 Today, we are going to continue our discussion on monsoon variability and agriculture. Last time we have seen that when the showing date is chosen to be the optimum one for a year which corresponds to the maximum yield, no moisture stress is experienced by the plant at the pot filling stage of 60 to 80 days after showing. So, for all the other showing dates, the moisture stress in fact is maximum at that point. And these are the other suboptimal showing that is when the moisture stress is maximum between 60 and 80 whereas, here for the optimal showing dates in fact the moisture stress is not at all there in this critical pot filling stage. So, this suggests that out of the different critical stages suggested in literature, the most critical is the pot filling stage which occurs about 60 to 80 or 65 to 85 days after showing. Now, this is tested by experiments with the model. See, this is a result that we got that it appears that pot filling stage is the most critical. So, then we tested it with the peanut grow model. What we did was we ran 3 years 83, 88 and 98. And firstly we ran them with the showing date taken as the optimum one and the actual rainfall that occurred in those years these were so called control runs. Now in addition to the control runs, we made runs by suppressing rain at different stages that is to say imposing dry spells at different stages. So, you see here this is the control run for 1998 and see in this case there is a dry spell imposed in the first month after showing. In this case the dry spell is imposed at this stage, in this one it is imposed at this stage, in this one at this stage and in this one at the pot filling stage. So, we introduced dry spells at different stages and then asked the question that would of course, not mean that the same amount of rain is suppressed because as you can see the rain varies from day to day. So, when you suppress this much more suppression occurs when you suppress this first part here then for example, when you suppress this part because here not too much rain occurred and so on and so forth. Now what are the results then? So at different stages vegetative stage which is up to 25 days after showing flowering stage which is 25 to 35, then the peg formation stage 35 to 50, then pod forming stage is 50 to 60, pod filling is 60 to 80 and pod maturity is 80 to 100. So these are the main life history stages of the plant, phenological stages of the plant and by suppressing rain during these different stages how much of a difference did it make to the total rainfall. So DR is the rainfall in that stage divided by rainfall during July to December into 100. So what was the fraction of the rain, total rain that occurred in that stage which we suppressed. So in 83 for example, vegetative stage 6.8 or 7 percent got suppressed same thing for flowering but peg stage more got suppressed 17 percent and pod filling in 83 45 percent got suppressed and similarly here and remember that they are in fact optimum showing date against which we are looking at things. Now similarly in 88 also we see in fact in 88 it is the pod forming stage that had maximum suppression of rain and so on. Now we also calculate what is the difference in the yield produced by this dry spell because we know from the control run what the yield would have been with the actual rain. Now we have suppressed part of the rain. So what has been the impact on the yield that is the question you ask and for example for 98 which is the figure you saw 7.4 percent yield got suppressed when it was in the first month the dry spell was imposed when it was imposed during flowering it was very little suppression then peg formation 13 percent then 23 percent got suppressed when you had the pod formation. But the maximum suppression in yield which is 57 percent occurs when we impose a dry spell at pod filling stage which is 60 to 80 days remember we have already identified that as a critical stage. So we saw with the rainfall that we got suppressed which was only 15 percent 57 percent was the impact on the yield. Similarly here also the rainfall suppressed is only 23 and in fact rainfall suppressed at the pod forming stage was 30 percent and the impact on yield was only 1 percent. But here the rainfall suppressed was 23 percent and impact of the yield is 40 percent. Similarly here the impact on the yield is 50 percent but there actually a lot of rain did occur in that stage which got suppressed 45 percent was the change in rain. But in any event for these 3 months sorry for these 3 years what do we see that amongst all the stages of the plant the impact of a dry spell is much larger than any other stage for the pod filling stage of 60 to 80. So by far the largest impact is of the dry spell at the pod filling stage 60 to 80 days after sowing. Now in fact this is a picture again to make the case that when you show makes a difference to what sort of moisture stress the plant will actually have and this is a case of 72 and when you have late sowing there is not much moisture stress here but when you have early sowing huge moisture stress at the pod filling stage. Same thing here in 83 early sowing would have led to this kind of thing whereas optimum sowing leads to no stress at the critical pod filling stage. So once these results were obtained and I must say that these are not results that we expected because when we looked at the literature there were papers saying every stage is separate paper saying every stage is critical some would say the first month is very important we should not get dry spell and so on and so forth. So we had no idea when we did this work that this pod filling stage would be the most critical but once we found it out through the model then the question was to see if this was consistent with the experience of the farmers. So what were the results we wanted to convey to them a that the broad sowing window that they choose 22nd June to middle of August is very reasonable for avoiding crop failure but within that sowing window if they could postpone the sowing to mid July and after that they would get even higher yields. This was the result from the model that we wanted to tell them and we told them this result and they were of course surprised because they did not quite expected. See right now they always show at the first opportunity in Poonarvasu which is to say early July that is what they prefer and we were trying to say that do not show at the first opportunity wait for a sowing opportunity a little later in the month in the next Nakshatra and they were a little skeptical about it in any event we thought let us also check with them about our result about which stage of the crop is more sensitive. So in order to test whether our results about the critical stage were consistent with their experience we asked them at which life history stage is the magnitude of the impact of a dry spell on the yield the largest. So we are asking the same question that we asked the model and we wanted them to tell us on the basis of their experience which is the most critical stage where if a dry spell occurs you would get a huge loss in the yield. Now Mr. Lakshman whom I have introduced to you in the last lecture who is by far the most knowledgeable of the farmers that we had immediately replied for filling stage. He also added that a dry spell in the first month after sowing has hardly any impact. This again is totally consistent with our result this is Lakshman again. So he is in fact the one who introduced TMV2 to this region the variety TMV2 which everybody is growing and he is a very careful observer and he actually supported the result we had that pod filling is the most critical stage. So the modeled results were found to be consistent with the experience of the farmers. Now the question is why does the pod filling stage being the most critical one imply that probability of very good yield is highest for sowing during 20th July to 9th August. See we have to understand this we have two results one is that we get very high probability of good yields when the sowing date is between 20th July to 9th August and second is pod filling stage is most critical. Now is there a connection between the two and let us just see here these are the probabilities you know of crop failure this is of reasonable yield that is more than 1000 and this is of more than 1500 and you see that right from 20th July onwards here onwards you get very good chance of getting very good yield it is over 60% and already the chance of getting reasonable yield is 20 this means 80% chance is there that you will make profit in this if you sow at this time if you get an opportunity to sow at that time. Now you remember these are the probabilities based on weekly rainfall this is 88 years of data at Anandpur and on the basis of that we have derived what is the chance of rain greater than 2 centimeters or 1 centimeter in the week this is wet spell likelihood and what is the chance of rain less than 0.25 centimeters or less than or 0 so this is the chance of 0 and so this is the probability of wet this is the probability of a wet week this is the probability of a dry week and what do we find we find that from around 20th September or so till early August is a huge peak in wet spells which coincides with a huge trough in the dry spells. So probability of wet spells is maximum around here and probability of dry spells is minimum around here so in some sense in terms of rainfall this is the most favorable period for plants. Now so it is seen that within the rainy season the probability of a wet week is the highest for the 3 weeks beginning around 21st September during this period the probability of dry week is also very low so if the showing date is 20th of July the pod filling stage falls exactly in this favorable period 20th September onwards for 10 20 days so 20th September to 10th October. So in fact if the showing is done on 20th July the pod filling stage falls exactly in this favorable period for later showing dates in the period 20th July to 9th August also the probability of a wet week in the pod filling stage is still reasonably high but chance of a dry week somewhat higher this is why for later dates you do not get as favorable as it is around 20th July. So we had found that this is a good period 20th July and a few days beyond that for showing that is when the yields are maximum we also found independently that the most sensitive stage is the pod filling stage and now we see that in fact the optimum showing date is such that the most sensitive stage pod filling stage occurs at a period when the probability of a wet week is highest and probability of a dry week is lowest that is to say it is most favorable thus the optimum showing window is associated with the most critical life history stage in the most favorable part of the rainfall profile. So this is what everything is now tying together nicely we have found the optimum showing window we have also shown that power filling is the most critical stage and we have seen why the showing window is optimum because if shown in at that time then the critical stage falls in the most favorable part of the rainfall profile of that region. So now the question is the following that farmers look for opportunities to show right from last week of June and they generally like to show in the in early Poonar Vasu which begins around 6th of July 6 to 20th July that is to say 15 days before our optimum. Now we tell them no no no if you were only to wait for two more weeks chance of you are getting high yields is much higher okay then they ask the question suppose we give up as showing opportunity earlier in July what guarantee is there that we will get another opportunity in later July this is a very very valid question see farmers believe that planting during Poonar Vasu which is 6 to 19 July leads to high yields if based on this analysis we suggest that they forgo planting opportunity in this period if they do not show although the showing rains have occurred and the soil is moist enough then we say no no do not show at that time just wait for the next nakshatra which is Pusha and that it begins on 20th July then they ask the question what is the probability of getting another opportunity later because this is a major worry expressed by the farmers that based on your analysis if we forgo an earlier opportunity you know in early part of July and wait for an opportunity in latter part of July or early August and it does not arise then we will have lost the whole crop trying to get maximum yield see this is not a risk that they would like to take so in order to address the problem of ascertaining the showing opportunities in a specific year what do we have to do as I mentioned before showing opportunities are decided by the soil moisture only when the soil has adequate moisture and in fact farmers tested by making a ball of the soil in their hand and seeing you know if the soil is not wet enough you won't be able to make a good ball you can make a good ball then they say okay the soil is reasonably wet we can go ahead and so so the important parameter is soil moisture only if the soil moisture is adequate can we get showing that now we want to know what are the opportunities for showing so what we need is a model for soil moisture because we don't have data on soil moisture like we have on rainfall we have data on rainfall we have some data on meteorological parameters but we don't have data on soil moisture if the criteria for when a showing rain will occur depends on soil moisture then we have to have a model for this and so in this semi-arid region the farmers do not plow the land or plant seeds unless they consider the moisture of the top layer of the soil to be adequate not only has the soil moisture to be adequate for showing but before that there are operations which involve land preparation they involve plowing and harrowing and so on now only for those operations also the soil cannot be dry it has to be sufficiently moist and they know how moist it has to be okay so what we need to do now is to generate a model which will give us how the soil moisture varies in a year when the rainfall variation is given and given the variation of the soil moisture then determine when the farmer would be able to plow when he would be able to harrow and when he would be able to sow all this has to be done if we need to ascertain how many showing opportunities arise in a year so based on the experience so far farmers have evolved rules for the periods during which they can undertake plowing and showing and these are the rules and SM stands for soil moisture and SA means when it's totally saturated okay and they have several rules that first plow will take place between 1st May and 25th June whenever the soil moisture exceeds 90% of saturation crisis plow if till 25th June no opportunity arises then they will undertake what they call a crisis plow and that will require only 75% of saturation of soil moisture you see and if that doesn't come then they had no opportunity to plow at all that year so cultivation is abandoned this is what they do then they try second plow that would be later of seven days after and so on and so forth and we will get into details of what they do after we learn about how to get the soil moisture so hence for ascertaining the timing of the different operations a hydrological model which can generate the soil moisture from meteorological data given the soil characteristics is required now extreme dry and wet conditions of the soil also promote growth of soil pathogens which can have considerable impact on the yield thus information on soil moisture is important for deducing the impact of pests and diseases on the yield I must also mention that models like peanut grow have built into it a soil moisture model so the growth of the plant in the model also depends on the soil moisture which the model itself derives okay so now we look for a simple soil moisture model so what we did was to develop a simple two layer models for deriving soil moisture okay evaporation and evapotranspiration is assumed to occur only from the top layer while moisture is retained in both the layers runoff is assumed to occur only when a part of the rainfall leads to saturation of both the layers now how deep are these layers see when the soil is bare the depth of the top layer is taken to be 15 centimeter because radiation cannot penetrate below that level that is the top layer of the soil which gets heated and so on so that is the layer that is going to take part in interaction with the atmosphere this is taken to be 15 centimeter now depth of the top layer is assumed to increase from 15 centimeter at the time of sowing to 60 centimeter at the time of flowering which is 35 days after sowing because the grounded crop is expected to have developed its maximum effective roof depth of 60 centimeter by this stage so initially the plant is developing its root and it is going deeper and deeper into the soil as it is going deeper and deeper into the soil the soil moisture of the layer level to which the roots have reached is important that is what we have to keep a tab on so what we do is we increase the top layer from 15 centimeters to 60 centimeters as the roots grow and the second layer is assumed to x 10 from the bottom of the top layer top layer is 15 centimeters to 60 so initially top layer is 15 centimeters bottom layer is 45 centimeters top layer keeps on becoming deeper bottom layer keeps on becoming shallower till a 35 days after sowing there is no bottom layer left at all it is just one layer which is 60 centimeter deep okay now we are going to look at change in soil moisture in both the layers and what brings about the change in soil moisture you know very well this is a simple water balance approach okay so change in soil moisture equals rainfall which is the input of water evapotranspiration see evaporation is what happens when there is a bare surface evapotranspiration why because when there are plants plants also transpire so the net amount of water going into the atmosphere is the evapotranspiration when there are plants so change in soil moisture depends on how much rain occurred above how much got evaporator okay now then there is also runoff so one if we can calculate all these we can actually calculate what is the change in soil moisture now as I say we did it by a sample water balance approach by this is a model developed by professor Amprasad that our institute a very simple minded one-dimensional water balance approach now since we are interested in rain fed conditions it is assumed that the rainfall is the only source and there is no irrigation obviously and the soil moisture is depleted by evaporation or evapotranspiration depending on whether it is bare soil or cropland okay now when does runoff occur now this is a big assumption that is made in most of the soil hydrological models that only when there is enough rain so that over and above the evapotranspiration the soil moisture gets enhanced to a level that it is saturated in both the layers only then whatever is left over will be sent off as runoff from the field now this is an assumption this is an assumption which need not be valid in fact we know very well that in intense thunderstorms that occur in a pre monsoon season for example the runoff there is a lot of runoff seen but deep layers of soil do not get saturated because the rainfall is so intense you know runoff is a highly nonlinear function of rainfall which depends on rainfall irrespective of what is happening to the soil also so this is an assumption that is made and it is a it has to be made because we do not have data on runoff we can estimate evapotranspiration in some ways and I will show you how but once evapotranspiration is estimated we know what is rainfall minus evapotranspiration that water we take first to saturate the soil and only if any water is left over and above saturation then that is sent off as runoff so this is the basic assumption of the model that rainfall is the only source and soil moisture is depleted by evaporation or evapotranspiration and runoff is assumed to occur if there is excess rainfall above that required for saturation of both the layers evaporation from bare soil is assumed to occur at a rate proportional to the moisture content of the soil evapotranspiration estimation from cropped land is the major component of soil moisture modelling it is one of the major unknowns here so daily evapotranspiration is modeled as and I am not going to get into details of that and actually evapotranspiration is calculated from what is a potential crop evapotranspiration potential means if there was plenty of water available to the plant how much would the evapotranspiration be that is the maximum possible evapotranspiration actually evapotranspiration is generally a fact fraction of that which we get by using daily pan evaporation measurements using the pan coefficient and crop coefficient okay so there is a way of getting this now the roots of the plant cannot suck the water from the soil if the soil moisture is below a certain value okay it does not have to be 0 for that even if it is below a certain threshold the plant cannot suck it up this is called the permanent wilting point because when the soil moisture is below the critical value since the plant cannot get any moisture from the soil it will begin to build so this is called a permanent wilting point the maximum water the soil can hold is the field capacity this again varies obviously from soil to soil and we know that for red sandy soils the field capacity is less than clay or black soils that we get in other parts of our country so PWP and FC are expressed as moisture per unit depth of the soil the maximum water available to the plant is therefore root depth into carrying capacity FC which is the maximum field capacity minus the wilting point because it simply cannot take up water if it is below the wilting point so FC-PWP is the available moisture per unit depth you multiplied by the depth of the root you get what is the moisture available to the plant which is SA for sandy loamy soils of Anandpur we take carrying capacity this is per unit volume to 24 percent wilting point is 10 percent so the available maximum available soil moisture is 14 percent now having built this model having put in all the numbers that we could get from the literature and so on and the characteristics of the soil and so on we want to see how good the model is and it is extremely important to do so because any model one uses has to be validated for the site you are using it for so this is the validation that we did for 1998 up here is the rainfall pattern for that year and this is the soil moisture pro pattern simulated by the model and the circles are the observations and you can see that in the observations also there is a big dip here and then the moisture builds up here then begins to decrease here builds up here and is decreasing here all of that is captured very nicely by the model so the model is able to capture the variation between wet spells and dry spells for our purpose this model is good enough is what we concluded now there is another hitch that the daily data on pan evaporation is not available for many stations for many years however monthly data is available so we tested the sensitivity of the simulated soil moisture to the daily data on pan evaporation by comparing the result for 1988 with that obtained using daily value interpolated from the monthly one see monthly is available we can always interpolate and generate daily values these are fictitious of course but the point is that this pan evaporation does not play a great role in deciding the variation of soil moisture the key variable for that is rainfall and therefore you can get away with these kind of things and we did this model experiment replacing the daily pan evaporation by interpolated monthly and what you see is the solid line is of course the one which uses daily pan evaporation data dashed line is the one which uses interpolated daily data obtained from monthly data and you can see that there is hardly any difference in the two patterns because what determines this huge fluctuations is really the rainfall so we do not have to worry even if we do not have daily pan evaporation we can still generate soil moisture using this model now I was talking of land preparation for sowing in order to identify the sowing opportunities for a specific year that is a specific rainfall pattern it is necessary to spell out the required conditions for undertaking the different land preparation operations and sowing for the TMV2 variety of ground nut in our study area these were compiled on the basis of farmers input on what is practiced in the region see I talked earlier about how critical it is if you want to get answers to scientific questions that you have a genuinely interdisciplinary group with interaction this is another instance in which we would never have been able to model you know occurrence of sowing dates in ground nut had we not got enormous input from the farmers in the region as to what are the criteria they use for deciding on different operations for land preparation now the first operation is plowing as I indicated earlier the depth to which the land is plowed in the study region is 20 centimeter hence the soil moisture of a layer of depth 20 centimeter which we denote as sm 20 is calculated by appropriate averaging of the top and bottom layers of the model the maximum soil moisture for such a layer is denoted by s a 20 and taken as 28 millimeters this is consistent with the assumptions we had made for the red sandy sandy soil so what we say is for this 20 centimeter layer which remember is bare right now right the land has to be plowed in fact this is the situation that the maximum available soil is 28 maximum available soil moisture would be 28 millimeter now conditions for the different land preparation operations are in the next slide so what do we what do the farmers do if the soil moisture at of 20 which means the top 20 centimeters of the soil is greater than 90 percent of this s a 20 which we had taken as 28 see we had here s a 20 is taken as 28 millimeters that is to say it is 2.8 centimeters 90 percent of that would be something like 2.7 centimeters little less than that so so if that much of moisture is available in that soil then they will have do the first plow somewhere between 1st May and 25th June whenever this moisture condition is satisfied if the above condition for plowing is not satisfied before 25th June then they say okay now we cannot keep on waiting for the moisture to be enough for a regular plowing let us do crisis plow and so from 26th June to 30th July whenever the soil moisture exceeds 75 percent of the maximum possible they immediately do a plowing and that is called crisis plow if the above condition for crisis plowing is also not satisfied cultivation for that year is abandoned so they take two chances but if rain is simply not enough till end of July to even do crisis plowing then there is no point in cultivating groundnut is what they say now the second plow which is seven days after the first plow or 7th June whichever is later till 25th June so they have to plow the field twice and this is the schedule for the second plow and for the second plow it is enough if the soil moisture is over 75 percent of maximum available now if the above condition for second plowing is not satisfied the first plow itself is taken as the final plow so second plow in some sense is optional they are not going to abandon the cultivation if they cannot plow second time this is the practice that they have so if second plow is not possible then it is final plow first itself is taken as final plow then they have a process called harrowing which is done when either final plow date or 15 June whichever is later to 16th August and here the condition is that the soil has to be sufficiently dry for two consecutive days for harrowing if land is not sufficiently dry to harrow cultivation for that year is abandoned because they simply cannot prepare the land see they may have plowed it but if the land never gets dry enough to harrow they cannot carry on with the groundnut cultivation now then comes sowing so sowing can be done on the day of harrowing or 25th June whichever is later to 25th July so they have a limit for every operation and the condition for sowing is that the soil moisture has to be greater than 0.6 of the maximum available more than 60 percent now if the land is not sufficiently wet to so till 25th July late or crisis sowing is undertaken this late or crisis sowing is done either on harrowing day so this will not be done till almost end of July till 26th July whichever is later to 16th August so due to different conditions then you may end up doing these different operations at different times now day of sowing there are further conditions that we have to impose that they cannot sown or any day so day of sowning sowing should be a non rainy day and here we use the IMD definition of a non rainy day that rainfall is less than 0.25 millimeter now soil depth consider is s 20 a 20 centimeters the saturation is 28 millimeters and sm 20 is the available soil moisture so this is what they practice now these conditions are applied to a certain whether an opportunity to plow harrow and so occur in any year and if so the plowing harrowing and sowing dates are determined using the same table so now what have we done we have collected from the farmers what are the conditions and criteria which have to be met for them to undertake plowing harrowing and sowing and these are all determined in terms of soil moisture to generate soil moisture from rainfall we have a poor man's model simple model of hydrological model which seems to work reasonably well and from that model we get the soil moisture then we check whenever those conditions that they have laid down are satisfied then we call that a plowing date harrowing date and sowing date and in case no such opportunity arises we say the sowing got abandoned so the cultivation itself is abandoned now how many sowing opportunities are there this is clubbing all the 88 years how many sowing opportunities occur on different days so in the percentage of years so see around 6 july large number of years almost 50 percent of the years you will get sowing opportunity and it increases slowly to 60 percent near 60 percent towards august and so on so initially in may june which is when the package of recommendation said you should do planting the sowing opportunities are very very small up to june they are less than 10 percent and even up to 20 second june this is the sowing window that they choose now see after 20 second june it has crossed 20 percent so it makes sense that if they feel like sowing there is some chance of being able to do so because remember it is not a matter of only sowing they have to have opportunities for plowing harrowing and so on after that okay now this is not a figure that you will be able to figure out but for every year we have calculated what are the sowing opportunities and in which year's crisis sowing was taken and in fact you will find that in some years cultivation just had to be abandoned like this year here no cultivation possible this year no cultivation possible the rainfall was so poor that it did not happen so out of 88 years in 6 years the sowing had to be abandoned out of these in 4 years there was no opportunity to plow whereas in other years there was no opportunity for harrowing and sowing so in the rest of the years there were several sowing opportunities in different parts of the rainy season in particular out of 88 now we are talking of in 62 years there was an opportunity to show in punarvasu which is the favorite sowing season for the farmers 6 to 19 july it's important to note that in all these years in each of these 62 years there was also a later opportunity of sowing either in pusha which is the next nakshatra 20th july 2 2nd august or ashlesha which is the nakshatra which follows which is 3 to 16th august so although they like to sowing punarvasu if they decided to forego that to you know maximize the chance of getting high yields in each of the years in which opportunity occurred in punarvasu a later opportunity also occurred this is a very very important result because it suggests that giving up an earlier opportunity in the early part of july for an opportunity in the optimum sowing window that we have determined with our model suggested earlier does not involve any risk this is what we are saying that as far as our data is concerned there is no risk involved in this operation now this is something that I will come to when we talk of indirect impact of weather events as I mentioned earlier you know wet spells and dry spells can also promote or trigger attacks by pests and diseases dry spells for trigger attacks by leaf miner wet spells trigger attacks by a thing called late leaf spot or tikka so this is a indirect impact of climate events or weather events see so far through peanut grow model what we looked at was only the direct impact of rainfall direct impact of rainfall on soil on how it impacted on growth of the plant on the yield and so on and so forth this is the direct impact but there is also an indirect impact of wet spells dry spells and so on which generates pests and diseases as I mentioned before now we have very large areas of monocropping and because of that this pests and diseases in the region and in almost every rainfed region in India have become endemic this means that a low level of low density of these is around all the time and if a favorable weather event occurs that is to say favorable for them pests or diseases immediately epidemic starts you know pest population increases and they start attacking the plant and the losses can be quite high so one has to also think of going beyond the crop models which we have already which are already developed and with us to think of how to take into account these indirect impacts of climate events and that is leaf miner and tikka I will not talk about how we determine this now because that will come when I talk about the model for these pests and diseases but this is only showing you the probability of incidence of leaf miner and late leaf spot depending on the showing date and there was a belief amongst the farmers that if you so late then the chances of pests and diseases are higher but actually we find that is not true at all leaf miner after 20th of July it very rapidly decreases chance of getting leaf miner and even tikka or late leaf spot also decreases right from about mid July onwards so as far as just the triggering of these attacks by weather events are concerned the probability of incidence of these is going to decrease with late showing window but farmers tell us that you know actually if somebody cannot so on time and so late the attack is much more on his field but that has to do with the fact that what we have not taken into account in our model is that these pests move from one place to another this is also spreads from one place to another so if you have patches in which plant is crop is planted earlier lying side by side with patches in which it is planted later then diseases or plants which attack the early sown crop can easily travel to the other field and attack it so this is something we have not taken and this could in fact lead to problems in having more pests and diseases when you do not have uniform showing over large areas so this is something we will have to worry about when we look at the model for pests and diseases so what are the conclusions now we have made a very detailed study in response to the question that farmers post to us what is the optimum showing rate of the region saying that what is in the package of recommendations is not right we have come up with another showing window and how good is that and can you make suggestions as to how to improve the yields by changing the showing window somewhat now as you know we did this exercise by taking a lot of rainfall data by taking a validated crop model and doing experiments to see how the yield changes with showing date and what do we find we found that the broad showing window used by the farmers which they have empirically hit on by trial and error they have come to this showing window for this region over the two or three decades in which it is grown so this is different from the one that was in the package of recommendation is the one which minimizes the risk of crop failure so it is a very rational choice they have already come empirically on to a choice of showing window which is very reasonable in that it avoids the chance of crop failure it minimizes the risk of crop failure now what we found was that within this broad window of 22nd June to 16th August the farmers generally use the first showing opportunity however we find that the overall yield is higher for showing in the period 20th July to 10th August again we had shown that the probability of obtaining high yields increases from about 30 percent for showing in the beginning of July to over 60 percent for showing at the end of July only in 25 percent of the years is the yield higher for showing prior to 20th July so 75 percent of the years the yield will be higher if one can show after 20th of July and again we have to note that if there is an opportunity to show in early July which is when farmers do like to show when they have an opportunity in all the years in which such an opportunity occurred in early July another opportunity did occur after 20th July also so it is possible for them to forego the first showing opportunity and so in this window that we are suggesting that is the latter part of the showing window that they have adopted we have also now this is something that we just found by running several years 88 years of the model we found that this is an optimum showing window but then we also looked into why this is optimum and we found that the most critical period with a large impact of dry spells is the pod filling stage so as far as the life history stages of the plant are concerned the most critical stage is the pod filling stage which is about 60 to 80 days after showing now dry spells in the first month after showing have relatively little impact this is also something that we showed in that experiment you remember that we did with 3 years so again it has gone wrong okay so generally so we have found that the sensitive period sensitive stage is the pod filling stage dry spells in the first month after showing do not have much of an impact now actually the peanut grow model also calculates the moisture stress experienced by the plant so using that we could show that if you see in general you know the moisture stress is highest in the pod filling stage because the moisture stress grows with the plant see as more and more leaves grow and the plant attains its maximum height then the need for moisture also becomes maximum and so when you have a dry spell the moisture stress is also maximum so by 60 days the moisture stress generally is the highest at pod filling stage but we found interestingly enough that when the showing date is chosen to be optimum for that year then it turns out that the plant experiences no moisture stress at all in the pod filling stage now this is very interesting this is could be done only because of the model that we could show that in fact the optimum showing dates which we discovered by analysis of how the yield varies with showing date actually implies that there is no moisture stress in the pod giving pod filling stage of the plant when the showing date is optimum now so this already tells us that we can do a lot with these results without actually having to run the model you see what is it saying now we have to remember that in this region the rainfall in any season you don't get continuous rain or heavy rain all the time like you do in rainy parts of the country this is a semi arid region so what happens every year you have some wet spells interspersed with dry spells the probability of occurrence of these spells varies in time right we have seen that also so probability of wet spells is high and dry spells is low during this special window 20th September to 10th October for this region whereas reverse case occurs in the latter half of June perhaps I have it here no I don't okay so given any region and given the rainfall we can easily derive you know what are the probabilities of wet spells and dry spells in different parts of the rainy season as we have done here now what do we want to do we want to maximize yield okay for that we would like to avoid dry spells at the critical stage where dry spell can have a very large impact on the yield so if for maximizing the yield the dry spells are to be avoided at some stage of the crop the showing that can be adjusted so that the probability of dry spell is least at that stage okay so you can you know move the thing around visa with the showing date move the rainfall pattern around visa with the showing date and thereby ensure that the probability of dry spell is minimum when the plant is at its most sensitive stage this one can do but you know dry spells just cannot be wished away so what you do is you can this is what one means by tailoring to the nature of the variability now we are saying since dry spells at the pod filling stage which occurs at 60 to 80 days after showing have a large impact on the yield and dry spells in the first month after showing have relatively little impact both these results have come from our model work then tailoring to the variability means that you should actually reduce the probability of dry spells in power filling stage that is how the showing date should be chosen while if required enhancing the probability in the first month in other words as you are shifting the thing around you know you cannot just make do with no dry spells at all in any stage of the crop dry spells will occur the point about tailoring is that you do not let them occur when the plant is most sensitive most vulnerable to dry spell that is what one means by tailoring and that is how showing date is used for tailoring now what is interesting about this study is it actually required a lot of modeling it required modeling using of crop models which were fortunately validated by some other group and it required getting the rainfall data and running the models analyzing it and so on and so forth but at the end of the day what we have learned is that for ground at power filling is the most critical stage and the way to get optimum showing is to somehow match the power filling stage with the minimum probability of dry spells in the rainfall pattern of that region that is what we have learned. So, this can be without running any models or anything if you have another variety of ground at that you want to grow in this region for which the power filling stage is at a different stage not 60 to 80 days as it is for TMV 2 and so on then it would be possible for us to right away conclude what the showing that should be. So, the thumb rule would be to choose a showing window such that the rainfall variability of the region for that rainfall variability that the region experiences the probability of dry spells at the power filling stage should be minimum see by using this kind of a thumb rule we will do much better than in terms of choosing the showing window of course, if you had a model it would help but having had the model the conclusions that we have come to are so so reasonable and so consistent with the experience that we using these the thumb rule we can derive and using this we can actually go ahead and determine the optimum showing date. Now, so far we have talked only of models as I mentioned before without pest and diseases and because of that actual model yield is much much higher than what farmers get on the field. Even then because the sensitivity of the yield to the showing date is similar we could derive a lot of results about optimum showing window using the peanut grow model, but eventually you would like a model which mimics reality in terms of variability of yield on the farmers fields. Now, to do that we cannot omit losses that occur due to pest and disease incidence. Now, there are no models none of the crop models incorporate pest and disease attacks. So, what we did was that we developed a heuristic model in which using again the enormous experience of the and knowledge of the farmers we could develop a model in which we could actually calculate by means of using the model what sort of pest and disease incidence would occur in any year and what would be the losses and thereby correct the yield generated by peanut grow by actually including losses that occur due to pest and diseases. This is what I will talk about next time. Thank you.