Six validation sites were chosen within the five country region of the LEWS project. These locations were:
Approximately, 20 diet:fecal pairs will be collected at all sites except Eritrea. LEWS is going to help rebuild infrastructure at the Eritrea location to better serve science issues in the project for that area.
It was agreed that there should be sufficient pasture size to allow approximately 50% use of the initial standing crop and not to exceed 70% over a 120 day grazing period. Generally, it was felt that the trial should start at the latter part of the growing season and extend into the dry season to allow quality profiles to drop during the period and capture variable growth rates of the vegetation and diet selectivity by the animals. At least 5 animals of a kind (cattle, sheep and goats) will be used in the pasture with no more than 15 head per kind. It was recognized that it may be difficult to run all three kinds of livestock but it is preferable to do so and run in the same pasture together to minimize the amount of vegetative sampling. This will require grazing areas of 5-40 ha depending on the productivity of the land, initial conditions and anticipated growth condition over the 120 growing period. A spreadsheet will be emailed to the country coordinators to be used for computation of pasture size given animal numbers, expected growth, initial standing crop, etc. The spreadsheet will be named valdsize.wk1 and saved in lotus 123 format compatible with all speadsheets (Lotus, Quattro, Excel). The sites can either be bounded by fence or flagged and herders used to contain the animals in the area. The initial forage conditions should be good in the sense that there has not been excessive grazing on the site and some time has been allocated for recovery of herbage standing crop. Recording of the GPS longitude and latitude readings for the centroid of the sites used in the validation study.
As stated the trial must have cattle, sheep, and goats. If more than one class of animal is used there must be at least five individual animals per class or that class should be eliminated. If more than one breed is used per class then there must be at least five individuals within the breed in the herd. We must keep in mind that we are not contrasting breeds on a site. Rather, we are trying to see if the analytical tools that we have assembled require modification and variability is coming from multiple locations across the region. Within class age structure must be chosen with care. For mature animals 5-10 years of age is all that is necessary. However, if younger then they must be within plus or minus 2 months of the average age of the group assembled for a given class.
Parasite management (dipping, spraying etc.) should be continued with routine health management of the site chosen for validation. Please record all events health incidences. We need to provide a budget to cover the animals during the experiment - from LEWS. If any supplements or minerals are provided, you must record what and how much is being supplemented with estimated daily consumption rates. You must retain a sample of the feed for quality analysis (need to know what kind and how much of concentrates are used). Please keep a log of the experiment as to what is happening in terms of weather, herders, animal health, other problems that emerged. Try to record the diseases or at least the symptoms, but note when veternary care is required and the actions taken. As to parasite counts (assume that these animals will be wormed at the beginning), do not need to routinely check for parasites, but rather do a count for an animal that is not apparently responding or is out of synch with its herd-mates.
For the animals used in the validation site, it would be helpful to measure hip height of the animals just above the hooks. Other body measures are optional. Also, we need to conduct an estimate of peak milk yield of all lactating females. The suggested method is weigh suckle weigh. Other measures are not essential other than noting the breed, sex, age, and physiological stage of each animal used in the trial.
All the animals will be penned on the evening before weighing and allowed access to water the entire time. Early in the morning they will be allowed to stir before weighing for at least 15 minutes to allow defecation and urination. They will be weighed rapidly and accurately and then turned out for normal grazing. The timing and technique should be repeated every time and communicated with the other validation teams. Scale calibration is required. Teams need to repair the scales and conduct lower and upper weight tests on the scales to make sure they are working correctly. It was agreed to use scales that weigh to one-half kilogram and we will tare the scales between animals to minimize scale drift between weighing.
At every weighing, body condition score each animal in the herd. Use the 1-9 system for cattle and the 1-5 Australian system for sheep. After debate, it was decided to use the 1-5 system of sheep for the goats. However, we can use the 1-3 system if we feel it does not work. The validation teams will exchange methodology between sites to assure standardization. The 1-9 ILRI system was proposed and accepted as the body condition scoring method for cattle. If a team does not have access to the ILRI blue book, contact should be made with the Iliad Ababa library for a copy. Pictures should be taken of the animals at each weighing and then exchanged with the other validation sites with notations as to how they were scored. Please provide your colleagues with your opinion to help standardize the regional estimations. When you take a picture, record the roll number, frame number, date, location, and description of subject.
Teams will collect a composited fecal sample from each kind of animal in the herd representing a majority of the individuals in the head on a weekly basis offset by 3 days of the weighing. Try to establish a routine that samples are always taken at morning release from the boma, say on Wednesday for instance. A heaping handful of sample is ample for the analysis. The samples should be dried at less than 60 C and well aerated during drying to prevent mold formation or excessive core temperatures. After drying, crumble the sample around in your hands or over a screen to allow the dried material to break loose from the pellets. After drying, place each kind of animal's composited feces in a bag and store out of the sun and in a place not subject to excess heat or where rodents and insects can get to them. We need to send the samples to the Debre Zeit lab on a continuous basis to allow them to grind them through their Trecator cyclone mill and not get overwhelmed. Debre Zeit will split the sample and ready for bulking to the USA at the end of the trial. TAMU's GAN Lab will train the NIRS technician and hopefully have the machine in place by late May to allow reading of the samples in East Africa shortly after the trials begin.
For each of the validation sites, they should be measured to get the size ad shape of the site grazed. Also, the vegetation should be mapped if more than one ecosite (range site) to minimize vegetation sampling. This means that a large contiguous area will have to be identified on one soil type.
One soil pit needs to be dug in each ecosite to characterize the horizon's thickness. If the soils are complex then that complexity needs to be measured. We will try to use the pedon description for the soil if available. However, each team is urged to bring in a soil scientist to help and derive the following data where possible: surface storage depth (% area ponded and average depth per ponded spot), slope, horizon thickness and for each layer cm/cm water holding capacity of each layer at 0, -1/3, -15 bars, saturated hydrologic conductivity, bulk density and percent of volume occupied by rock.
It was decided by the LEWs team that 15, 50m permanently marked transects would be established on each validation site and marked at both ends with durable markers. The transects should be allocated in some systematic manner to allow rapid relocation at each sampling date. All of the data required for vegetation validation will be acquired along these transects. The transects should be read at the beginning of the trail, and then within plus or minus 2 days of the weighing (every 28 days) and at the end of the trail.
Prior to initiating the trial, measure the cover, density and physical dimension of the shrubs on the transects. Stretch the 50-m tape between the permanent markers of each transect. Walk along the transect and measure the interception interval of each plant's canopy. Record the meters of interception interval of each plant's canopy. Record the meters of interception to the nearest cm.
As you walk along the tape, record all woody plants rooted within 1-m to one side of a 1x50-m belt transect. This provides the means to determine plant density and scale up the measured canopy volume to m3/ha. For each tree encountered in the 1x50-m belt, measure the following dimensions: diameter at the widest part of the canopy (D1), diameter at the base of the canopy (D2), height of plant canopy (H1), height to widest part of canopy (H2) and height to the lowest part of the canopy (H3). Using the pedistal parabolic frustrum model (formula will be provided), compute the volume of each plant's canopy (m3). Knowing the total volume of canopy of a species in the belt multiplied times the plot factor (10000 m2/50 m2)=200. This allows you to compute m3/ha but not kg/ha. This is only needed at the beginning of the trial as error in measures more than override differences due to grazing pressure.
The next step is to determine the weight per unit volume of each woody species on the site (kg/m3). Construct a 3-d frame of 30 wide x 30 eight x 50 depth cm. Select 5 plants of each species in the plots. Insert the frame to the surface of the woody plant's canopy below 1.5 meters height and clip out all current year's growth (leaf+stem) and then go above 1.5 meters in height and repeat the same process. This allows you to determine the changes in weight per unit over the trial. This should be repeated at each measurement period (5 times over 4 months). We originally thought we would use a reference unit concept but after further study, it was determined that measures of thinning woody plant foliage due to grazing had to be accounted in the sampling method.
In order that PHYGROW can be tested properly, there needs to be characterization of the initial mass of current years and old years growth standing in the grazed validation site. The prior measurements allow you to measure current year's growth but does not measure old growth. To measure old growth requires use of allometric techniques. Essentially, we are attempting to find some easy measure of geometry of the plant to estimate how much dry woody bio mass is in the tree canopy and trunk. First, we will determine if allometry equations are already present for many species and determine their adaptability to the research sites. However, for the dominant trees, we must pursue creation of this measure.
To create an allometric equation, we must identify 12 trees of representative heights (small to large), and then measure the canopy dimensions (as before) and the basal diameter of the stems and the number of basal stems. Then the tree is cut and current years growth removed and the wood is weighed. Small subsamples need to be taken, weighed and then dried to determine the weight on a dry matter basis. We then will use a series of different mathematical equations, usually log-log, to generate a regression equation to predict wood content from other measures such as canopy diameter. We have numerous equations for sub-tropical species in southern Texas and there are sources within ICRAF as well. Other sources for allometric equations should compile a unified list of equations by species. In Ethiopia and Eritrea, cutting of trees is probably not possible due to local laws and customs. We will probably have to use equations for species that appear similar to those where we have equations.
A double sampling technique will be used to characterize standing crop by species with the validation site during each sampling period (initial plus every 28 days). A 1 x .5 m quadrant (0.5m2) will be used to sample the herbaceous component. We will systematically sample four quadrants along each of the 50-m transects. Total herbaceous vegetation will be estimated in each of the quadrants every 4th one clipped, resulting in 60 plots read and of those 60, 15 are clipped to develop a regression for correcting the standing crop estimates. After you have estimated the bio mass in the quadrant, then visually estimate composition by weight of each species. The weight estimates require well trained people. Remember to offset which quadrant you clip (every 4th quadrant - from a different beginning).
All data sheets will be entered into an electronic spreadsheet and the files backed up on external disks. This allows records in paper, on the hard drive of the computer and on floppy disk. Data collected and entered into the computer will be sent to TAMU electronically as well for back up once a month. All data must be summarized as it becomes available and not held until the end of the trail. Monthly reports should be made to all the other team members to allow exchange of information, problems emerging and progress each month.
The minimum amount of daily weather data that should be collected on the validation site is rainfall (cm), maximum (C) and minimum temperature (C). If possible, total daily radiation and average wind speed would be helpful but not essential. This should be recorded daily and reported monthly to TAMU modelers. Also, prior to initiating the study, all historical weather data at the site must be sent to TAMU for processing and creation of weather generation parameters.
This component of the program is targeted toward increasing the robustness of the fecal NIRS profiling system in East Africa. We have shown that the system from sub-tropical Texas seems to work in the region. Also, we have developed an excellent database for the type of vegetation found in diets of cattle, sheep and goats in the highlands of Ethiopia, primarily crop residues and fodder trees. The role of the additional diet: fecal pairs is to broaden the vegetation types to assure ourselves that the fecal NIRS technology can move on out to pastoral households.
After considerable discussion in Nairobi, we agreed that diet: fecal pairs would not be generated in Eritrea due to financial strain on that facility. We also agreed that we would like to generate 20 pairs at each of the remaining five locations. However, if it is not possible, we will generate as many as we have time, money and human resources.
Remember that a diet: fecal pair is one animal one ration. We need to get the variation in diets not in animals to increase the power of the NIRS fecal profiling system. Each ration should be fed at least 7 days for small stock and 10 days for large stock with gradual quality transitions. Depending on the facilities, one to 5 animals need to be placed on a concrete floor stall which can be easily cleaned of feces and feed placed in bunks that do not allow a lot of wastage during feeding.
Create a ration of logical groupings of plant species that are easy to harvest and represents changes in both types of vegetation (grass, forb, browse) and season of vegetation (rapid growth, dormancy, dead). The key is to create up to 20 rations that represent the variety of expected conditions in your region. It is not essential that you do all 20 but it is essential to have good, well thought out rations.
To assemble a ration, collect a large quantity of forage of some major species. Then have it chopped up or coarsely ground (1-cm screen if dry). It is good to mix several types of species (grass, forb, browse). To minimize picking among parts and species it is good to spray it with or mix in molasses. Weigh in the amount allocated (excess of demand) and then weigh out the orts (refusals) and determine the moisture content of the allocated ration. The weigh all fecal output and subsample for moisture content. By knowing the amount consumed and the amount of dry matter output, you can determine the in vivo digestibility. It is critical that you keep at leas a heaping handful of manure during the last 2 days of the ration to be used for moisture correction and for fecal NIRS scans. Dry the same as the validation trial fecal material. When feeding the ration subsample it over the last three days of the feeding period and place in a bag. Dry it and ready it for wet chemistry determination of nitrogen content and subsequent derivation of crude protein. We will also run in vitro DOM analysis on the rations as well. Please share your protocol with the entire group when you get ready to initiate the trials.