4619.0.55.002 - Land Management Practices in the Great Barrier Reef Catchments, Experimental Estimates, 2008-09  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 23/04/2010  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All
  • Explanatory Notes



1 This release presents experimental estimates from the ABS Land Management Practices Survey of the Great Barrier Reef catchments for the 12 months ending 30 June 2009.

2 This survey collected information from the land owner and/or manager on a range of land management practices applied to their agricultural land. These were soil testing, fertiliser use, chemical use (including weed, pest and disease control), riparian management, surface water management and irrigation water management. Information was also collected on practices specific to a particular agricultural activity. The majority of estimates measure the proportion of holdings undertaking a particular management practice. There are also estimates showing weight of fertiliser applied as well as the area over which selected land management practices have been applied.

3 The survey methodology differs from that of most previous ABS surveys that collected information about agricultural activity or land management practices in this region. This will affect comparisons with past data. In particular, this survey employed a land based methodology to collect information from land holdings (see paragraphs 9 to 12). Previous ABS surveys have used a business methodology to collect information from agricultural businesses. The scope and coverage of this survey also differs from previous ABS surveys (see paragraphs 4 to 8).


4 This survey covers land parcels in 28 of the 30 catchments draining into the Great Barrier Reef lagoon. The two catchments in the Cape York NRM region have been excluded because they have minimal agricultural activity.

5 Information was collected for all land areas where the primary land use was classified to one of four agricultural activities (sugar cane growing, horticulture, broadacre cropping or beef cattle grazing). Other agricultural activities, or land where agriculture was undertaken as a secondary land use activity, were not in scope.

6 Land used for agricultural activity (primarily grazing), located in Queensland Parks and Wildlife Services estate (forests, national parks and reserves etc) was excluded from the survey as access arrangements to this land are highly regulated, and the potential for the land manager to influence or determine land management practices is minimal.

7 The survey reference period is the 12 months ending 30 June 2009. That is, all reported practices were carried out within this time period.

8 As the total contribution of small land holders to agricultural production in a region can be significant, especially with respect to horticultural activity, there was no minimum holding size cut off for selection in the survey.


9 The statistical unit from which information about land use practices was reported is the 'holding'. A 'holding' represents all the land parcels under common ownership/management within an individual catchment.

10 Queensland Valuation and Sales Data (QVAS), and the Digital Cadastre Database (DCDB), were used in combination to define spatial land parcel boundaries, and to source information about each land parcel including ownership details, contact information, land use information and a common ownership indicator (property indicator). A number of other information resources were also used to augment the available information about land parcels.

11 Information about the land area of the holding and a breakdown of land use of the holding was also collected in the survey to validate land ownership/management data from QVAS. The information collected on agricultural land use has been published as part of the survey output.

12 This survey is the third ABS survey to use a land parcel based methodology. A previous survey based on land parcels was conducted over part of the Great Barrier Reef catchment. Results from that survey are available in Land Management: Fitzroy and Livingstone Shires Queensland, 2004-2005 (cat. no. 4651.0). Although the 'holding' was also used as the statistical unit in that survey, the scope of this survey was broader.


13 The sample was stratified by agricultural commodity, area of farm and catchment to achieve low relative standard errors for estimates of practices for the primary agricultural activity (commodity) within each catchment. A sample of 4,502 holdings was selected. The 4,502 holdings accounted for approximately 43% of the in-scope land area and approximately 23% of the total population of holdings.

14 The sample design assumed a minimum response rate of 85% and for the responses received to cover 90% of land surveyed. The collection process achieved a response rate of 92% of forms which represented 89% of the area of the selected holdings. After accounting for holdings which, after collection, were deemed to be out of scope, the final response rate was 90%.


15 The survey reports estimates for Catchment and Natural Resource Management (NRM) regions. The 28 catchments in the survey aggregate to five NRM regions covering the majority of the Great Barrier Reef Catchments and NRM regions have been developed by the Australian and State Governments to manage priority resource management issues.


16 A land manager may be responsible for land parcels in multiple catchments. As noted in paragraph 9, this would result in separate reporting units (holdings) in each catchment, with the land manager potentially providing multiple survey responses (for those holdings selected).

17 Where holdings have reported more than one activity (for example sugar cane and beef cattle grazing), it is not always possible to determine if a reported land management practice is related to the management of a specific commodity. For example, some proportion of the population of sugar cane holdings may report the construction of cattle ramps. This is because some holdings with sugar cane also manage cattle on their holding.

18 Some practices (for example herbicide application) could be carried out multiple times on the same piece of land in the survey reference period. More than one practice may also be carried out on the one piece of land (for example when managing broadacre crop stubble, an area of stubble could be grazed, following which the remaining stubble is then incorporated into the soil). In these cases, the area of land where the practice was applied will be counted multiple times for each application or each practice undertaken.

19 Practices are strongly influenced by climatic and other external conditions (eg price and availability of inputs). This is particularly relevant for comparing data with any future iterations of this survey. As an example, the application of irrigation water management practices will be strongly influenced by the availability of irrigation water in the survey period.

20 Holdings in the survey may have significant land area dedicated to conservation purposes, or include agriculturally unproductive areas. Such land has not been included when defining the total area of land being used for agricultural production. However, this information has been collected in the survey, and is reported in other outputs from this survey.

21 The reported figure for 'total area of catchment' has been obtained from using the digital layer of Australian River Basins provided by Geoscience Australia. The resulting figures may not match estimates of area of catchments obtained from other sources. It should also be noted that 'total area of catchment' does not include off-shore islands.


22 The estimates in this release are based on information obtained from a sample drawn from all of the holdings in scope of the survey, and are subject to sampling variability; that is, they may differ from figures that would have been produced if all holdings had been included in the survey.

23 One measure of sampling error is given by the standard error (SE), which indicates the extent to which an estimate might have varied by chance because only a sample of holdings was included. There are about two chances in three that a sample estimate will differ by less than one SE from the figure that would have been obtained if all holdings had been included, and approximately nineteen chances in twenty that the difference will be less than two SEs.

24 In this release, the 'sampling' variability of the estimates is measured by the relative standard error (RSE), which is obtained by expressing the SE as a percentage of the estimate to which it refers.

25 Estimates in the data cubes included with this release are annotated with a caution when the RSE of the estimate is greater than 10%. Where the RSE of an estimate included in this release falls in the range of 10% to less than 25% the estimate should be used with caution as it is subject to sampling variability too high for some purposes. Where the RSE of an estimate is 25% to 50% the estimate should be also be used with caution as it is subject to sampling variability too high for most practical purposes. Where the RSE of an estimate exceeds 50% the sampling variability causes the estimate to be considered too unreliable for general use.


26 Non-sampling errors are inaccuracies that occur because of non-response, errors in reporting by respondents and interviewers, and errors made in coding and processing data. These inaccuracies may occur in any enumeration, whether it be a full count such as a census or a sample. Every effort is made to reduce the non-sampling error to a minimum by careful design of questionnaires, intensive training and efficient processing procedures.

27 Likely potential sources of non-sampling error identified in this survey include:

  • Errors made when amalgamating land parcels to create the statistical unit (holding) due to limitations of land parcel ownership data.
  • Coverage errors (parcels wrongly included or excluded from the population), that resulted from land parcels with incorrect, incomplete or outdated ownership or land use information.
  • Land parcels that span the boundaries of multiple catchments were assigned to an individual catchment in which the largest percentage of the area of the land parcel was located, thus creating small discrepancies in the estimated area under management within a catchment.
  • Errors resulting from respondents failing to report for parcels of land under their ownership/management during the survey reference period. The converse was also true, that some respondents may have reported for land that was not owned or managed by them during the survey reference period.
  • Errors resulting from clerical estimation of selected non-responding units.
  • Errors in rates of sensitive practices resulting from non-response bias (i.e where the characteristics of non-respondents differ from that of respondents).

28 When producing the experimental estimates featured in this release, the weights of the sampled units were not adjusted to represent the total number of units in the sub-populations of interest (Burnett sub catchments and land under crop/grazing by size). This may introduce some bias into the estimates of proportions undertaking certain practices.


29 A range of recent environment publications produced by the ABS, include:

30 Other relevant ABS environment publications include: