How to do a DIY ex­per­i­ment on your block

NZ Lifestyle Block - - Feature -

to mea­sure. It’s the ones that get turned into profit such as the stuff you har­vest: lambs, grapes, ap­ples, let­tuces, wheat grains. Dis­ap­point­ingly, this is of­ten the mea­sure­ment that gets missed by sci­en­tists.

How­ever, it is also im­por­tant to mea­sure ‘in­ter­me­di­ate’ pa­ram­e­ters, such as growth dur­ing the whole sea­son, plant nu­tri­ent lev­els etc, as these are im­por­tant for help­ing you to un­der­stand what is go­ing on.

There is a mantra in sci­ence that ‘cor­re­la­tion does not im­ply cau­sa­tion’. That is, if you only mea­sure yield, you don't know why the yield in­creased so you only have a cor­re­la­tion which is weak sci­ence. If you mea­sure other pa­ram­e­ters, these can point to how the in­crease was caused, giv­ing you stronger sci­ence.

What time frame do they cover?

For prod­ucts such as bios­tim­u­lants that have an im­me­di­ate and rel­a­tively short term ef­fect, trial du­ra­tion is typ­i­cally one crop cy­cle, based on the as­sump­tion that there is lit­tle or no resid­ual ef­fect; if you stop us­ing the prod­uct, then the ef­fect stops af­ter a week to a few months.

How­ever, it is rare for ef­fects to be truly short term so, if re­sources al­low, the ex­per­i­ment should be run for three to five years to see what the long-term ef­fects are.

For prod­ucts such as biofer­tilis­ers or any­thing that im­pacts on soil pro­cesses, du­ra­tion should be as long as pos­si­ble be­cause soil pro­cesses and per­for­mance change very slowly. It re­ally can take decades for the long-term ef­fects to be fully shown. Truly long-term soil ex­per­i­ments around the world have now been run­ning for over a cen­tury and data from these shows that it takes up to 50 years for soil to truly reach a new equi­lib­rium. When the first 10-30 years data from these ex­per­i­ments are an­a­lysed they of­ten give quite dif­fer­ent re­sults com­pared with anal­y­sis af­ter 50 years. If sci­en­tists are be­ing re­ally hard core about such tri­als, they will throw out the data from the first five years, have a look to see if there are any trends in the next five years, and then con­sider data af­ter the first decade as start­ing to be­come re­li­able. If you are run­ning or look­ing at data from ex­per­i­ments that af­fect the soil, a trial should re­ally be kept run­ning for five years at a min­i­mum, but ideally a decade.

How to do your own ex­per­i­ments

Agri­cul­tural ex­per­i­ments are among the sim­plest, and the value of DIY ex­per­i­ments is that they are done on your crop or pas­ture so the re­sults are 100% mean­ing­ful for your op­er­a­tion. All you need to do is fol­low a few sim­ple rules.

Limit what you’re go­ing to test

Treat­ments are the dif­fer­ent prod­ucts you want to test. More is not al­ways mer­rier as the amount of work in­creases con­sid­er­ably.

De­cide on a regime

It is im­por­tant to de­cide the ap­pli­ca­tion regime from the start: is the prod­uct to be ap­plied once at the start of the trial, or sprayed on weekly? The ap­pli­ca­tion regime should match what would be done in the real crop.

Have a ‘con­trol’

You need to have some­thing to com­pare, a con­trol area, where noth­ing is ap­plied to the crop, and/or you use your cur­rent prac­tice, eg your cur­rent fer­tilis­ers. The con­trol needs to be repli­cated and ran­domised just the same as the treat­ments.

De­cide on a du­ra­tion

Get­ting the ex­per­i­men­tal du­ra­tion right is re­ally im­por­tant. For bios­tim­u­lants that is typ­i­cally one crop cy­cle but ideally three or more, while for biofer­tilis­ers the du­ra­tion should be as long as pos­si­ble, ideally five years but a decade is much bet­ter.

Re­peat, re­peat, re­peat

Like the farmer spray­ing sev­eral strips of sea­weed fer­tiliser on his peas from page 24, you need to have sev­eral ap­pli­ca­tions of the treat­ments. Tra­di­tion­ally the min­i­mum is four repli­cates but in a per­fect world six to eight is best.

Be as ran­dom as you can

It is im­pos­si­ble to em­pha­sise how im­por­tant proper ran­domi­sa­tion is. The pea farmer sprayed al­ter­nat­ing strips up his field, but it would have been bet­ter if he’d flipped a coin at the end of each row – head for spray, tails for a con­trol – and kept go­ing un­til he had enough repli­cates (spray strips) of the sea­weed

and un­sprayed (con­trol). Ran­domi­sa­tion helps take chance out of the ex­per­i­ment, so you know you didn’t ac­ci­den­tally add all the treat­ment you are test­ing on an area that by chance had higher or lower fer­til­ity any­way. The stan­dard lay­out for field tri­als is the ran­domised com­plete block (RCB). Fig­ure 1 shows a RCB ex­per­i­ment lay­out with four treat­ments (a, b, c, d) and four repli­cates. The key to block­ing is that each of the four treat­ments (or how­ever many there are) is found in ev­ery one of the blocks, cre­at­ing a com­plete block. Plots need to be big enough so that the nat­u­ral vari­a­tion found in agri­cul­ture is min­imised, so big­ger is bet­ter. Don't make the mis­take the pea grower did of tak­ing his har­vester’s per­for­mance as a mea­sure of pea vol­umes. It is es­sen­tial to mea­sure the fi­nal prod­uct, the thing you sell to make money. That's easy in agri­cul­ture or hor­ti­cul­ture but harder with live­stock (very large plots and lots of stock are re­quired), so with an­i­mals, the sur­ro­gate mea­sure of pas­ture growth and lab­o­ra­tory anal­y­sis is mostly used. Sta­tis­tics are typ­i­cally the most con­fus­ing part. For­tu­nately the ANOVA test (anal­y­sis of vari­ance) found in most spread­sheets is typ­i­cally used. How­ever, if you are not com­fort­able with sta­tis­tics get help from some­one who is. While the ba­sics of an ex­per­i­ment, as out­lined, are re­ally pretty straight­for­ward, there are niceties in the de­tails that take ex­pe­ri­ence to get right so talk­ing to a sci­en­tist can help you. If the prod­uct you are ap­ply­ing costs $200/ ha to use but only in­creases in­come by $100 you are $100 out of pocket (profit has re­duced $100), although there may be some other ben­e­fit, like an in­crease in soil or­ganic mat­ter over the longer term which re­sults in big­ger yields in fu­ture. But the ul­ti­mate mea­sure­ment of an ex­per­i­ment is not yield, it is profit, so it is crit­i­cal that gross mar­gins for all the treat­ments are cal­cu­lated to test for the level of profit or loss.

Newspapers in English

Newspapers from New Zealand

© PressReader. All rights reserved.