PastureFunctions module

author: Young

Pasture is the primary livestock feed source because in an extensive farming system it is a cost-effective source of energy available for the entire year. Different pasture types can be represented by altering the inputs for each pasture type. The default pasture type is “annual pasture”. However, by altering the inputs, perennial pastures and mixed swards can be represented.

Pastures are often included in a rotation to provide a break from cropping, which can rejuvenate soil conditions, provide disease and pest management and provide a cheap feed source for livestock. The pasture module generates the pasture production (as discussed below) and the costs and labour associated with seeding, monitoring, fertilising and spraying. Similar to cropping, the inputs vary based on rotation history, soil type and weather. AFO can then optimise the area of each pasture phase to include on each land management unit (LMU).

Pasture feed sources can be supplemented with concentrates, and in a mixed crop-livestock farm system the pasture can be complemented with dry residues from crop production (stubbles). The biology and logistics of pasture growth rate that are represented in AFO is:

  • Pasture growth rate (PGR) is dependent on pasture leaf area, which is quantified by the level of feed on offer (FOO, kg of DM/ha). Additionally, PGR for each pasture type varies with the phase during its life cycle, soil moisture, sunlight and level of growth modifier applied. All are quantified by their land management unit (LMU), time of year and weather-year.

  • The average FOO during a feed period depends on FOO at the beginning of the period, the grazing intensity and the PGR during the period.

  • The mobilisation of below-ground reserves (germination) of annual pastures at the break of season is dependent on the seed bank. The seed bank is controlled by the rotation in which the pasture is grown and varies with LMU.

  • The mobilisation of below ground reserves of perennial pastures at the break of season can also be adjusted by rotation. However, perennials usually are not grown in rotation with crops.

  • The maximum intake of animals grazing pasture depends on FOO and diet dry matter digestibility (DMD). Intake can be less than maximum which implies that the optimum solution can include rationing of animal intake via rotational grazing.

  • The digestibility of the diet selected by animals grazing green pasture depends on the sward digestibility and the animal’s capacity for selective grazing. Sward digestibility varies depending on the pasture species, the time of year, the LMU and the FOO of the pasture. Selectivity depends on FOO and grazing intensity.

  • Dry pasture that is not consumed is deferred to later in the year, with a reduction in both its quality and quantity. Livestock can select a higher quality diet when first grazing the dry pasture but quality reduces with extra grazing.

  • Livestock trample both green and dry pasture while foraging in proportion to the amount consumed.

  • The risk of resource degradation increases when ground cover is lower so there is a user defined minimum limit to ground cover during both the green and dry phases of the year.

The decision variables optimised in AFO, that represent the above biology are the:

  • rotation phases in which pasture can be grown on each LMU.

  • FOO profile during the year that is represented by a discrete range of FOO levels at the start of each feed period.

  • grazing intensity and the variation across feed periods during the year is represented by a discrete range of the severity of defoliation in each feed period.

  • level of growth modifiers (nitrogen or gibberellic acid) applied to the pasture.

  • quantity of dry feed consumed from each of 2 dry feed quality groups in each feed period.

The nutritive value of pasture is determined by the metabolisable energy per unit of dry matter, the relative ingestibility and the relative availability. This varies with:

  1. Feed period.

  2. The level of FOO. The greater the FOO, the lower the average digestibility of the sward. Lower digestibility of a high FOO sward is associated with the lignification that occurs in older foliage. Higher digestibility of a low FOO sward is associated with the higher digestibility of new growth that constitutes a higher proportion of the sward. There can be some error associated with this assumption if the low FOO was generated by grazing a high FOO sward back to a low FOO, in which case most of the DM would be stalk at the base of the plant which compares to a sward maintained at a low FOO level since the break of the season.

  3. Grazing intensity. With heavy grazing there is little scope for selection, so the diet digestibility equals the sward digestibility. With light grazing there is scope for selection and diet quality that approaches that of high quality leaf. Note, increasing the energy content of the feed also improves the ingestibility of the feed (Freer et al., 2007).

Pasture on non-arable areas in the crop paddocks is modelled as above with a few additions. Firstly, pasture on non-arable area is represented as a continuous annual pasture. Secondly, non-arable pasture on crop paddocks is not available for grazing until after harvest and therefore it goes into the low-quality dry feed pool. Accordingly, pasture on non-arable areas of the crop paddocks does not receive any farm inputs.

Pasture is split into the following functions (see PastureFunctions):

  1. Germination (f_germination)

  2. Reseeding of pasture (f_reseeding)

  3. Requirement for machinery for reseeding (f_pas_sow)

  4. Green pasture: consumption, growth (f_grn_pasture)

  5. Senescence of green pasture (f_senescence)

  6. Dry pasture: consumption and decay (f_dry_pasture)

  7. Pasture consumed on crop paddocks (f_poc)

  8. Limit on grazing for soil conservation (f_erosion)

AFO can handle multiple pasture types. The user simply needs to create a copy of the inputs and calibrate them to the new pasture.

AfoLogic.PastureFunctions.f_dry_pasture(cu3, cu4, i_dry_dmd_ave_p6zt, i_dry_dmd_range_p6zt, i_dry_foo_high_p6zt, me_threshold_fp6zt, i_me_eff_gainlose_p6zt, mask_dryfeed_exists_p6zt, i_pasture_stage_p6zt, nv_is_not_confinement_f, i_legume_zt, i_hr_scalar_zt, n_feed_pools)

Calculate the quality and quantity of dry pasture available throughout the year.

Dry pasture is represented by a low and high-quality decision variable. When green feed senesces a proportion of the feed enters each pool based on the digestibility of the senescing feed. This representation allows some diet selection to occur, with the higher quality component grazed by different sheep than the lower quality. The ‘high’ pool has digestibility of the 25th percentile and the ‘low’ group the 75th percentile. If the nutritive value of the feed is greater than the livestock’s requirement the quality is reduced by a given factor (see fsfun.f_effective_mei).

Non-arable areas of crop paddocks grow pasture all season, calculated as an ungrazed green annual pasture (see f_grn_pas). After harvest pasture on non-arable areas can be consumed. All pasture grown on the non-arable areas of crop paddocks is allocated to the low-quality dry pasture pool because it has grown all year without being grazed.

Dry pasture that is not consumed is passed to the same pool in the next period and the average quality and quantity reduces each period as it decays. Dry pasture does not transfer into the new season because once green feed is available stock will not graze old dry feed. Consumption of the high quality/high FOO component further reduces the average as the feed available skews towards the low-quality pool.

Note

There is not a constraint that ensures that the high-quality pasture pool is grazed prior to the low-quality pool (unlike the grazing of stubbles where it does exist).

Parameters:

  • cu3 – params used to convert foo for rel availability.

  • cu4 – params used to convert height for rel availability.

  • i_dry_dmd_ave_p6zt – average digestibility of dry feed. Note the reduction in this value determines the reduction in quality of ungrazed dry feed in each of the dry feed quality pools. The average digestibility of the dry feed sward will depend on selective grazing which is an optimised variable.

  • i_dry_dmd_range_p6zt – range in digestibility of dry feed if it is not grazed.

  • i_dry_foo_high_p6zt – expected foo for the dry pasture in the high quality pool.

  • me_threshold_fp6zt – the nutritive value above which scaling will occur for the NV pool.

  • i_me_eff_gainlose_p6zt – Reduction in efficiency if M/D is above requirement for target LW pattern

  • mask_dryfeed_exists_p6zt – Boolean array stating which periods dry pasture exists.

  • i_pasture_stage_p6zt – maturity of the pasture (establishment or vegetative as defined by CSIRO)

  • nv_is_not_confinement_f – boolean array stating which nv pools are not confinement feeding pools.

  • i_legume_zt – legume content of pasture.

  • i_hr_scalar_zt – Scalar for height of the pasture

Returns:

Quantity and quality of dry pasture.

AfoLogic.PastureFunctions.f_erosion(i_lmu_conservation_p6lzt, arable_l, pasture_rt)

The minimum FOO at the end of each period.

Due to sustainability bare paddocks are often avoided by farmers. To represent this in the model, pasture paddocks have a user defined erosion limit which specifies how much FOO must still cover the paddocks in the following year. This stops the model consuming all the dry feed and leaving the paddocks bare and exposed to wind erosion.

Parameters:

  • i_lmu_conservation_p6lzt – Minimum foo at end of each period to reduce risk of wind & water erosion.

  • arable_l – Proportion of arable land on each LMU.

  • pasture_rt – Boolean array linking pasture type to rotation phase.

Returns:

The minimum FOO at the end of each period for each rotation phase.

AfoLogic.PastureFunctions.f_germination(i_germination_std_zt, i_germ_scalar_lzt, i_germ_scalar_p6zt, pasture_rt, arable_l, pastures, phase_germresow_df, i_phase_germ_dict, i_destock_date_zt, i_break_z, rzt)

Calculate mobilisation of the below ground reserves of the pasture for each rotation phase.

For annual pastures, the below ground reserves is the seed bank which is mobilised through germination. For perennial pastures it is the root reserves that are mobilised at the beginning of the growing season.

The mobilisation is calculated for each pasture rotation phase based on the LMU and rotation history. The phase history is assumed only to impact pasture seed bank and hence pasture establishment. Phase history impacts the seed bank for three main reasons. Firstly, spraying out the weeds during the crop phase reduces seed set. Secondly, pasture manipulation in the prior years can reduce seed set. Thirdly, reseeding in the previous year can have carry forward benefits. For LMUs that have a component of non-arable area, the mobilisation on the arable area is affected by the phase history, whereas the non-arable areas are assumed to be the same as the continuous annual pasture rotation. Additionally, resown pastures or perennial pastures (e.g. Lucerne & Tedera) are only established on the arable areas, the non-arable areas are assumed to be growing annual pasture.

Parameters:

  • i_germination_std_zt – standard germination level for the standard LMU in a continuous pasture rotation.

  • i_germ_scalar_lzt – scalar for germination on each lmu.

  • i_germ_scalar_p6zt – scalar for germination in each feed period.

  • pasture_rt – Boolean array linking pasture type to rotation phase.

  • arable_l – Proportion of arable land on each LMU.

  • phase_germresow_df – array of rotation phases.

  • i_phase_germ_dict – Dictionary with inputs of rotation germination for each pasture type.

  • i_destock_date_zt – Date resown pasture is destocked - used to determine if resown pasture has volantary germination.

  • i_break_z – Date of the break of season - used to determine if resown pasture has volantary germination.

Returns:

pasture germination in each rotation phase.

AfoLogic.PastureFunctions.f_grn_pasture(cu3, cu4, i_fxg_foo_op6lzt, i_fxg_pgr_op6lzt, c_pgr_gi_scalar_gp6zt, grn_foo_start_ungrazed_p6lzt, foo_grn_reseeding_p6lrzt, i_foo_graze_propn_gt, grn_senesce_startfoo_p6zt, grn_senesce_pgrcons_p6zt, i_grn_senesce_eos_p6zt, i_base_p6zt, i_grn_trampling_ft, i_grn_dig_p6lzt, i_grn_dmd_range_p6zt, i_pasture_stage_p6zt, i_legume_zt, i_hr_scalar_zt, me_threshold_fp6zt, i_me_eff_gainlose_p6zt, mask_greenfeed_exists_p6zt, length_fz, nv_is_not_confinement_f)

Pasture growth, consumption and senescence of green feed.

The green pasture decision variables combine the representation of FOO at the start of the period, FOO at the end of the period, animal removal, energy per unit of dry matter and volume. Aggregating the decision variable to include all these factors allows representation (and optimisation) of:

  1. The intake capacity of livestock is affected by the level of FOO at the start of the period & level of grazing intensity. Both affect average FOO which controls the intake by stock. When there is more FOO, animals can eat more and achieve higher growth rates.

  2. Livestock diet quality changes with grazing pressure (eg. by running a lower stocking rate livestock can improve their diet quality through increased diet selectivity). This selectivity can be important for finishing animals for market or fattening animals for mating.

  3. The digestibility of pasture decreases as the length of time from the last defoliation increases (i.e. older leaves are less digestible). Having consumption and FOO in the same activity allows a drop in digestibility associated with old leaf to be included by linking digestibility to FOO. This is especially important for species such as kikuyu that drop in digestibility rapidly if pastures are grazed laxly and FOO increases.

  4. The pasture growth is reduced with higher grazing intensity because the average leaf area is reduced during the growth period.

These issues are likely more important in a system producing meat, where growth rate of animals and hence diet quality is critical to profitability. In a meat system the trade-off between quantity of feed utilised and quality of feed is quite different than the trade-off for a wool system.

For a given period, the decision variables are defined by starting FOO level and grazing intensity. There are three FOO levels; low, medium and high starting FOO and four grazing intensities; no grazing, low, medium and high. Green pasture decision variables represent the total green pasture on the farm in each period. The level of the decision variables at the start of the growing season are determined by the area of pasture and its level of establishment (see f_germination).

Gross pasture growth rate for each activity is calculated as a linear interpolation of the inputs of PGR by FOO using the input FOO level for the growth/consumption activity and grazing intensity during a period reduces growth rate during the period.

Diet digestibility is the overall quality of the pasture consumed by the livestock. The input value is the quality of the high quality component of the sward when animals have capacity to graze selectively (25% grazing intensity). At higher grazing intensity, the reduction in diet quality depends on the (input) range of digestibility within the sward. The reduction in digestibility for the 100% grazing intensity decision variable is half the range of digestibility within the sward. If the nutritive value of the feed is greater than the livestock’s requirement the quality is reduced by a given factor (see fsfun.f_effective_mei).

When the pasture senesces, it is removed from the green pasture pool and allocated into the dry pasture pool. The calculation of senescence is done in 2 parts. The first part relates to the older leaves of the plant senescing in the process of new growth. This is calculated as a proportion of the average FOO in the period. Thus, it varies depending on new growth and grazing. The second part of senescence is the death of the whole plant at the end of the reproductive cycle or in response to water deficit (termed end-of-season senescence). It is modelled to occur at the end of the specified period and the user specified proportion of FOO prior to end-of-season senescence is passed into the dry feed pool.

Parameters:

  • cu3 – params used to convert foo for rel availability.

  • cu4 – params used to convert height for rel availability.

  • i_fxg_foo_op6lzt – each level of starting FOO used in conjunction with PGR.

  • i_fxg_pgr_op6lzt – PGR at each level of starting FOO if the pasture is not grazed for each soil type (typically derived from a simulation model).

  • c_pgr_gi_scalar_gp6zt – The impact of grazing intensity within the period on the PGR achieved.

  • grn_foo_start_ungrazed_p6lzt – FOO at the start of the period if the pasture has been ungrazed from the start of the growing season.

  • i_foo_graze_propn_gt – proportion of the FOO available (greater than the base level) that is grazed for each level of grazing intensity.

  • grn_senesce_startfoo_p6zt – The proportion of the green feed at the start of the period that senesces.

  • grn_senesce_pgrcons_p6zt – The proportion of the green feed that grows during the period that senesces (or the reduction in senescence if FOO reduces due to grazing).

  • i_grn_senesce_eos_p6zt – The proportion of the green feed at the end of the period that senesces because it is the end of the growing season.

  • i_base_p6zt – The base FOO level which represents the level below which the pasture can’t be grazed. This base level is set with several criteria in mind; the physical limit at which animals can graze, an erosion limit below which farmers wouldn’t or shouldn’t graze; the optimum level of FOO based on the trade-off between PGR and FOO.

  • i_grn_trampling_ft – amount of feed that is trampled while green feed is being consumed as a proportion of the feed consumed.

  • i_grn_dig_p6lzt – DMD of the green feed that would be consumed if animals graze with 25% grazing intensity a sward that has medium FOO level in each feed period on each LMU.

  • i_grn_dmd_range_p6zt – Range of DMDin the sward between the 25th percentile to the 75th percentile.

  • i_pasture_stage_p6zt – Maturity of the pasture (establishment or vegetative as defined by CSIRO)

  • i_legume_zt – Legume content of pasture.

  • i_hr_scalar_zt – Scalar for the height ratio of this pasture.

  • me_threshold_fp6zt – The nutritive value above which scaling will occur for the NV pool.

  • i_me_eff_gainlose_p6zt – Reduction in efficiency if M/D is above requirement for target LW pattern.

  • mask_greenfeed_exists_p6zt – Boolean array stating which periods dry pasture exists.

  • nv_is_not_confinement_f – boolean array stating which nv pools are not confinement feeding pools.

Returns:

Quantity and quality of green pasture.

AfoLogic.PastureFunctions.f_poc(cu3, cu4, i_poc_intake_daily_p6lzt, i_poc_dmd_p6zt, i_poc_foo_p6zt, i_legume_zt, i_hr_scalar_zt, i_pasture_stage_p6zt, nv_is_not_confinement_f, me_threshold_fp6zt, i_me_eff_gainlose_p6zt)

Calculate energy, volume and consumption parameters for pasture consumed on crop paddocks before seeding.

The amount of pasture consumption that can occur on crop paddocks per hectare per day before seeding - adjusted for lmu and feed period The energy provided by the consumption of 1 tonne of pasture on crop paddocks. - adjusted for feed period The livestock intake volume required to consume 1 tonne of pasture on crop paddocks. - adjusted for feed period

Green pasture can be grazed on crop paddocks if destocking for seeding occurs after pasture germination. Pasture grazed on the crop paddocks in the period before destocking for spraying and seeding is represented as a pre-specified quality and maximum quantity available each day on the area that is yet to be seeded, with the additional requirement that pasture must be destocked 10 days prior to seeding to allow time for an effective knockdown spray. Grazing days are provided by the seeding activity which are calculated in mach.py and depend on the time between the break of season and destocking prior to seeding.

The main assumption/limitation of the poc representation is that the pasture consumption per lmu per day per ha is a fixed value per feed period.

Parameters:

  • cu3 – params used to convert foo for rel availability.

  • cu4 – params used to convert height for rel availability.

  • i_poc_intake_daily_p6lzt – maximum daily intake available from 1ha of pasture on crop paddocks

  • i_poc_dmd_p6zt – average digestibility of pasture on crop paddocks.

  • i_poc_foo_p6zt – average foo of pasture on crop paddocks.

  • i_legume_zt – legume content of pasture.

  • i_hr_scalar_zt – Scalar for the height of the pasture

  • i_pasture_stage_p6zt – maturity of the pasture (establishment or vegetative as defined by CSIRO)

  • nv_is_not_confinement_f – boolean array stating which nv pools are not confinement feeding pools.

Returns:

  • poc_con_fl - tonnes of dry matter available per hectare per day on crop paddocks before seeding.

  • poc_md_fp6 - md per tonne of poc.

  • poc_vol_fz - volume required to consume 1 tonne of poc.

AfoLogic.PastureFunctions.f_reseeding(i_destock_date_zt, i_restock_date_zt, i_destock_foo_zt, i_restock_grn_propn_t, resown_rt, feed_period_dates_fz, i_restock_fooscalar_lt, i_restock_foo_arable_t, dry_decay_period_p6zt, i_fxg_foo_op6lzt, c_fxg_a_op6lzt, c_fxg_b_op6lzt, i_grn_senesce_eos_p6zt, grn_senesce_startfoo_p6zt, grn_senesce_pgrcons_p6zt, max_germination_flzt, length_fz, n_feed_periods, p6lrzt, p6zt, t_idx, z_idx, l_idx)

Generates the green & dry FOO that is lost and gained from reseeding pasture.

Resown pastures germinate as a usual pasture (see f_germination) the FOO is then set to 0 for the periods when the paddock is ungrazable due to reseeding. At the time of restocking the FOO on the arable areas is set based on user specified inputs and the FOO on the non-arable area is calculated from the FOO at destocking plus the growth over the destocked period. Pastures that are resown after the break of season can be grazed until user specified destocking date, prior to reseeding. To give the resown pasture time to establish, the paddocks remain destocked until a user specified restock date.

Parameters:

  • i_destock_date_zt – Date of destocking prior to reseeding.

  • i_restock_date_zt – Date of restocking post reseeding.

  • i_destock_foo_zt – Level of FOO at destocking prior to reseeding.

  • i_restock_grn_propn_t – Proportion of the FOO that is green when pastures are restocked after reseeding.

  • resown_rt – Boolean array denoting which rotations phases are resown.

  • i_restock_fooscalar_lt – Scalar for FOO between LMUs when pastures are restocked after reseeding.

  • i_restock_foo_arable_t – FOO at restocking on the arable area of the resown pastures.

  • dry_decay_period_p6zt – decline in dry foo for each period.

  • i_fxg_foo_op6lzt – each level of starting FOO used in conjunction with PGR.

  • c_fxg_a_op6lzt – coefficient a for the FOO/growth/grazing variables. PGR = a + b FOO

  • c_fxg_b_op6lzt – coefficient b for the FOO/growth/grazing variables. PGR = a + b FOO

  • i_grn_senesce_eos_p6zt – proportion of green feed that senesces in period (due to a water deficit or completing life cycle).

  • grn_senesce_startfoo_p6zt – proportion of start foo that senesces during the period (due to leaf drop).

  • grn_senesce_pgrcons_p6zt – change of senescence over the period (due to growth and consumption).

  • max_germination_flzt – maximum germination of annual pasture across all rotations.

Returns:

Change in FOO due to reseeding pasture.