English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

The DairyNZ Modelling Group has developed a model called the Conservation Demo to assist in identifying and conserving surplus pasture on dairy farms. The hypothesis was that a computer algorithm would be able to make decisions concerning timing and amount of conservation similar to an experienced pasture manager. Weekly grazing decisions on DairyNZ farmlets, with stocking rates of 2.6 to 3.5 cows/ ha, were made by a team of research staff over three years. The model was run for each farmlet to compare its recommendations with the conservation decisions made by the research team. In spring on all farmlets and in summer on the higher stocked farmlets, the model was more conservative than the team in recommending paddocks to be closed in all seasons. In summer 2011/12 the recommendations from the model were more variable week-to-week (paddocks being closed one week and not the next), and on the lower stocked farmlet the model was more aggressive than the team. The Conservation Demo is a useful tool for identifying surplus pasture if the output is moderated by the farm manager, and the use of a feed wedge in conjunction with the model would improve the decisions. However the model requires pasture covers for each paddock, ideally at least weekly, and currently <20% of dairy farmers collect these data on a regular basis. Introduction Accurate and timely identification of surplus pasture is important on dairy farms for conservation of feed for periods of deficit, and for maintaining the quality of pastures grazed by the milking herd. Being aggressive and closing too much of the farm area for conservation results in underfeeding of the cows or supplements having to be fed. Being conservative and closing too little of the farm area results in higher post-grazing residuals that lead to decreased pasture quality at subsequent grazings (Lee et al. 2008) and this can reduce milksolids production (Hoogendoorn et al. 1988). The DairyNZ Modelling Group has developed a model called the Conservation Demo to assist in identifying surplus pasture. The model determines if a surplus exists, and whether it is a true surplus, and recommends which paddocks to close for conservation. This study set out to test the hypothesis that the Conservation Demo would be able to make decisions concerning timing and amount of conservations equally as well as an experienced pasture manager. Methods The Conservation Demo The Conservation Demo follows the pasture conservation rules developed by Macdonald (1997). These rules involve testing three conditions to identify if there is a pasture surplus and if the surplus can be harvested. Condition 1: Is the average pasture cover (APC) on the grazing area greater than the required cover? The required cover is the mid-point between the target grazing residual and the target pre-grazing cover. This condition determines if a surplus exists. Condition 2: Given the rotation length, will there be enough days for a break to grow back to the target pregrazing cover for the following rotation? This condition determines if there is a true surplus, i.e., if there will still be a surplus in the near future. Condition 3: Is the cover of the highest yielding paddock (excluding paddocks already closed) greater than the target pre-grazing cover? This condition determines if the surplus can be harvested without compromising animal intakes. If all three conditions are met, the Demo rechecks if the conditions are still met with the highest yielding paddock closed. If so, the Demo closes the paddock, and then rechecks the conditions for the next highest yielding paddock, and so on, until one of the conditions is not met. Model inputs The Conservation Demo requires the following data: Paddock information: Paddock name/number, area and current pasture cover (kg DM/ha) of all paddocks. Animal information: Herd size and feed required (kg DM/cow/day). Pasture information: Target post-grazing residual, rotation length, and expected pasture growth during the next rotation (kg DM/ha/day). In addition, a policy option for managing break size and rotation length when paddocks are closed for conservation needs to be selected: Policy option 1: Light cut – fixed break size. After 250 249-254 (2012) Proceedings of the New Zealand Grassland Association 74: closing, a paddock is immediately cut and returned to the rotation. Therefore, the rotation length and break size do not change. Policy option 2: Bulk cut – fixed break size. A closed paddock is left to accumulate growth. With this option, break size is maintained on the remaining grazing area. This means that rotation length has to be shortened because part of the grazing area is closed for a period of time. Policy option 3: Bulk cut – variable break size. A closed paddock is left to accumulate growth. With this option, planned rotation length is maintained on the remaining grazing area. Therefore, break size has to be reduced because the grazing area is now smaller (excludes the closed paddock). The model can be run using each of the policies in turn to compare the outcomes quickly. Model output The model runs the calculations for the three conditions, and lists which paddocks can be closed for conservation. A “step-by-step” button can be selected so that the user can see the calculations and outcomes for each condition, or if the “fast forward” button is selected, the model scrolls straight to the recommended list of paddocks to be closed. The user can change an input variable (e.g. policy option or expected pasture growth) and “fast forward” to see the impact on the number of paddocks recommended to be closed. Model assessment Each Tuesday, a team of five experienced pasture managers conducted a farm walk on the trial farmlets at DairyNZ Scott Farm. Members of the team used visual assessment, calibrated against quadrat cuts, of pasture cover in each paddock. After each farm walk the paddock covers were entered into a spreadsheet and a feed wedge (DairyNZ 2010) generated for each farmlet. The team then made the grazing decisions for the next week for each farmlet, including deciding which paddocks, if any, were to be closed for conservation. During the 2009/10 season there were eight farmlets on Scott Farm (Table 1). The model was run for each farmlet immediately following the weekly walk, prior to the team making the grazing decisions. The recommendations were compared with the team’s decisions, and as a result of this, it was suggested that the farmlets could be used to test the model. In subsequent seasons, the model was run for each farmlet after the team had made the weekly grazing decisions so that the team’s decisions were not influenced by the model’s recommendations. The results presented in this paper are from the subsequent seasons (2010/11 and 2011/12). The eight farmlets from the 2009/10 season (Table 1, Farmlets A, C, G to L) continued in the 2010/11 season, and an additional farmlet (E) with some cropping was also included. Data collection from all the farmlets finished at the end of May 2011. In the 2011/12 season, two new larger farmlets were established: Farmlet X in pasture and similar to the previous Farmlet A, and Farmlet Y, also in pasture, but with a lower stocking rate and using less nitrogen fertiliser (Table 1). From early September until early April each year, after the weekly grazing decisions had been made, the model was run for each farmlet. Actual values were used for the herd size, feed required, target post-grazing residual and rotation length inputs. Policy option 2 was used (bulk cut – fixed break size). For farmlets where crops were being grazed, or were about to be grazed, the feed required input was entered as the amount of pasture required. The expected pasture growth value was generated from a pasture growth model for each farmlet. The paddocks closed by the model were recorded. If the model closed no paddocks, or fewer paddocks than the team, the conditions not met were recorded. The hypothesis was that the model would be able to make decisions concerning timing and amount of conservation similar to those of the team. Where the recommendations of the team and model differed, it needed to be determined whether the team or the Table 1. Description of farmlets on DairyNZ Scott Farm in 2011 and 2012 seasons. Farmlets G to L are described by Clark et

Farmlet testing of the DairyNZ Conservation Demo