Calculation of diversification indicators and other covariates

Summary

Indicators

Indicator	Data	Format
perimeter and area of the field	RPG*	vectoriel
mean field size within buffer	RPG*	vectoriel
crop rotation (N-4:N)	RPG*	vectoriel
hedgerows around field	RPG* + Liu 2023	vectoriel
% land cover within buffer	RPG* + CLCplus	raster 10m
edge density	Not available	NA

Datasets

Registre Parcellaire Graphique (RPG): annual field crop data for the period 2015-2024 available for France on IGN website: https://cartes.gouv.fr/rechercher-une-donnee/dataset/IGNF_RPG. The dataset differentiates 366 crop categories
Geodienste - nutzungsflaechen (Nutzung): Swiss equivalent of the RPG, not entirely open yet but raw data provided by Selma Cadot: https://geodienste.ch/services/lwb_nutzungsflaechen. The dataset differentiates 146 crop categories
Corine Land Cover plus (CLCplus): land cover at European scale with 10m resolution and 11 basic land cover classes, available on Corine Land Monitoring Service https://land.copernicus.eu/en/products/clc-backbone.
Liu et al. (2023): European tree cover and biomass map at 30m resolution for Europe https://zenodo.org/records/8154445 derived from 3m resolution PlanetScope imagery of 2019.
Land cover class harmonization: list all classes from RPG, Nutzung and CLCplus and how to categorize them. This file should be completed by expert and customized for the project objectives.

Figure 1: Number of observations per year and per project

Table 1: Number of observations per year and per project

	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024	TOTAL
BACCHUS_OPERA	0	0	0	0	40	38	40	40	38	38	38	272
BIOMHE	0	0	0	0	0	0	40	0	0	0	0	40
BISCO	0	0	0	27	0	0	0	0	0	0	0	27
DIVAG	0	0	0	0	0	40	0	0	0	0	0	40
DURUM_MIX_GM	0	0	0	0	226	0	0	0	0	0	0	226
EXCLU_BVD	0	0	0	0	0	0	0	7	12	16	0	35
FRAMEwork_BVD	0	0	0	0	0	0	0	36	0	0	0	36
lepibats	0	0	0	0	0	0	0	33	0	0	0	33
muesli	0	0	29	0	0	0	0	0	0	0	0	29
OSCAR	0	0	0	0	21	44	50	80	100	102	115	512
PestiRed	0	0	0	0	0	0	62	68	66	63	0	259
SEBIOPAG_BVD	0	0	0	0	0	0	20	21	20	0	0	61
SEBIOPAG_Plaine de Dijon	20	20	20	20	20	20	20	20	20	20	20	220
SEBIOPAG_VcG	20	19	17	17	17	17	17	17	17	17	0	175
SEBIOPAG_ZAAr	0	0	0	0	20	0	0	17	0	16	0	53
SERIPAGE	0	0	9	0	0	0	0	0	0	0	0	9
TOTAL	40	39	75	64	344	159	249	339	273	272	173	2027

Figure 2: Map of field observations

More detailled overview of the field observations can be found here

Identification of the crop field

Because of data availability, we will only focus on the observations made in the period 2015-2024 with field coordinates. The project DURUM_MIX_GM has only one coordinates leading to the entrance of the Institut Agro-Montpellier, without registered crop field within 1500m from the coordinates so it was also excluded.

We identified the crop field from RPG or Nutzung dataset corresponding to the observations based on the coordinates and the year of the sampling.

Table 2: Number of observations in fields from RPG

	Nobs	in_RPG	Perc
BACCHUS_OPERA	272	224	82.35
BIOMHE	40	39	97.50
BISCO	27	26	96.30
DIVAG	40	40	100.00
EXCLU_BVD	35	34	97.14
FRAMEwork_BVD	36	30	83.33
lepibats	33	30	90.91
muesli	29	29	100.00
OSCAR	512	459	89.65
PestiRed	203	185	91.13
SEBIOPAG_BVD	61	52	85.25
SEBIOPAG_Plaine de Dijon	200	200	100.00
SEBIOPAG_VcG	155	152	98.06
SEBIOPAG_ZAAr	53	53	100.00
SERIPAGE	9	9	100.00

In total, 92 % of the fields observations are covered by national crop data.

There are disparities among projects with BACCHUS_OPERA, FRAMEwork_BVD and SEBIOPAG_BVD having a lower coverage than 90%.

Figure 3: Case of BACCHUS_OPERA observations in 2023 with overlaied RPG

Remarks:

It is important to know and acknowledge that RPG is incomplete.

Field size

We calculated the area and the perimeter of the crop fields corresponding to the samplings.

boxplot
table

Figure 4: Field area in ha per project. The dashed line show the median area.

Table 3: Summary statistics per project of the area (in ha) of crop fields

	Min.	1st Qu.	Median	Mean	3rd Qu.	Max.
BACCHUS_OPERA	0.26	1.06	2.38	4.57	6.16	39.27
BIOMHE	0.68	1.59	3.69	4.48	6.29	13.99
BISCO	0.50	1.10	1.71	3.22	3.15	23.69
DIVAG	0.97	2.21	2.98	3.17	4.02	6.01
EXCLU_BVD	0.35	0.71	1.29	2.23	3.81	6.56
FRAMEwork_BVD	0.36	0.56	1.36	3.83	5.01	18.15
lepibats	1.40	2.77	6.18	11.89	13.96	55.47
muesli	0.41	1.90	3.84	5.46	7.61	34.05
OSCAR	0.22	0.48	1.01	1.73	1.90	21.35
PestiRed	0.10	0.96	1.08	1.36	1.43	5.19
SEBIOPAG_BVD	0.36	0.83	3.18	5.65	5.20	29.02
SEBIOPAG_Plaine de Dijon	0.53	5.11	6.83	7.46	8.92	17.82
SEBIOPAG_VcG	0.85	2.90	4.27	5.56	7.00	24.06
SEBIOPAG_ZAAr	1.21	3.16	4.73	4.96	6.94	16.22
SERIPAGE	1.55	2.19	3.51	4.04	5.19	7.60

boxplot
table

Figure 5: Field perimeter in m per project. The dashed line show the median perimeter.

Table 4: Summary statistics per project of the perimeter (in m) of crop fields

	Min.	1st Qu.	Median	Mean	3rd Qu.	Max.
BACCHUS_OPERA	231.25	436.61	802.36	1091.35	1334.82	6803.79
BIOMHE	354.03	592.12	835.73	907.71	1072.08	1807.17
BISCO	307.54	503.97	593.59	743.08	846.90	2550.45
DIVAG	485.39	654.90	799.78	816.22	937.89	1275.08
EXCLU_BVD	236.87	448.13	552.11	737.27	925.69	1853.08
FRAMEwork_BVD	259.83	456.27	514.11	1009.81	1130.31	4940.46
lepibats	488.44	824.10	1260.50	1789.15	2670.28	6828.99
muesli	257.24	608.61	939.47	1060.52	1338.17	4021.15
OSCAR	200.02	371.36	493.46	589.05	749.38	2728.54
PestiRed	299.90	461.58	514.09	555.85	628.35	1024.14
SEBIOPAG_BVD	259.83	464.28	762.95	1129.45	1205.19	4980.60
SEBIOPAG_Plaine de Dijon	393.53	1075.19	1240.09	1297.87	1558.72	2220.91
SEBIOPAG_VcG	377.85	815.28	1171.63	1202.13	1505.30	4249.39
SEBIOPAG_ZAAr	451.97	793.52	949.02	1061.93	1207.41	3466.03
SERIPAGE	494.30	750.53	776.57	928.22	1227.43	1466.38

Figure 6: Relation between area and perimeter (in log scale)

There is a strong relation between area and perimeter (Figure 6). In median, field size is 2.1 ha and field perimeter is 727m.

Outliers

Figure 7: Field with small area and large perimeter

Figure 8: Field with large area

Remarks:

Some fields are defines as a strip (large perimeter, small area), e.g. Figure 7
Some fields include many sub-plots, e.g. Figure 8.

Field size within buffer

Using the coordinates of the sampling sites, we calculated the average area of all crop fields within a buffer (500m, 1000m, and 1500m).

Table 5: Summary statistics of the field area (in ha) within different buffer size

	B_500m	B_1000m	B_1500m
Min.	0.22	0.26	0.30
1st Qu.	1.28	1.27	1.31
Median	2.12	2.07	2.08
Mean	2.95	2.84	2.65
3rd Qu.	3.70	3.20	3.06
Max.	35.09	153.56	110.46
NA’s	2.00	1.00	0.00

We see that some observations don’t have crop field within 500m (N=2). These observations (listed in Table 6) would need to be checked and ensure that they are close to an agricultural field.

Table 6: Observations with no fields within a 500m buffer.

	Study_ID	Plot_ID	Year
181	OSCAR	33_2011_00002	2018
481	OSCAR	11_2023_00004	2023

Figure 9: Correlation among field areas per buffer size

Figure 10: Average field size with buffer of 500m

Figure 11: Average field size with buffer of 1000m

Figure 12: Average field size with buffer of 1500m

Outliers

Figure 13: Highest average field size within 1500m buffer

Outlier due to large SPL = pasture land of 1320 ha.

Figure 14: Lowest average field size within 1500m buffer

Figure 15: Large differences between 1000 and 1500m buffer

Figure 16: Large differences between 500 and 1000m buffer

Remarks:

The classic issue of average. Using median would solve the issue with high outliers or we should make sure to remove large pastoral plots.

Crop rotation (N-4:N)

barplot
table

Figure 17: Length of the land cover time series per project.

Table 7: Number of observations per length of crop rotation data (in years) available

	0	1	2	3	4	5
BACCHUS_OPERA	35	6	11	9	4	207
BIOMHE	1	0	1	0	0	38
BISCO	1	0	0	26	0	0
DIVAG	0	0	0	0	0	40
EXCLU_BVD	0	0	1	1	1	32
FRAMEwork_BVD	6	0	0	2	0	28
lepibats	2	1	4	0	1	25
muesli	0	0	29	0	0	0
OSCAR	34	7	23	39	24	385
PestiRed	18	31	45	49	25	35
SEBIOPAG_BVD	9	0	0	1	0	51
SEBIOPAG_Plaine de Dijon	0	20	20	20	0	140
SEBIOPAG_VcG	1	18	17	17	2	100
SEBIOPAG_ZAAr	0	0	0	0	0	53
SERIPAGE	0	0	9	0	0	0

The number of observations in the period N-4:N in Table 7 is a results of the number of observations per year (Table 1). Most observations have 5 years of data on crop rotation (67%, N=1134).

Let’s have a look at the crop rotations over the 5-year period (N:N-4).

barplot
table

Figure 18: Number of different crop groups cultivated in the period N:N-4. Only observations with complete information on land cover for the 5 years are included.

Table 8: Number of different crops grown in 5 year period

	1	2	3	4
BACCHUS_OPERA	203.00	4.00	0.00	0.00
BIOMHE	2.00	11.00	17.00	8.00
DIVAG	2.00	9.00	25.00	4.00
EXCLU_BVD	32.00	0.00	0.00	0.00
FRAMEwork_BVD	28.00	0.00	0.00	0.00
lepibats	17.00	7.00	1.00	0.00
OSCAR	171.00	142.00	64.00	8.00
PestiRed	2.00	33.00	0.00	0.00
SEBIOPAG_BVD	50.00	1.00	0.00	0.00
SEBIOPAG_Plaine de Dijon	5.00	20.00	82.00	33.00
SEBIOPAG_VcG	2.00	30.00	43.00	25.00
SEBIOPAG_ZAAr	2.00	24.00	20.00	7.00
percentage	30.26	16.48	14.78	4.99

From 1134 observations with complete crop rotation information in [N-4:N], 516 have the same crop for the whole time period, while 85 fields have four different crop groups in the past 5 years (Figure 18).

Remarks:

Because crops are not grouped yet, it is hard to visualize crop rotations.
Crop rotations depends a lot on the project targeting either annual vs perrenial cropping systems.

Hedgerows length

We overlaid the field as defined in RPG* (+ a 10m buffer) with the tree cover density from Liu et al. 2023.

boxplot
table

Figure 19: Hedgerow cover (%) per project. The dashed line show the median cover.

Table 9: Summary statistics per project of the hedgerow cover (%)

	Min.	1st Qu.	Median	Mean	3rd Qu.	Max.
BACCHUS_OPERA	0.00	0.63	2.19	3.99	5.66	17.54
BIOMHE	0.41	3.72	5.65	8.13	8.27	41.39
BISCO	1.48	5.78	9.27	11.88	18.33	27.58
DIVAG	0.48	4.63	7.73	8.43	11.40	34.71
EXCLU_BVD	11.64	27.60	34.89	46.12	71.06	90.29
FRAMEwork_BVD	9.07	14.51	26.59	36.45	61.78	90.29
lepibats	0.00	0.50	1.70	3.45	5.02	14.17
muesli	1.38	4.13	6.29	7.68	11.33	18.89
OSCAR	0.00	0.47	2.84	7.71	14.14	45.05
PestiRed	0.00	0.15	1.53	4.02	5.41	29.36
SEBIOPAG_BVD	1.06	12.39	28.07	29.74	33.14	90.29
SEBIOPAG_Plaine de Dijon	0.00	0.00	1.08	2.22	3.38	10.68
SEBIOPAG_VcG	0.24	4.45	7.09	8.34	11.49	21.71
SEBIOPAG_ZAAr	0.31	2.42	4.80	5.67	8.42	17.95
SERIPAGE	2.69	5.99	10.43	9.37	12.12	14.80

boxplot
table

Figure 20: Hedgerow biomass density (kg per ha) per project. The dashed line show the median biomass.

Table 10: Summary statistics per project of the hedgerow biomass density (kg per ha)

	Min.	1st Qu.	Median	Mean	3rd Qu.	Max.
BACCHUS_OPERA	0.00	0.63	2.19	3.99	5.66	17.54
BIOMHE	0.41	3.72	5.65	8.13	8.27	41.39
BISCO	1.48	5.78	9.27	11.88	18.33	27.58
DIVAG	0.48	4.63	7.73	8.43	11.40	34.71
EXCLU_BVD	11.64	27.60	34.89	46.12	71.06	90.29
FRAMEwork_BVD	9.07	14.51	26.59	36.45	61.78	90.29
lepibats	0.00	0.50	1.70	3.45	5.02	14.17
muesli	1.38	4.13	6.29	7.68	11.33	18.89
OSCAR	0.00	0.47	2.84	7.71	14.14	45.05
PestiRed	0.00	0.15	1.53	4.02	5.41	29.36
SEBIOPAG_BVD	1.06	12.39	28.07	29.74	33.14	90.29
SEBIOPAG_Plaine de Dijon	0.00	0.00	1.08	2.22	3.38	10.68
SEBIOPAG_VcG	0.24	4.45	7.09	8.34	11.49	21.71
SEBIOPAG_ZAAr	0.31	2.42	4.80	5.67	8.42	17.95
SERIPAGE	2.69	5.99	10.43	9.37	12.12	14.80

Figure 21: Correlation between hedgerow cover (%) and biomass density (kg/ha).

There are 143 NA’s corresponding to observations without corresponding fields in RPG*.

Outliers

Figure 22: Field with high hedgerows cover

Figure 23: Field with high hedgerows biomass

Figure 24: Field with low hedgerows cover

Remarks:

Liu et al. 2023 doesn’t distinguish hedgerows from orchard.
Yet it appeared to be highly correlated with Grain Bocager

Land cover within buffer

Table 11: Summary of the land cover classes composition in 1500m buffer

	n_classes	av_perc_rpg
BACCHUS_OPERA	63	44.71
BIOMHE	68	70.92
BISCO	62	70.17
DIVAG	66	73.03
EXCLU_BVD	117	41.17
FRAMEwork_BVD	90	38.25
lepibats	73	40.13
muesli	69	71.84
OSCAR	171	43.25
PestiRed	117	54.13
SEBIOPAG_BVD	104	42.08
SEBIOPAG_Plaine de Dijon	122	65.18
SEBIOPAG_VcG	120	73.88
SEBIOPAG_ZAAr	98	73.54
SERIPAGE	50	74.77

Without any grouping, there are 354 different categories covered by the 1500m buffers (Table 11). Theses categories need to be simplified before the land cover can be analyzed.

In average, roughly half of the buffer areas are filled with land cover classes from RPG (and the other half are CLC classes). The proportion of RPG classes vary greatly by project.

Table 12: Top 10 most frequent land cover classes

   frac1500_1003_Woody.Broadleaved.deciduous.trees 
                                             14.68 
                           frac1500_212_vigne_cuve 
                                             12.27 
                frac1500_1006_Permanent.herbaceous 
                                             11.71 
            frac1500_1005_Low.growing.woody.plants 
                                              7.84 
                       frac1500_1_ble_tendre_hiver 
                                              6.39 
                              frac1500_1001_Sealed 
                                              5.98 
frac1500_186_prairie_permanente_herbe_predominante 
                                              4.03 
           frac1500_1002_Woody.needle.leaved.trees 
                                              2.73 
             frac1500_188_autre_prairie_temporaire 
                                              2.23 
                             frac1500_41_tournesol 
                                              2.00

Figure 25: Average land cover per project with buffer of 500m

Figure 26: Average land cover per project with buffer of 500m

Figure 27: Average land cover per project with buffer of 500m

Figure 28: Average land cover per project with buffer of 500m

Remarks:

All buffers include some amount of forest (5-20%) and sealed area (5-10%), while the crop composition is highly project specific.

Edge density

This section is not covered yet because it can be sensitive to the way fields are defined and the resolution of the raster (if rasterized).

To be discussed

Summary and questions about metrics

Figure 29: Correlation among field diversity metrics

Remarks:

Most observations have a corresponding crop field in RPG dataset (Table 2).
Could calculate elongation shape indicator (e.g. \(\frac{perim}{area}\), \(\frac{0.25*perim}{\sqrt{area}}\), \(\frac{perim}{2*\sqrt{pi*area}}\))
Further work on land cover class homogeneization is needed to make use of the extracted information. This will be done independantly from the GIS data extraction.
The edge density needs further thinking to decide which kind of edges should be quantified, and at what scale.
The sampling location within the field might influence the results (different impact of hedgerows, or of agricultural practices). We might want to add an indicator reflecting the distance to the center of the field and/or the distance to the closest field boundary?