOF: hatch of fertile - the ratio of hatched chicks to the number of fertile eggs
* HOS: hatch of set - the ratio of hatched chicks to the number of eggs in the incubator
Empirical evidence is information acquired by observation or experimentation. It requires accuracy and integrity of the data so the research is considered valid and unbiased. When executing comparative trials, you can achieve 2 types of empirical data: quantitative and qualitative.
Quantitative data is numerical data such as percentage HOF or HOS. Qualitative data involves using the human senses, such as an assessment of chick quality. Both types of data are valid, but in the case of chick quality assessment, we recommend you assign this task to the same personnel every trial to get consistent results.
Interpretation is key
When analysing your data, make sure you interpret them correctly. Raw data can be misleading if you read them out of their context. For example: normal venting of humidity and the introduction of oxygen to the environment during the later stages of incubation will result in a distortion of the normally homogenous temperature profile within the cabinet. If you are just looking at the data logging results without understanding the reason for this effect, you will get to a misinformed conclusion.
- Keep in mind what you’d like to improve and formulate it in a SMART way
- Make sure all conditions stay the same except the one parameter you are testing
- Look for trends in your data
- Always look at your data in the right context
When you keep these basics in mind, you will have no trouble setting up a small trial to improve your hatching results.