Impacts of Land Use Change - Pastoral Farming

4.0 PASTORAL FARMING - STATISTICS AND FARM MODEL DEVELOPMENT

_______________________________________________________________________

Key points

Approximately 60% of the total land area of the District was in primary productive use, predominantly farming (48%).

Farm woodlots accounted for less than 2% of the total farmed area.

There were slightly less than 300 pastoral farms in 1993. These carried 1.5 million stock units and employed 684 persons (including owner operators).

Two model farms were developed reflecting average farms for their respective classes. The large model farm represented 60 farms carrying on average 13 925 LSU per farm. The small model farm represented approximately 230 farms carrying on average 3568 LSU per farm.

4.1 Source of Statistics

All historical agricultural statistics presented or referenced in this report are derived from Statistics New Zealand publications unless otherwise stated. These statistics are gathered annually through an agricultural census of farmers conducted shortly after 30 June in each year, all data provided being as at that date (i.e. 1981 statistics for livestock relate to numbers on hand at 30 June 1981). Statistics relating to female livestock mated/used for breeding purposes refer to livestock mated in the previous statistical year for the production of offspring in the statistical year concerned. These statistics are not 100% accurate for any one region owing to farmer inaccuracies and uncompleted census returns. However, they do provide the best data set available for the WDC area, and therefore have been used for the review of pastoral land use and production in the 1981-1993 period. It is noted that some regional boundaries on which statistics are prepared changed in 1990. However, the WDC area did not change significantly, and a historical review can be undertaken without adjustment to the statistics.

The variability of the average annual statistics is quite large and does give cause for concern. Specific concerns include recent data for plantation areas, the variability of area in grass, and the decline in both the total occupied area and the area classified as 'other land'. The variability between years and between sources of data relating to employment statistics is also of concern. In the absence of an alternative data source, these statistics are the best available. However, the study makes some small departures from them in instances where the authors consider that a better estimate can be made (e.g. average stocking rate per hectare, as discussed in subsection 4.5).

New technology in remote sensing and GIS may in future provide more accurate and consistent data, but was prohibitively expensive for this study.

4.2 Land use in Wairoa District Council area 1981-1993

The Wairoa District Council Area is described as predominantly steep to very steep hill country, much of which is currently subject to erosion damage. Approximately 48% of the total land area is in grass, and 60% in use. The Land Use Capability (LUC) classes for the area have been mapped by Landcare Research (see Map 2 and Table 3.5). Land above class (5) is hill country and it is mainly class (6) and (7) areas on which land-use changes from pastoral to forestry use are occurring.

A summary of WDC land-use statistics is presented in Table 4.1. Important factors associated with these statistics are summarised as:

A farm is defined for statistical purposes as any area of land used for or potentially usable for horticulture, cropping, livestock or exotic forestry operations. In the WDC area, the number of holdings and total area in use peaked in 1986, and then declined to the lowest level in the survey period in 1992/93.
The total land area encompassed by farms steadily decreased over the period to 1993. It is noted that the category 'other land' also declined over the period. Other land includes conservation plantings and indigenous vegetation (as well as horticulture and non-productive uses), and it is possible that census respondents are understating these areas as they are no longer judged potentially usable for pastoral farming purposes. This conclusion is supported by the comparatively static area of the District reported as the 'grassed area' (which includes grassland, tussock/danthonia and lucerne).
The most important land-use statistic is the grassed area. This is the area used for pastoral farming, and peaked in 1982 (following uptake of Government-funded land development incentive schemes). The grassed area reported has remained relatively static during the period (1981 to 1991) after allowing for minor fluctuations in the total number of census returns annually. The decline between 1991 and 1993 is an expected trend arising as a result of increased afforestation of pastoral areas during that period. The fluctuation recorded in 1988 may have resulted from the aftermath of Cyclone Bola, respondents taking into account the pastoral damage incurred. For purposes of this study, the statistic reported for the 1993 year of 194 149 hectares of grassland in the WDC area is used as the starting point for analysis.
The average area and average grassed area per holding are of general interest only in that the total number of holdings includes smallholdings, forests and horticultural and cropping units as well as traditional pastoral farms. The range about this average is probably greater than 100% (see Table 4.1 for the average areas per holding over the period 1981 to 1993).
The area recorded as plantations relates to exotic trees planted for timber, and includes exotic species shelter belts planted in three rows or more. Statistics indicate continuing forest development until 1987. Between 1988 and 1991 the plantation area remained relatively static (a factor predominantly associated with high interest rates and changes to forestry taxation occurring at that time).

Table 4.1 Summary of land-use trends in the Wairoa District Council Area 1981-1993

Year Ended 30 June	1981	1982	1983	1984
Number of holdings	472	481	502	496
Total area	282 792	280 753	282 808	284 185
Grassed area	197 361	207 107	206 106	199 213
Plantations	17 118	18 831	20 676	23 206
Other land	68 313	54 815	56 026	61 766

Av. holding area	599.1	583.7	563.4	573.0
Av. grass area/holding	418.1	430.6	410.6	401.6

Year Ended 30 June	1985	1986	1987	1988
Number of holdings	502	507	493	477
Total area	283 674	283 955	279 106	277 116
Grassed area	193 199	194 098	195 417	189 742
Plantations	24 490	26 032	28 011	28 348
Other land	65 985	63 825	55 678	59 026

Av. holding area	565.1	560.1	566.1	581.0
Av. grass area/holding	384.9	382.8	396.4	397.8


Year Ended 30 June	1989	1990	1991	1992
Number of holdings	475	453	447	441
Total area	264 862	264 580	266 198	253 113
Grassed area	194 072	195 083	198 686	191 388
Plantations	26 239	26 804	28 338	22 778
Other land	44 551	42 693	39 174	38 947

Av. holding area	557.6	584.1	595.5	573.9
Av. grassed area/holding	408.6	430.7	444.5	434.0

Year Ended 30 June		1993
Number of holdings		443
Total area		260 536
Grassed area		194 149
Plantations		33 419
Other land		32 968

Av. holding area		588.1
Av. grassed area/holding		438.3

Note: From an estimated 300 pastoral farms in 1993, the average grassed area per pastoral farm is calculated as 647 hectares.

Statistics for 1992 indicate a substantial decrease in plantation area, and this is difficult to explain other than through a possible non-return of census information by some forest owners. The 1993 statistics appear to be more in line with expectations, given known increases in plantings between 1991 and 1993. (The statistical data indicate an increase in the exotic forest area of 5081 ha in this two-year period.) A similar decrease in pastoral land occurred between the 1991 and 1993 years (a decrease of 4537 hectares), and a decrease of 49 674 LSUs is also recorded for this period (10.9 LSU/grassed hectare, if all related to forestry expansion).

The important statistic for this study is the 1993 grassed (pastoral) area of 194 149 hectares.
Information obtained from the MAF Quality Management division of the Ministry of Agriculture and Fisheries indicated that, at 30 June 1994, there were 289 separate pastoral farm holdings carrying sheep and/or beef cattle in the WDC region with an area greater than 5 hectares. If this figure is adjusted to 300 pastoral farms at 30 June 1993 (thus allowing for dairy farms and any specialist deer farms), the average pastoral farm area in 1993 would be as follows.

Average farm area (excluding plantations) 757 ha

Average grass area per farm 647 ha (85.5%)

(The statistic of 289 pastoral sheep and/or beef farms at 30 June 1994 was obtained from the MAF Quality Management Agribase Farm Location Project.)

The difference between the total area farmed and the total District Council area of 412 750 ha is mainly made up of urban areas, roads, very small rural holdings and DOC estates.

4.3 Total Livestock Numbers 1981-1993

Livestock numbers for specific years in the period 1981 to 1993 have been sourced from Statistics NZ. The years chosen include 1981, 1986, 1991, 1992 and 1993, and the related statistics are contained in Appendices 3.1.1 to 3.1.5. The important factors arising from these statistics are summarised below.

(i)	The total livestock units (LSUs) changed over the period as follows.

Year	Appendix	Total LSUs	LSU's/grassed ha
1981	3.1.1	1 799 944	9.12
1986	3.1.2	1 816 449	9.36
1991	3.1.3	1 552 375	7.81
1992	3.1.4	1 505 741	7.86
1993	3.1.5	1 502 701	7.74

Livestock numbers increased until the mid 1980s encouraged by Government-funded land development schemes, livestock product pricing policies (SMPs) and other subsidised assistance provided to the agricultural sector. Following the commencement of deregulation and restructuring of the NZ economy post 1984, the economic conditions in the primary sector at the farm gate level deteriorated markedly, and a decline in livestock numbers began. Main reasons for the decline included falling fertility due to severe cutbacks of fertiliser on marginal land, low product prices, high interest rates, changes to livestock taxation, Cyclone Bola damage, the 1989 drought, and, more recently, changes in land use from pastoral to forestry enterprises. This decline is expected to continue at least until 30 June 1995 for reasons of land-use change to forestry. The decrease in livestock numbers between 1986 and 1993 amounted to 17.3% in total, and is greater than the New Zealand average of about an 8% decline over that period.
Sheep and beef cattle comprise 96.7% of total livestock units at 30 June 1993, with only marginal variations to this ratio over the 1981-93 period. The sheep-to-cattle ratio has changed from 61:39 in 1986 to 54:46 in 1993, owing mainly to product price changes in favour of beef cattle, and also to farm management factors (including those associated with decreased labour requirements). In total, sheep LSUs have decreased by 27.2% and cattle LSUs by only 3.0% in the period 1986-1993. Sheep and beef cattle accounted for an average stocking rate per grassed hectare of 7.47 LSU/ha out of the total 7.74 LSU/ha in 1993.
Dairying accounts for 1.2% of LSUs at 1993, and the expansion or contraction of this small pastoral activity is unlikely to be directly affected by future land-use changes to forestry.
Deer numbers increased to a peak in 1992, accounting for 1.7% of total LSUs in that year. In 1993 a decrease in deer numbers is recorded (of 18%), this coming about as a result of falling product prices, a higher-than-usual stock carryover in the 1992 year and, to a smaller degree, through land-use change.
Goat numbers increased markedly in the mid 1980s, but significantly decreased thereafter owing to the decline in market demand for both live feral goats and their cashmere products. However, it is noted that there are much greater numbers of goats in the WDC area than are recorded as being 'farmed' in the agricultural statistics. These are feral animals, retained on many properties for reasons related to weed control rather than productive output, and are probably not considered to be farmed by some census respondents.

4.4 Farm owners and employees

Statistics NZ agricultural data (Table 4.2) identifies the number of people employed on farms in the Wairoa District area.

Factors of note include the following.

A large decrease in paid full-time employees occurred in the 1991 year, without a compensating increase in the number of self-employed workers, as national statistics may indicate.

Table 4.2 Farm employment in the Wairoa District

Year
No. of farms Working owners, lease-holders and sharemilkers No. of paid full time employees Total

1986
1989

1991

1992

1993
507
475

447

441

443
465
408

426

415

382
332
256

211

270

302
797
664

637

685

684

Agriculture Statistics do not include paid part-time employees, but data for working owners, leaseholders and sharemilkers do include part-time employment statistics.
The definition of farms includes all holdings irrespective of their use, and thus encompasses horticultural holdings, forest areas, etc. It is noted that there are approximately 300 livestock farms in the District.
In the case of family partnerships, more than one working owner per property may be recorded (e.g. in a husband-and-wife partnership).
A separate analysis of 1991 Industry Statistics indicates that a total of 609 people worked on sheep, beef and mixed farms in the district (out of a total of 753 working on all farms). It is noted that these statistics do not agree with the Agricultural Statistics for 1991.
On the basis of the employment input/output data analysed in this study and derived from WDC model farms and farmer surveys, the total number of full-time equivalents on sheep and beef farms would be about 590 workers (0.34 per 1000 LSUs, assuming an average 9 LSU/grassed ha).
While of interest, the absolute or correct number of farm workers in the District is not an important factor, as the study analyses the labour loss with land-use change on a separate basis.
Industry statistics 1991 records that 36 workers were employed in forestry and logging in the WDC area.

4.5 Describing farm businesses - development of representative farm models

Two whole-farm models have been developed to represent Wairoa pastoral farming properties. Base data has been derived from representative farms used by Agriculture New Zealand consultants in the preparation of the MAF Farm Monitoring Report. These farms are all located in the Wairoa District Council area, and are not amalgamated with the Gisborne representative farms as occurs for purposes of the MAF Farm Monitoring Report. Production indices are set at average levels, and prices and costs used are those current in the 1994/95 financial year.

From these two farm models a weighted average model has been calculated, and is used in the analysis of the impact of land-use change in section 8.0 (note that the computer software associated with this study can very quickly analyse impacts using only the small or large farm model data). The two models developed are summarised as follows.

(i)	Large farm model

	Represents approximately 20% of farms and 53% of the total pastoral (effective) area.

Farm area:		1654 hectares (effective)
Livestock units:	Sheep	8 603 (61.8%)
Cattle		5 322 (38.2%)
Total		13 925

Average stocking rate:	8.42	LSU/effective hectare
Lambing percentage:		88%
Calving percentage:		90%

(ii)	Small farm model

	Represents approximately 80% of farms and 47% of the total pastoral (effective) area.

Farm area:		367.5 hectares (effective)
Livestock units:	Sheep	2 202	(61.7%)
	Cattle	1 366	(38.3%)
	Total	3 568	LSUs

Average stocking rate:	9.7 LSU/effective hectare
Lambing percentage:	109%
Calving percentage:	88%

The weighting given the two models (20% and 80% of total farms) encompasses a total grassed area of 187 500 hectares. This correlates relatively well with the Statistics New Zealand total of 194 149 grassed hectares in 1993 when considering the relatively large farms sold to forestry between June 1993 and December 1994, and grassed areas in other uses. (Note: a ratio of 21% large farms and 79% small farms would correspond to a grassed area of 191 300 hectares.)

Both models are restricted to sheep and cattle farming. Other livestock types in the district account for only 3.3% of total livestock units:

Dairy		1.1%
Deer		0.8%
Goats		0.9%
Total		3.3%

(1993 Agricultural Statistics)

Dairy farms are unlikely to be converted to forestry, and are therefore excluded from analysis. Goats are operated mainly as a feral enterprise for weed control, and deer are farmed on a minority of properties, with total numbers including breeding stock only 12 300 in 1993. However, it is noted that the analyses of farm models include all costs of deer and goats in the budgets prepared, and the item 'other farm income' will include income arising from these livestock types (and is about 1% of total farm income).

The farm models are analysed with constant stock numbers (status quo state). No off-farm income is included.

4.5.1 Model Development

The models' physical and financial details were developed on the basis of 1000 livestock units, with total farm costs directly apportioned on a cost per LSU basis. Farm income was calculated separately for sheep and cattle through livestock reconciliation analysis, also on the basis of 1000 LSU groups. The results of these two model analyses are then amalgamated into a weighted average model.

The use of a 1000 LSU base permits easier analysis of the effect of differing levels of reduction in livestock numbers at both the whole-farm and marginal levels, given that the model farms from which they were developed are 'averages' and don't reflect the range of farm sizes and stock carried within each of the two groupings. However, the 1000 LSU models can be grossed up to actual farm size by simple multiplication. Further, the effects of differing stocking rates and effective farm areas can also be analysed using the 1000 LSU base methodology.

The sheep-to-beef ratios of approximately 62:38 as contained in the derived models do not reflect the ratios recorded by Statistics NZ in their 1993 Agricultural Statistics. The official statistics record a ratio of 54:46. In order to more closely align the models to actual statistics, the sheep-to-cattle ratio was adjusted to the statistical levels for income assessment purposes and for calculating livestock numbers of each type available for sale. (The actual adjustment made was a ratio of 53.5:46.5 sheep to cattle, and was based on provisional 1993 statistics made available before their official release.) Costs other than those directly related to livestock (e.g. shearing, purchases and transport) were not adjusted. To indicate that this adjustment had an insignificant effect on income levels, the following summarised figures are provided.

	Small farms	Large farms
Income from sheep and wool (nett of
purchases, transport and shearing)	$35.73/LSU	$28.34/LSU

Income from cattle (nett of
purchases and transport)	$34.54/LSU	$27.50/LSU

Difference between sheep
and cattle	$1.19/LSU	$0.84/LSU

Effect on income of changing
the ratio	-$0.10/LSU	-$0.07/LSU

The effects on income per LSU (decrease of 10 cents and 7 cents per LSU) are insignificant. This is a necessary adjustment in order to better estimate the reductions in stock numbers available for slaughter by livestock type as land-use change occurs.

Full details of the physical and financial structure of the two models and a weighted average of the two (based on 1000 LSU groups) are contained in Appendices 3.2 and 3.3. These appendix tables are referenced below for ease of access.

			Small farms	Large farms

Details of assumptions made			App 3.2	App 3.2
Whole farm stock numbers			3.2.1	3.2.7
Sheep reconciliation per 1000 sheep LSU			3.2.2	3.2.8
Wool account per 1000 sheep LSU			3.2.3	3.2.9
Cattle reconciliation per 1000 cattle LSU			3.2.4	3.2.10
Financial budget per 1000 total LSU			3.2.5	3.2.11
Stock numbers per 1000 total LSU (including sales data and destination)			3.2.6	3.2.12

	Weighted	average farm

Weighted average model budget	App.	3.3
Stock nos per 1000 total LSU		3.3.1
(including sales data and destination)
Average farm capital repayment, income tax, and disposable surplus		3.3.2

Livestock salvage value		3.4

4.5.2 Summary of average farm model indices

This subsection briefly summarises the main indices in the weighted average farm model on a per 1000 LSU basis.

Total sheep		586
Total cattle		90

Sheep sold to slaughter		250
Cattle sold to slaughter		16.7

Sheep sold as store stock		99
Cattle sold as store stock		16.2

Gross income per LSU ($)		34.93
Farm expenditure per LSU ($)		24.38
Cash Farm Income per LSU ($)		10.55

Disposable surplus (after capital
repayments and income tax) per LSU ($)	6.08

Full-time labour equivalents on farm	0.30	per 1000 LSU
Full-time labour equivalents shearing	0.04	per 1000 LSU

Lamb equivalents slaughtered	372	per 1000 LSU

Wages paid per 1000 LSU ($)	5740

4.5.3 Stocking rates and total effective area

The model farm developed for this study has an average stocking rate calculated as 9 LSU per grassed hectare. As previously noted, this is higher than the estimated Statistics NZ level of 7.7 LSU per hectare. A number of other estimates of average stocking rate per hectare were obtained, and all are summarised below.

Wairoa Survey (91 farms)
Large farms	8.9	LSU/grassed ha
Small farms	10.9	LSU/grassed ha
Average	9.8	LSU/grassed ha

Valuation NZ (farms sold)	7.6	LSU/grassed ha
Farm Monitoring	9.0	LSU/grassed ha
Agricultural Census (statistics)	7.7	LSU/grassed ha

In addition, the question of average District stocking rate was discussed with a number of Wairoa farmers, whose estimates varied between 9.0 and 9.5 LSU per grassed hectare.

Given the wide variation between sources, and the wish to avoid underestimating the agricultural impact of land-use change to forestry, the average stocking rate of 9 LSU/grassed hectare as calculated in the model has been accepted for the purpose of this study.

A further adjustment is required to account for land on farms not currently used for farming, but which is likely to be planted in forest at the same time as the pastoral areas. Farmers surveyed in the District estimated their ineffective area to be 18% of the farm. This aligns with official statistics data associated with 'other land' of 15%. Thus, for purposes of this study, it is assumed that for every 1000 hectares of planted forest, 85% (850 hectares) is currently in pastoral use and 15% (150 hectares) in clearable scrub. On an adjusted basis, the average stocking rate over the total area reduces to 7.7 LSU per hectare. This is the level of reduction of livestock numbers per hectare of planted forest used in this study. (Variations about this average are analysed and discussed in section 8.3.1, which evaluates the impacts on the agricultural sector of land-use change.)

It is also noted that the stocking rate calculated using Statistics New Zealand livestock numbers is not calculated on a class-by-class basis. The livestock numbers are very aggregated, and the average stock units applied to each group may be marginally lower than actual for some groups (particularly cattle). This would contribute to a lower stocking rate than actual being calculated, and would explain - at least in part - the variation between the District average and that modelled.