I suggest that the answer depends on the harvested size of the plot relative to the size of the whole plot/treatment. I recently received a paper in which the author had sampled one plant per replicate plot and presented the results in kg/ha. Clearly multiplying up the results of one maize cob to the production of a field in kg/ha is ridiculous and misleading. In the example at the beginning of this discussion, 10m2 was harvested, so this is 1/100 th of a hectare, so reporting results in kg/ha involves 1000 times multiplication - is this realistic? A small error multiplied 1000X becomes very big.. The other issue is border effects. In the example of 10 m2 harvested from a plot 15 m2, depending on the shape of the plot and sample area, there may border effects that could bias the measurement. These are the main reasons why journals request yields in g/m2 rather than kg/ha for small plots. Of course if you are talking to farmers it is a different situation - they want to know the results in t/ha or bushels/acre!

It is true that yield extrapolation from 10 m sp. to kg or t/ha is wrong as the methods is accounting for large scale bias from small plot. Unfortunately, this is what most people do to save time and resources. You can argue certainly argue about the moethod if you want to compare your results to others.

Cheers

Rodrigue Diogo

Keshav sir, Nice to see you and your question in research gate. Yes you can do. I would suggest you to provide yield in Kg/m2 and put t/ha in bracket.

In agronomic experimentation the plot size is usually function of soil variability and the treatment. For example, experiment about genotype, fertilizer and pesticide, small plots can be used (10-50 m2). For experiment about soil tillage and crop rotation, large plots are recommended (50-500 m2). Larger is the plot, and closer is the crop response to field conditions.

In any case, the presentation of crop yield data in kg/ha or t/ha is usually accepted from scientific journals.

Yes, using a elementary unit (1m2 or ha or even km2, depending of the variable) is still the best way to compare different data from different studies. The unit (kg/m2 or t/ha or...) will depend of the journal policy (commonly SI unit, e.g. kg per m2)

It is better to say kg per square meter for reporting yield from small plots.

All data can be extrapolated, indicating of course that some differences should be obtained when you make a real scale up of experments

A tricky but answerable question. Scale matters, it is true, but depends on the question under investigation if extrapolation of some values derived from small experimental plots to higher scale is justifiable. For example, pesticide dose can be expressed from Litre per hectare to ml/ square meter and vice versa whereas carbon sequestration, even if it possible to expressed as t/ha for a specific location, cannot be used (easily and without cross references) to other. Yield can be extrapolated, as someone else already mentioned, from kg per square meter to t/ha but with caution (historical data about the yields of the area can assist and secure the estimation).

It would be better to report g m-2 or g plant-1

If your plot is standard size like those planted in standard breeding trials, then you are ok in reporting in that format because it is a convention. Otherwise, by small plot you mean couple of lines of 1 or 2 meters long, then it would be more appropriate to use units like gr/plot then defining what plot is.

I agree with the difficulty of extrapolate results since kg/plant to kg/ha, but it depends on the kind of trial and the cultive. For instance, in woody crops are well considered experimental plots with 10 trees or plants per unit, and the results can be expressed in terms of kg/ha, and its represent no more than 10-15 lineal meter, but in tomatoe for instance you need more than 100m in row to desing your plots.

Ya, it is true in general, we are accustomed to the manner /method as mentioned in a question itself. The objective of small area experimentation is well known to improve precision as one can do and simultaneously to improve accuracy as much as possible. Representing data as kg / M2 is equivalent to t/ha. Is any difference between these two units. In my opinion kg/ha is most likely to be used for high value and less volume crops, which productivity is in general range below 1000 kg (ton/ha). High value crops like sugarcane,potato , yield may be depicted in terms of t/.ha

In agronomic experiments it is common to convert the plot yield to hectare basis. Usually we have replicates, so it is possible to calculate the variability and to express this as a standard error. I prefer to use SI units, which is this case would be kg/ha or Mg/ha.

You would be better off to express yield in kg/m2 because area that you harvested is small. Extrapolation to larger areas from small experimental plot can over estimate yield. Specify the area of harvest in your methods.

Yes, We Can present values ​​in terms of kg or t / ha, but performed poorly These extrapolations can lead to overestimate production / ha. I mean, it is necessary to consider situations that occur in the field, such as missing plants, diseased plants, unproductive plants and others related to the development of the plant, as soil conditions

Statistical analyzes must be made with the values ​​of the plots!

Data can be extrapolated from small scale experimental plots to hectare basis, provided standard units are used. It is important that there can be huge variations, as one gram error in calculation (at small scale) multiplied and turn to be significant at hectare basis.

Agreed with Silvio Bastidas that statistical analysis must be performed with original values obtained from the plots before converting into hectare basis.

It is correct to express the values ​​in t / ha (considering that one has are 10000 square meters) provided that you explain the methodology

I don't see why not, especially not if both output and surface are precisely measured. There is however potentially an issue when measuring small plots depending on how they are measured. New research from from developing countries seems to suggest that self reported measures of small plots tend to be larger than they actually are (as per GPS measures), while self reported measures of large plots tend to be smaller than they actually are (overestimate size of small plots, underestimate size of large plots). This affects yield measures, leading to steeper yield curves (and thus even more efficient small farmers) than what is generally perceived in developing countries. See recent paper http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2011/12/12/000158349_20111212131701/Rendered/PDF/WPS5908.pdfs

Many of the respondents have righty suggested to keep the yield at kg/10sq.m level. It is convention to present the yield in kg or mt/ha. For high yielding crops like sugarcane, cabbage,and multi-picking crops like brinjal, Okra this may result in over estmiate of yields. Therefore it is important to have a reality check whether the extrapolation results in yields beyond a cultivar's potential yield. In such cases it is better to confine to the original data from a given plot size.

Dear Colleague.

In Science you are allowed to produce data as you like it. You are allowed to do with it what ever you like.......however you have to tell what you did and how you did it.

In your case you harvested the experimental data from your experimental plot. All what you discuss in your investigation are facts which refer to the experimental lay out.

To compare your data with others its fully legitimate to propose or speculate with scale up or down calculations as long as you indicate how the original data were found.

Every day, worldwide, the Wetherforecast is done this way, what does not mean that an ombrella is meaningless after sunshine was predicted.

Good luck, Reto J. Strasser (Geneva, Switzerland)

I agree with Pilar. I don't see any problem with expressing in t/ha, as long as it is clearly explained what you did. Thanks, Gabriele

Yes, of course you can express your results in kg/ha. It is often used to compare results obtained in small areas (experimental condition) in relation to optimized agronomic conditions. That's why we use repetitions in our experiments. It is also common in horticultural crops and grain crops experiments. Furthermore I think that is almost mandatory, otherwise in order to compare your results, when publish in kg/square meters the lector will have to do the conversions to have an idea how significant was your production in relation to real crop production. Just make sure that you are explaining how you convert your data.

In fact, 10 m2 net plot size for final grain harvest is not small. In agronomic field experiments, researcher often layout plots much smaller to the one you have. but, having said that, if you have replicated plots (min 3 reps) then you better present yield in kg/ha (Indian Journals) and t/ha or Mg/ha (International Journals) with their respective std errors.

I agree with Anil Singh. In many of these agronomic field studies, apart from yield (grain crops) or total fresh biomass (sugarcane or vegetables), researchers also look at economic analysis and cost:benefit ratio of production systems. In those cases, it is more meaningful and practical to express gross and net profit at acre or ha basis than plot basis. So, at the end, irrespective of whether you present the data at plot scale (m2) or field scale (acre or ha), the endusers, especially farmers would like to know at Acre or Ha basis. Of course, it is more appropriate to do statistical analysis on data at plot scale.

Presenting in t/ha helps the reader because that is a familiar unit. I would present in this unit and then describe the details in the method + discuss possible extraploation issues. If you want you can of course put the alternative unit in brackets.

If you have multiple plots you could of course do some statistical analysis to estimate the variability for the extrapolation. But I think real-world variability is likely much higher than what you get from experimental field plots. Hence a discussion on this topic is warranted in any case.

Yes, you can extrapolate kg/m2 to ton/ha, just watch out if you numbers are in average with the crop in your region.

I guess you may have to use Mg/ha in a scientific journal, that´s because a Mg (mega gram) is the same number of a ton, but it´s in a International System that, muts be used.

I agree with Matheus and Pilar. This can facilitates comparisons anywhere in the world. good luck, Andre

According to my opinion you should use kg/10sq. m for clarity sake and to be adherent to SI Units. But feel free to use t/ha, if you like it. My (little) experience tells me to always explain clearly to the reader the units of measurement you are using in all text, tables and charts. This is the most important point at all.

Keshav

different journals have different conventions for this. strictly speaking you should express the finding in kg/m2, but some journals will allow you to express it in t/ha as long as you expressedly say this in the materials and methods. that is describe exactly what you have measured and then say the results have been expressed in t/ha equivalents in order to make them comparable with previous work. then you leave it to the editor to decide

Mick

It is allowed but more often when the rsults are converted, they show exaggerated figure.

I think there is any problem with expressing in t/ha.

Viliana

You can

It is better to express your field experimental yields in kg or ton/ha for ease of comparison, though I appreciate your fear of extrapolation error when expressed in economic terms

You should minimise the chances of a reader misinterpreting the results. Suppose someone only spent 15 minutes reading your paper before quoting the results in a summary document of their own, would they quote your results correctly ? Units of kg/ha and tonne/ha are widely understood and minimise the chances of users misinterpreting the results. Most of these values were obtained from small plots like yours.

There is another question inside. Usually production is compared with soil properties. Then, units for production usually is kg/m2, but soil analysis mg/kg or mg/Mg; i. e., you refer units to a little plot.

But better than this is to refer both soil & production parameters to Mg/ha or kg/ha, even for soil (macro- & micro-nutrients respectively); i. e., to refer units to the total system. For that you need compulsorily to know the soil bulk density and some people do not have this soil property. And for that prefer to refer to plot, not to the system that is the better option to compare with other referred ecosystems or agrosystems.

This is my personal experience..., fortunately I can determine very easily soil bulk density, o calculate the production in open forests.

mmm... I'd like to put another idea. Because I've learned to use always SI units. So g cm² or kg m² or Mg m². It should be right. But you need to think about the super or sub estimation.

If you are working with plant breading or sth in this area, there are some "rules" of sample size. Such as a plot w/ some measure or some number of lines or number of plants. And you need to cut the border near the next plot.

What specie are you working? Did you search this issue in a high factor journal? Or Did you ask some editor?

If I were you, I will think in the ideas of all colleagues that right before and maybe these question.

Gizele

Since the output is given for ha, it is better to show the result in terms of t/ha

Better to use the least value i.e. Kg instead of t/ha. As this will not only minimize the error but will be more realistic.

If the data taken on plot basis what plant basis the least value will be a true representative.

There is nothing wrong in expressing the yield data in t/ha, if the experiment is conducted in the field and sampling is done as per standard procedure and with due care for uniformity.The other parameters like dry matter, nutrients etc. may be given as per the norms of journal which you have chosen.

DVS Reddy

i fully agree with D Reddy regarding conversion of output of experimental plots to tons/ha provided due care has been taken while sampling,

I also agree with D. Reddy. It is a must to follow the proper sampling to minimize the errors in the experiment. If not the data become very unrealistic when extrapolation is done.

Keshav Adhikari

It dosen't mater, you can. It is beter given as Kg/ha.

You can either use g/m2 or t/ha and as your yield area is 10m2 the latter would not be a problem

Some journals do not accept units of kg/ha ot t/ha from small plots. What I do is use g/m2 in tables and figures,but in the text put in brackets what this translates to in kg/ha or t/ha for at least the key results.

In science, for all measures that come from small plots, and carried out ​​with scientific precision, in order to minimize the experimental error, their values ​​must be specified in accordance with SI units. Thus, the expression t / ha or Mg / ha is correct.

It depends on your interpretation of data, and its use in testing your hypothesis. 10 m2 is very small area, and one should express in kg/m2 and not in t/ha, as the later will involve tons of errors. so take a wise decision.

Take the data on plot basis. Analyze it. Then take the extrapolated data. Analyze it. The results will be the same. So it's up to you whether you take one or the other.

I think it will not be true, but for recearch it will be more valed

If we multiply raw data by a constant (as we do in extrapolation) the resultant is the same but extrapolated. However, if we change the scale of the raw data (as in log, square root or sine transformation) then the results will be different.

You may consider to express the plot results in g/m2 based on the magnitude of crop yield results. It helps to take care of variation factor. Since you mentioned it to be net plot size, the yield of the total plot need to be accordingly calculated. You may convert the data to kg/ha or t/ha based on the crop yields for the key results and for discussion.

You can use kg or t/ha. It would be better if you use g/m2 in bracket

It all depends on the user of the information. Normally researches are carried out in small plots, and extrapolation gives an idea of what would befallen an expand of land. Therefore, actual plot size should be expressed & extrapolated/relative values in parentheses.

I would like to mention here is there is no problem with units, because here you are comparing different treatments and it is best way to express in Standard SI units rather than plot size units. Since in publication point of view, you/audiance will be comparing this with other crop or other treatments or with locations, hence standard unit is required rather than plot size based units.

In the case you set up experiment in shade-house or glasshouse, the unit is used as g/pot. In this case, you carried out the experiment in the field, it should be expressed in kg or t/ha or Mg/ha (Mg=t).

Hi, if you have got higher yield for example in rice if you got 5000 kg/ha then if go for t/ha is good due to higher values.

if in case of sugar cane the yield levels are so high like 40,000 kg/ha in this case we have to express yield in t/ha is good.

so it is better to go for kg/ha or q/ha. best of luck.

You can use t ha-1, or Mg ha-1, but not kg/plot

Hi, if you have got higher yield for example in rice if you got 5000 kg/ha then if go for t/ha is good due to higher values.

if in case of sugar cane the yield levels are so high like 40,000 kg/ha in this case we have to express yield in t/ha is good.

so it is better to go for kg/ha or q/ha if values are small. it helps you for better interpretation of data. best of luck.

Yes you can express them in t/ha or kg/ha. All you need to do is to find the ratio of your small plot to ha and your the ratio of your unit to tons or kilogram. It would be better understood by many readers in the standard units of ton per ha.

The use of samples to represent a whole involves some issues, as if the plot is representative of the treatment analyzed and assessed on all parcels sampling was performed uniformly. If, overall, this was accomplished, we can only use such data for comparison between treatments, unless that is also used a control treatment in which it is representative of commercial material whose productive potential is well known in the conditions evaluated, and then analyze the plots in relation to that, to an estimated production data per hectare more realistic.

First, I am confused by the units of the plot: 15 m sq would mean 15 m x 15 m (= 225 sq m) whereas 15 sq m is total area. The unit for expressing the results should depend upon the applicability of the study to lrager areas. Small plot experiments have large "edge effects" and extrapolating the data to hectares may yield large errors. On the other hand, when the experiemental plot is large and sampling is adequate (larger number of samples - even if each one is small), the data may be representative of large-scale field situation. One has to think over the point: if one gets an yield of 1000 g/sq m, theoretically one should get 10 tons/ha. If the value is realistic, the experimental data are good and may be expressed in any manner. A farmer is not interested in yield per sq m but per ha. a Forester cannot rely on tree biomass per sq m - even if the sample saize was 5x5 m or 10x10 m.

Also note that the SI Unit for t (ton) is Mg (megagram).

The values should be expressed in a way that a reader can understand . The use of kg/ha or so is fine, but if you give them per plot it is hard to extrapolate your results. Some journals use kg ha-1 or Mg ha-1, but this is just a matter of journal format.

I do not have a problem with the units as long as there are random replications of the various plots according to treatments

I have done it both ways. I used to present the data in terms of a per plot as well as a per hectare basis because the Analysis of Variance and mean comparison will give us the same answer. Make sure that you use the standard plot size of the crop you are working, or even larger plot size.The more important matter is that you must control all external factors (which are not one of your sources of variation) to the best you can to make sure that the yield level and ranking of the treatments are reliable.

If it is a non-tillering crop variety (for eg. Maize/cotton) you may ensure that the plant population is uniform in all the plots and in all the treatments; and then the collected data in grams per plot can be converted to kg per ha or Mg per ha. There will be no difference in the answers of ANOVA you obtain by subjecting the data collected either in grams per plot or kg per ha.

Yes you can, so long as you have taken care of your statistics to perform your research with a high precision and accuracy, e.g. having at least three, but preferably four replications with good plot size. The good plot size depends on your crop, planting style, number of samples you will take during the course of your research, and different parameters you will measure. So, you have to be careful how small you mean by "small plots".

Both measures (Kg/ha or t/ha), are acceptable in standard reporting, however , it is advisable to use the unit that will enhance the readability of your report. Instances where you have a yield of more than three digits, then simply convert to t/ha. in the same version avoid excessive decimal places eg 0.0034 t/ha is not easily readable as such simply present it in kg/ha. But endeavor to use the same unit in a single report.

Yes, You can do that. .Actually is the best way for me. The assumption is that a whole hectare is treated in the same way than your little plot.

You put pot experiment yield data in g/pot only not in kg or t/ ha.

O yes1

this is the way of reporting, all yield -related experiments in the breeding programmes are used such this estimations, estimation of yield per ha through experimental plots.

However, as Mohsen highlighted, you have to carefully care about the statistical consideration in running the experiment. Ok the plot size is one important point, but taking into account of the competing plants is a very important point. Thus, if you are in a beginning work, try to establish a pre-experiment using different number of competing plant to evaluate their yield. These info will help you later in correcting the plot yield and extend the results to the large scales. Also think about co-variance analysis.

No, it is not wrong to express in kg or t/ha to submit in the journals. Problem is as make the extrapolation for carry to hectare, and not overestimate the yield harvest.

It is necesary to consider that a small stripe border of the 10 m sq plot is shared with the neighboring plants. For calculation purposes, the parcel is less than 10 sq m, depending on the distance between plants.

It is necesary to consider too, the number of disease plants, empty sites and the others

yes, but first it should be calculated per plant and then extrapolate to plant population per ha based on spacing provided for the crop. I hope this would satisfy your requirement.

I think it shouldn't be taken that seriously if we talk statistically. In the field we get more spatial variability than small plots where the conditions are more uniform. Of course it is extrapolation and not a depiction of actual yield one would get from a larger (ha or more) area. As I already said, It doesn't matter if we change the scale of measurement from grams per plot to tons per hectare as they come up with the same results as for as an analysis of variance is concerned.

The best way to collimate the plot to field yield (t∙ha^ -1) is to have zero bias in your plotting. Always is possible, register a macroplot yield. If the partial and the total values disagree then the macroplot prevails, the small are proportioned to the big, and statistics as Friedman test for coupled samples may be carried out. Individual variables are valid per-se.

I agree with all contributors to this question but there should be a caution to be observed. If the net area harvested is too small, it would be advisable to present your results on per sq. m basis. Too small plots are those less than 15 sq. m. Many researchers, however, do not observe this in their conversions of per plot yields into kg or t per per ha.

I would add one more comment that I think is important. I wouldn't worry so much about the units per se, UNLESS you will use ton/ha to extrapolate your conclusions to a large scale commercial level without cautioning about the possible caveats mentioned before (plant competition, spatial variability, harvest method, etc.)

If the plant population is uniform, in all the plots harvested, as you have already reduced out the border effect, it no problem to scale-up your data to kg or t/ha. because you are considering your experiment.

Comments about competing plants if talking about weeds, they are so varied in density, frequency as well as floral composition that in a different local/environment there can be little relation; competitiveablity changes drastically.

Yes I agree. It is common to scale-up your experimental data to t/ha, If the plant population is uniform, in all the harvested plots.

However, the optimum size of plot depend on the crop which should be consider in your research.

To follow-up a bit on my previous note and that of Mohammad's, testing "corn" in longer plots have shown increased yield repeatabilities and smaller CVs (possibly caused by a reduction of the "aile effect" where the 1-2 plants closer to the aile have higher yields). The downside is that costs can double if you double the lenght of the plot :(. In any case, my point is that it is not just a matter of the converting units of measure. It is also very important to be aware of other considerations that might highly bias your conclusions.

If you want to be sure, please use better g (or kg) m-2.

But sometimes you have units referred to g m-2, meanwhile others referred to kg ha-1. THIS IS THE TRUE PROBLEM.

Then, I recommend you works in small-plot units as possible.

But if your study is complex and you have other parameters referred to kg ha-1, then convert all them in that unit, but taking in account possible errors inbcluded, as indicated.

That is my poor opinion....

Express your data in any version of Weight per surfave per harvest. What ever is comfortable for you and comparable for others. However add somewhere the originally measured weight harvested on the corresponding surface, given in units as they are indicated in your note book. This is the original unchangable experimental information of your experiment. Good luck and regards, Reto

Thanks for all valuable comments.

Normally research trial in Agronomy and allied sciences carried out in specified design viz., RBD, SPD, FRBD and so on in replicated means in smaller plots of 15 or 20 sqm area and the final yield should be represented in terms of Kg or tons per unit area (acre or hectare) with mean values and SEd and CD values.

This is the recommended practice for any agronomical trial and values should be expressed in kg or ton

In our work where work with international organization we analysis the data after turned to Kg/ha, the lowest plot area 10m2 and the harvested area 1m2

I wonder why Nawal has large plots but harvest area is only 1 m2? We need to be enlightened on that! Usually we need to leave some guard rows of just few plants or at most 1 m width on either side of the plot. The data obtained would very accurately be converted to yield per ha in kg.

Problems that I encounter every field season when extrapolating small plot estimates to field or landscape scales is that small plots yields do not account for in-field variability associated with landscape position that affects hydrology, weed pressure, soil characteristics etc. And, most often, these variables interact with treatment response. Further, a 100 gram discrepancy in DM yield on a 1 m-2 plot results in a 1 Mg difference in per hectare estimates. That being said, I am always tepid when using small plots to calculate farm wide yields as they often tend to over- or under-estimate yields. However, small plot research works very well for the intended purpose of comparing treatment effects. Nonetheless, I always have to report results on a per ha basis during manuscript preparation and presentations and often I do think that the information can be misused and misinterpreted by readers.

My understanding of scaling up units like that is that if you have true replication and randomization, then that accounts for errors and allows you to be able to size up your results.

Yes, these errors are part of your statistical error assumptions. However, if you sit in a combine and watch a yield meter, you will see drastic differences in yields across the same field. I have seen anywhere from 60 to 110 bu/ac soybean in the same 5 acre field. So, if you know this ahead of your study implementation you can block your field and minimize this within field variability in your study design. I suppose the point I am trying to convey is that while you should scale up to bu/ac or Mg/ha to report results of treatment comparisons within your study design, you must use caution in interpreting whether your estimates truly are indicative of the average yield of the farm.

In our experiments with quinoa, with small plots (9 sq.m) we always obtained higher yields than farmers' larger areas. This must be due to more light received per plant and better control of growth of other plants. Acknowledging these reasons we mostly use our data to compare among treatments. Scaling-up the results should always be done with care. Thanks to all for your time and comments.

Your data are only as accurate as your ability to measure. For example if your scale is accurate within 0.1 pound or kilo and your conversion factor is say 5.0 ( based on area of the small plot and the test weight) then your accuracy is only 0.5 bushels/acre or 0.5 kg/ha. Thus reporting the S.E. is critical. Reporting in bushels/acre and kg/ha is something that a farmer can relate and therefore useful. Reporting data in kg/ area and that area is not something a farmer can visualize is providing a disservice. It is important that the reader understand that you are reporting relative yield data. Some researchers advocate reporting relative yield (compared to the highest yielder) and this has merits. When combining data over locations/environments the highest-yielding locations will not carry more weight than a lower-yielding location.

It is normal practice to conduct an agronomy trial, which would consist of experimental units in the dimension you describe or even smaller, and then present those data in units such as kg/ha or Mg/ha. I prefer Mg/ha as the unit is smaller and differences are more apparent. Refer to journals like Agronomy Journal of the Canadian Journal of Plant Science. This all assumes that the data was subjected to a rigorous statistical analysis.

I tend to agree with James Houx, extrapolating may have some setbacks. despite blocking our experiments, the in-field variability in a "real life" situation presents challenges. probably the important thing to do is 'cut and paste' all the management practices from the small plots into the field situation. A small plot may give you 2 ton/ha of cowpea grain but farmers may challenge a researcher that this yield is not realistic in a field situation. but nonetheless, we still need to present our research results from small plots in tons/ha because that is how farmers are going to measure their yield. rigorous statistical subjection is imperative to minimise error though..

Generally as a small plot study we tend to maintain optimum plant population without any field gaps. But when it comes to a hactare area, we will find poor population patches which we may not be considering while we calculate the yield per hactare. This leads to variation in the extrapolated values

Maruthi Vegapareddy

Central Research Institute for Dryland Agriculture, Hyderabad, India

Yes maruthi,that is why spacing, fertilizer rates,weeding, thinning and insect pest management are crucial components that need to be upheld in order to attain good yields such as those in small plots bearing in mind that a hactare is larger and higly variable and may not be so easy to manage compared to smaller plots..

To avoid mis-understanding and mis-interpretations it is important to state in the method section how the yields are calculated from the experiment. Not only as a ratio but in a "historical" way concerning your experiments: e.g. crop yield per pot if you used pots. Or dry mass per pot. Or fresh weight per pot IN YOUR EXPERIMENTAL SET UP. It is fully legitimate to present converted yields as CALCULATIONS like kg or tons per ha. Every student is allowed to scale up his weekly pocket money to kg gold per 100 years. This makes however sense only if many others are using such units as well. Good luck and remain realistic. Reto

I fully agree with Reto that as far as we express clearly in method section, we can express result in per hectare area. However, we should be able to capture as much variability as possible to make the realistic estimate of landscape scale conclusion from plot scale observation

It is always preferred for the yields obtained from plots in the field to be converted to production per hectare in order to be understood including by the professionals working directly in production. In this way, most of the field research results are presented like this, being more easily to process.