You don't adopt the results as it is but rather integrate it with other results to have a bigger picture of the system.

This is rather philosophical question with enormous practical influence. Of course, the results from these two types of experiments cannot be directly related. However, in laboratory conditions we eliminate a lot of unknown influence, making the results more explainable. As for your considerations, I fully agree to the 2nd one and partially to the first, but not at all to the 3rd, as this can be easely adjusted. In practice, it is advisable to use experiments in controlled conditions for screening of general effects, concentration- or intensity-dependence of some treatments etc. It is well known that if we can see the effect in controlled conditions, it does not mean that the same effect will be seen also on field (usually it is less or absent), but this can give evidence on general relationships in a relatively shor time.

Dr. Raman

Experiments in controlled conditions is carried out to isolate factors that affect some component of plant growth. This is necessary for a better understanding of plant production. Under field conditions, I suggest a plot field where you can apply the fertilizer, for example, that you studied and thus verify the consequences of your conclusions. In my page of RG, the book Stevia rebaudiana: seed production you can see the images I attached long time ago. This procedure is necessary when the researcher is investigating a new crop plant. In soybean and maize, it is not always necessary because the vast literature about the crop can complete the understanding under field condition. Otherwise, the volume of the plug size has significant influence in the flowering time of Stevia rebaudiana, for example. Only one experiment is not enough to decide the field practices about one specific crop plant.  I hope these considerations can help you. Yours Sincerely. Walter.

There is a lot of heterogeneity in the field of the factors that influence crop growth. The essence of experiments in controlled conditions is to afford the researcher the opportunity to better understand the effects of various factors by controlling them. The famous nutrient solutions is a typical example of how data from controlled conditions that can be used to determine the approximate requirements in the field on soils of similar characteristics.

Prof. Aflakpui,

As you mentioned that there is a lot of heterogeneity in the field that influence crop growth and experiments conducted in controlled conditions can help researcher better understand the effects of various factors by controlling them. I agree with you. But farmers grow crops in the field where there is lot of heterogeneity. So, how would you correlate the one done in controlled conditions to the farmers' field.

I believe it is quite difficult to generalise results obtained in controlled conditions to the field and would recommend doing experiments in the field whenever possible. Though it is true that the complexity of natural conditions make it difficult to understand the effect of different factors without controlling them, it is not impossible. If you want to learn about a method to unravel individual contributions of multiple environmental factors in the field, you can take a look at this paper that I published recently: https://elifesciences.org/content/4/e08411/article-metrics

It is almost impossible to mimic the field environment in the controlled condition however extrapolation of the result in the control condition to the field could be also a function the trait(s) that you want to work on. 

Farmers grow crops in the field where there is a lot of heterogeneity. Experiments in controlled conditions give us the basic understanding of how these factors influence crop growth. With this basic understanding we can then apply  the relevant statistical principles of blocking and randomisation to help compute the error mean square that we can use to test for significance or otherwise of factors of interest in our experiments. This way we are able to explain to a greater extent the reasons for the differences we observe in the field with the results of controlled conditions as a guide.

Controlled conditions can be considered as an Academic investigation to detect the effects of limiting factors for instance temperature, light, watering, and others. The results of controlled cabinet do not match precisely the field results, owing to the natural  alterations of limiting factors. Photo period for instance you can never achieve th sun light intensities intermingling with cloud during day and night, Thermoperiodisms, day/night  alteration, Humidity and wind speed interference ... These limiting factors all reflected on cell expansion growth rate and cell differentiations,    

As Gederts Levinsh said, this is also a rather philosophical question. To me, it is a trade-off between getting a mechanistic understanding of the system (a plant, or interactions between plants and their environment) and the external validity (how important are these mechanisms in 'real' field conditions?). The results in a field experiment would be the result of the total amount of interactions, which make it impossible to point out which interaction plays which role. In other words, the whole system is complex, making reductionist explanations (dividing the system into smaller pieces and putting it together from these building blocks) impossible. In the end, both controlled and field experiments are a valuable addition to the advancement of science, but differ in their questions and answers, and it's thus important to choose carefully  which type you use to answer your research questions. 

You don't adopt the results as it is but rather integrate it with other results to have a bigger picture of the system.