POLLINATORS POLLINATION AND FOOD PRODUCTION

THE ASSESSMENT REPORT ON POLLINATORS, POLLINATION AND FOOD PRODUCTION
determine land cover, which can be logistically impeditive
at national and global scales. Finally, the spatially explicit
information available for valuation is usually obtained from
censuses and aggregated at municipality, state or national
levels by national bureaus of statistics, a procedure that per
se causes some loss of information (Vermaat et al., 2005).
Furthermore, increasing the spatial scale means using data
collected by different researchers or agencies using distinct
protocols, which frequently are not directly comparable
(Lautenbach et al., 2012, Leonhardt et al., 2013). By
contrast, GIS data are gathered by pixel or cell. Inserting
such reported administrative data (crop type, production
area, yields) into mapped units frequently involves several
calculation steps and many assumptions (Monfreda et al.,
2008) that may decrease estimate accuracy at large scales.
Some studies used GIS to calculate pollination service
value at the local (including landscape) scale (Lonsdorf
et al., 2009, Ricketts and Lonsdorf, 2013), but the most
comprehensive attempt to map pollination benefits at the
global scale was conducted by Lautenbach et al. (2012).
These authors used the geographic distribution of crop
areas and crop yields made by Monfreda et al. (2008) with
latitude-longitude grid cells of 5 minutes x 5 minutes made
possible by the use of the use of satellite. Despite the fine
resolution (approximately 10 km x 10 km at the equator),
this approach has some limitations, because the distribution
of yield statistics into raster cells (i.e., a grid containing
values that represent information) eliminates some crops
for such cells (Lautenbach et al., 2012). Thus, accurate
estimates of pollination benefits at national and global scales
can be strongly influenced by evolving low-cost satellite
technology to distinguish different crop types, and countries’
adoption of standardized frameworks to collect crop data
(e.g., Vaissière et al., 2011; Ne’eman et al., 2010).
An alternative to the lack of detailed data for pollination
valuation at larger scales is the use of benefit or value
transfer-based mapping (Troy and Wilson, 2006; Eigenbrod
et al., 2010). This procedure consists of determining the
value of the pollination service for a given crop type at
a local scale, and using this as a proxy to estimate the
value of the same crop type at other locations or at the
regional or national scale. However, this procedure has
several limitations related to the lack of correspondence
between locations (Troy and Wilson, 2006; Plummer, 2009;
Eigenbrod et al., 2010), leading to generalization errors
that can only be overcome with improved spatial data and
increasing the number of local replicates used for calculating
the value of pollination services. A review of spatially
explicit tools for pollination service valuation is available in
Chapter 6 (see also a summary in Table 4.7), and details
on geographic differences on pollinator availability, efficiency
and dependency are given in Chapter 3.
3.3.2.2 Landscape design
The general effects of landscape design (spatial heterogeneity,
connectivity, isolation, and proportion of natural habitats) on
pollination by managed and wild species are addressed in
Chapters 2, 3 and 6. Several studies have demonstrated
positive effects of the pollinator habitats maintenance on
agricultural yield (Ricketts et al., 2008; Garibaldi et al., 2011;
Ferreira et al., 2013; Kennedy et al., 2013). However, sparing
natural vegetation in a given farm incurs an opportunity
cost from not using that area for crop production or other
TABLE 4.7
Summary of factors that affect valuation methods across scales and the tools to apprehend such effects.
Temporal scale
Rationale: different
demands across
institutional levels
(e.g., farmers x
government)
Spatial scale
Rationale: micro vs.
macroeconomics
valuation
Factors affecting valuation across scales Tools to apprehend scale effects Examples
- Price dynamics
- Production effect
- Discount rate
- Availability of long term data sets
- Loss of data quality at large scales
- Landscape design
- Time series analysis
- Scenarios
- GIS techniques
- Spatially-explicit frameworks
- Regression methods 1
- Stochastic simulations 2
- Forecasting models 3
- SRES 4
- MEA 5
- ALARM 6
- UK NEA 7
- Maps 8
- Landscape metrics
(fragmentation, connectivity) 9
- Polyscape 10
- InVEST 11
- ARIES 12
- Envision 13
- Markovian models 14
- Niche modeling 15
235
4. ECONOMIC VALUATION OF POLLINATOR GAINS
AND LOSSES
References
[1] Gordo and Sanz, 2006; Aizen et al., 2008; 2009; Aizen and Harder, 2009; Lautenbach et al., 2012; Bartomeus et al., 2013; Leonhardt et al., 2013; [2] Keitt, 2009; [3] Clark
et al., 2001; [4] Nakicenovic et al., 2000; [5] MEA, 2005; [6] Spangenberg, 2007 ; Gallai et al., 2009b; Spangenberg et al., 2012; Settele et al., 2012; [7] Haines-Young et al.,
2014; [8] Schulp and Alkemade, 2011; Lonsdorf et al., 2009; Lautenbach et al., 2012; Kennedy et al., 2013; Ricketts and Lonsdorf, 2013; [9] Ricketts et al., 2004; Garibaldi et
al., 2011; Ferreira et al., 2013; Kennedy et al., 2013; [10] Jackson et al., 2013; [11] Lonsdorf et al., 2009; Nelson et al., 2009; Tallis et al., 2011; Ricketts and Lonsdorf, 2013;
Zulian et al., 2013; [12] Bagstad et al., 2011; Jackson et al., 2013; [13] Bolte et al., 2007; Hulse et al., 2008; [14] Satake et al., 2008; [15] Settele et al., 2008; Giannini et al.,
2013; Polce et al., 2014.