Fast and precise leaf area index (LAI) estimation in terrestrial ecosystems is crucial for a wide range of ecological studies and calibrating remote sensing products. Presented here is the protocol for using the new LP 110 optical device for taking ground-based in situ LAI measurements.
Leaf area index (LAI) is an essential canopy variable describing the amount of foliage in an ecosystem. The parameter serves as the interface between green components of plants and the atmosphere, and many physiological processes occur there, primarily photosynthetic uptake, respiration, and transpiration. LAI is also an input parameter for many models involving carbon, water, and the energy cycle. Moreover, ground-based in situ measurements serve as the calibration method for LAI obtained from remote sensing products. Therefore, straightforward indirect optical methods are necessary for making precise and rapid LAI estimates. The methodological approach, advantages, controversies, and future perspectives of the newly developed LP 110 optical device based on the relation between radiation transmitted through the vegetation canopy and canopy gaps were discussed in the protocol. Furthermore, the instrument was compared to the world standard LAI-2200 Plant Canopy Analyzer. The LP 110 enables more rapid and more straightforward processing of data acquired in the field, and it is more affordable than the Plant Canopy Analyzer. The new instrument is characterized by its ease of use for both above- and below-canopy readings due to its greater sensor sensitivity, in-built digital inclinometer, and automatic logging of readings at the correct position. Therefore, the hand-held LP 110 device is a suitable gadget for performing LAI estimation in forestry, ecology, horticulture, and agriculture based on the representative results. Moreover, the same device also enables the user to take accurate measurements of incident photosynthetically active radiation (PAR) intensity.
Canopies are loci of numerous biological, physical, chemical, and ecological processes. Most of them are affected by canopy structures1. Therefore, accurate, rapid, non-destructive, and reliable in situ vegetation canopy quantification is crucial for a wide range of studies involving hydrology, carbon and nutrient cycling, and global climate change2,3. Since leaves or needles represent an active interface between the atmosphere and vegetation4, one of the critical canopy structural characteristics is leaf area index (LAI)5, defined as one-half of the total green leaf surface area per unit of horizontal ground surface area or crown projection for individuals, expressed in m2 per m2 as a dimensionless variable6,7.
Various instruments and methodological approaches for estimating terrestrial LAI and their pros and cons in diverse ecosystems have already been presented8,9,10,11,12,13,14,15. There are two main categories of LAI estimation methods: direct and indirect (see comprehensive reviews8,9,10,11,12 for more details). Mainly used in forest stands, ground-based LAI estimates are routinely obtained using indirect optical methods due to the lack of direct LAI determination, but they usually represented a time-consuming, labor-intensive, and destructive method9,10,12,16. Moreover, indirect optical methods derive LAI from more easily measure related parameters (from the viewpoint of its time-demanding and labor-intense nature)17, such as the ratio between incident irradiation above and below the canopy and the quantification of canopy gaps14. It is evident that Plant Canopy Analyzers have also been widely used to validate satellite LAI retrievals18; therefore, it has been considered a standard for LP 110 comparison (see Table of Materials for more details about employed instruments).
The LP 110, as an updated version of initially self-made simple instrument ALAI-02D19 and later LP 10020, was developed as a close competitor for Plant Canopy Analyzers. As a representative of indirect optical methods, the device is hand-held, lightweight, battery-powered, without any need for a cable connection between the sensor and data-logger that uses a digital inclinometer instead of a bubble level and enables faster and more accurate positioning and value reading. In addition, the device was designed to note immediate readouts. Thus, the time estimate needed for collecting data in the field is shorter for the LP 110 than Plant Canopy Analyzer by approximately ⅓. After the export of readouts to a computer, the data are available for subsequent processing. The device records irradiance within the blue light wavelengths (i.e., 380-490 nm)21,22 using an LAI sensor for making an LAI calculation. The LAI sensor is masked by an opaque restriction cap with 16° (Z-axis) and 112° (X-axis) fields of view (Figure 1). Thus, light transmittance can be noted using the device held either perpendicularly to the ground surface (i.e., zenith angle 0°), or at five different angles of 0°, 16°, 32°, 48°, and 64° to be able also to deduce canopy elements' inclination.
Figure 1: Physical features of the LP 110. The MENU key enables the user shift up and down throughout the display, and the SET button serves as the Enter key (A).The zenith view under different inclination angles (±8 due to the side view) and the horizontal view is fixed for LP 110 to 112° (B) similarly to the Plant Canopy Analyzer (modified by restrictors). Please click here to view a larger version of this figure.
Due to the higher sensitivity of the LAI sensor, its restricted field of view, in-built digital inclinometer, automatic logging of reading values at the correct position indicated by sound without a button press, the new instrument is also suitable for above-canopy readings at narrow valleys or even on broader forest roads to measure a wide range of sky conditions. Besides that, it enables quantification of mature stand canopies above the relatively high regeneration, and it attains higher accuracy of irradiance values than Plant Canopy Analyzer. Moreover, the price of LP 110 equals about ¼ of the Plant Canopy Analyzer. Contrariwise, the utilization of LP 110 in dense (i.e., LAIe at stand level over 7.88)23 or very low canopies as grassland is limited.
The LP 110 can work within two operating modes: (i) a single sensor mode taking both below-canopy and reference readings (above the studied canopy or in a sufficiently widespread clearing located within the vicinity of the analyzed vegetation) performed before, after, or during below-canopy measurements taken with the same instrument and (ii) a dual sensor mode using the first instrument for taking below-canopy readings, whereas the second one is employed for automatically logging reference readings within a regular predefined time interval (from 10 up to 600 s). The LP 110 can be matched with a compatible GPS device (see Table of Materials) to record each below-canopy measurement point's coordinates for both the modes mentioned above.
The effective leaf area index (LAIe)24 incorporates the clumping index effect and can be derived from measurements of solar beam irradiance taken above and below the studied vegetation canopy25. Thus, for the following LAIe calculation, transmittance (t) must be calculated from irradiation both transmitted below the canopy (I) and incident above the vegetation (Io) measured by the LP 110 device.
t = I / I0 (1)
Since the irradiation intensity exponentially decreases as it passes through a vegetation canopy, LAIe can be calculated according to the Beer-Lambert extinction law modified by Monsi and Saeki9,26
LAIe = – ln (I / I0) x k-1 (2),
Where, k is the extinction coefficient. The extinction coefficient reflects each element's shape, orientation, and position in the vegetation canopy with the known canopy element inclination and view direction9,12. The k coefficient (see equation 2) depends on the absorption of irradiance by foliage, and it differs among plant species based on the morphological parameters of canopy elements, their spatial arrangement, and optical properties. Since the extinction coefficient usually fluctuates around 0.59,27, equation 2 can be simplified as presented by Lang et al.28 in a slightly different way for heterogeneous and homogenous canopies:
In a heterogeneous canopy
LAIe = 2 x |ln t| (3),
or
In a homogeneous canopy
LAIe = 2 x |ln T| (4),
Where, t: is transmittance at each below-canopy measurement point, and T: is the average transmittance of all t values per measured transect or stand.
In forest stands, LAIe must be further corrected due to a clumping effect of the assimilation apparatus within the shoots29,30,31,32,33,34 to obtain the actual LAI value.
The protocol is devoted to the practical utilization of the LP 110 optical device for estimating LAIe in a selected example of Central European conifer forest stands (see Table 2 and Table 3 for the site, structural, and dendrometric characteristics). LAIe estimation in a vegetation canopy using this device is based on a widely used optical method related to the transmittance of photosynthetically active radiation and canopy gap fraction. The paper aims to provide a comprehensive protocol for performing LAIe estimation using the new LP 110 optical device.
NOTE: Before beginning to take planned field measurements, sufficiently charge the battery of the LP 110 device. Connect the instrument (USB connector, see Figure 1) and the computer through the attached cable. Battery status is shown in the left-upper corner of the device display.
1. Calibration before measurement
NOTE: For the LP 110, perform a dark calibration of the LAI sensor and in-built inclinometer calibrations before beginning each field measurement campaign.
2. Single sensor mode for LAIe estimation
Figure 2: Optimal weather conditions for taking LAIe measurements using the LP 110. The optimal weather conditions when using the LP 110 are uniformly overcast skies with no direct solar radiation (A), or use either before sunrise or after sunset (B). Please click here to view a larger version of this figure.
3. Dual sensor mode for estimating LAIe
4. An example of field measurement and LAIe calculation
Figure 3: Transect's layout for estimating LAIe in homogenous vegetation cover. Transect I-IV: transect's number; Χ: measurement point for taking the below-canopy reading. The first ten positions are labeled (1Χ-10Χ). Transects must be oriented perpendicularly to the rows of plants. Please click here to view a larger version of this figure.
The spatial structure obtained from both tested devices obviously differed in all studied plots, i.e., thinned from above (A), thinned from below (B) and a control without any silvicultural intervention (C; see Table 2 for more details). At the stand level, similar differences in LAI values obtained from the LP 110 and the Plant Canopy Analyzer were confirmed between thinned plots with various densities (A vs. B) using ANOVA and Tukey's test. For the Plant Canopy Analyzer, significantly higher LAI values were observed in the control plot with no silvicultural intervention than in the thinned ones (A, B). However, the values significantly exceeded LAI obtained from the LP 110 in the control plot. For the LP 110, LAI did not significantly differ in the C and B treatments. Contrariwise, a significant difference in LAI values between the C and A plots was found. Generally, LAI significantly decreased after applied thinning treatments in the studied stands. LAI estimated using the LP 110 (LaiPen LP110) decreased more evidently in plot A, whereas the LAI values obtained from the Analyzer (LAI-2200 PCA) decreased more in plot B. Nevertheless, these recorded differences were slight (Figure 4).
Figure 4: LAI values estimated using the LP 110 and the Plant Canopy Analyzer optical devices in Norway spruce pole stands under different silvicultural treatments. For estimating LAI, 81 below-canopy readings were taken in each studied stand. A: Thinning from above; B: Thinning from below; C: Control plot. The dots signify the mean LAI value. The whiskers display the standard deviations. Various letters indicate significant differences (p < 0.05) among the silvicultural treatments and different optical instruments using Tukey's Post-hoc test. This figure has been modified from Černý et al.20. Please click here to view a larger version of this figure.
The LAI values' spatial variability is illustrated in Figure 5 for each thinning treatment in pure Norway spruce pole stands.
Figure 5: Spatial heterogeneity of LAI estimated using the LP 110 and the Plant Canopy Analyzer at the level of individual measurement points under studied spruce canopy. A: Thinning from above; B: Thinning from below; C: Control plot. The numbers above arrows signify the lateral side length and spacing of measurement points within the regular grid. This figure has been modified from Černý et al.20. Please click here to view a larger version of this figure.
The LP 110 underestimated LAI by 7.4% and 10.6% in plots A and C, respectively. Contrariwise, this device overestimated the LAI stand value obtained from the Plant Canopy Analyzer in plot B by 3.7%. If the total averages from all LAI values regardless of the thinning treatment applied were calculated and subsequently compared (LP 110 vs. Plant Canopy Analyzer), the LP 110 device underestimated LAI obtained by the Plant Canopy Analyzer by 5.8%. Subsequently, differences in specific LAI values measured above individual points arranged within the regular grid were calculated for both instruments, and these deviations were subsequently expressed as a percentage. Under these circumstances, the LAI values measured by the LP 110 and the Plant Canopy Analyzer differed profoundly (Table 1).
Silvicultural treatment | Forest stand LAI | Relative differences (%) among LAI from LaiPen LP 110 compared to LAI-2200 PCA at the level of individual measurement points | |
LaiPen LP 110 (m2 m-2) | LAI-2200 PCA (m2 m-2) | ||
A | 7.05 ± 1.73 | 7.61 ± 2.29 | 1 ± 37 (-58; 156) |
B | 7.76 ± 1.36 | 7.48 ± 1.75 | 8 ± 30 (-33; 183) |
C | 8.35 ± 1.23 | 9.34 ± 2.51 | -5 ± 26 (-48; 115) |
Table 1: Mean LAI at the stand level and LAI differences expressed as a % between the LP 110 and the Plant Canopy Analyser at the level of individual measurement points. A: Thinning from above; B: Thinning from below; C: Control plot. This table has been modified from Černý et al.20.
For all LAI data measured at a particular point level using the LP 110 and the Plant Canopy Analyzer, linear regression between both the employed devices was performed. The linear regression of y = 0.8954x (R2 = 0.94; RMSE = 2.11438) was found for all LAI data from both the tested instruments (Figure 6).
Figure 6: The linear regression among LAI values coming from the LP 110 and the Plant Canopy Analyzer at the level of individual measurement points in studied Norway spruce pole stands. This figure has been modified from Černý et al.20. Please click here to view a larger version of this figure.
Geographic coordinates | 49°29'31" N, 16°43'30" E |
Altitude | 610-625 m a. s. l. |
Mean annual air temperature | 6.5 °C |
Mean annual precipitation | 717 mm |
Table 2: Characteristics of the study site. This table has been modified from Černý et al.20.
Plot | Age of stand (years) | Stand density (trees ha-1) | Height (m) | DBH (cm) | BA1.3 (m2·ha-1) | Growing stock (m3·ha-1) |
A | 36 | 1.930 | 14.14 ± 3.73 | 14.84 ± 6.13 | 36.60 ± 0.25 | 250.02 ± 2.00 |
B | 36 | 1.915 | 16.33 ± 2.37 | 15.81 ± 4.47 | 43.41 ± 0.17 | 290.07 ± 1.32 |
C | 36 | 4.100 | 12.72 ± 2.68 | 10.97 ± 4.81 | 36.96 ± 0.19 | 287.12 ± 1.39 |
Table 3: Dendrometric and structural characteristics of the studied stands covering an area of 25 m x 25 m in 2014. In each studied stand, 81 below-canopy readings were taken within a regular grid (3 m x 3 m) under standard overcast skies (for more details, follow Černý et al.20). All LAI measurements were conducted in July and August when LAI values are most stable9,38. A: Thinning from above; B: Thinning from below; C: Control plot; DBH: stem diameter at breast height; BA1.3: the basal area at breast height. For BA1.3 at the stand level, the basal areas of each tree presented in the studied stand, calculated as: BA1.3 = (∏*DBH2)/4, was summed up. This table has been modified from Černý et al.20.
What are the differences between the LP 110 as a newly presented device for estimating LAI (or taking PAR intensity measurements) and the LAI-2200 PCA as an improved version of the previous standard LAI-2000 PCA for estimating LAI via an indirect method? Beyond the price being about fourfold higher for the Plant Canopy Analyzer compared to the LP 110, the number of output parameters, measurement conditions, methodological approaches, and possibilities of estimating LAI for different canopies, accuracy of results, etc., can be compared.
When comparing the hardware, the LP 110 seems to be more user-friendly. The LP 110 is a lighter device and does not require any cable connections between the sensors and the data-logger. Both sensors (i.e., for LAI and PAR measurements; see Figure 1) are integrated within the body of the device, allowing the operator to move easily throughout the studied ecosystem (e.g., in shrubs or dense forests). To ensure the reading value accuracy, a correct sensor position and value storage are essential. This position (either in the zenith or pre-set angles) is identified by a changing sound frequency if the sensor is close or far from the target position. Even under the most intensive sound (the volume can be corrected), the LP 110 held automatically saves the reading value. Contrariwise, finding the correct sensor position for the Plant Canopy Analyzer must be done with a manual bubble level on a hand-held stick. The operator must press the button to save the reading value simultaneously while checking the bubble level. However, the correct sensor position is routinely lost when pressing the button, resulting in decreased accuracy of the reading value. Since visually checking a bubble level is not necessary for taking LP 110 readings, there is also the possibility to hold the instrument on an extension rod, enabling the user to measure above canopies of natural or artificial regeneration, tall herbaceous or shrub layers. In this case, the correct sensor position can simply be found based on the changing sound frequency.
There are differences between the LP 110 and the Plant Canopy Analyzer in respect of LAI sensor construction, especially with regard to sensor sensitivity and the sensors' fields of view (FOV). If the LAI sensor of the Plant Canopy Analyzer is exposed to open-air, it can fog up under high air humidity conditions, which commonly occur in the early morning in open areas. Contrariwise, the LAI sensor of the LP 110 is fog-free as it is located inside the restrictor view cup (Figure 1). Although the restrictor of the LP 110's LAI sensor is removable, it has a fixed FOV; however, the FOV of the LAI sensor of the Plant Canopy Analyzer can be modified both in the azimuthal and zenith directions using different restrictors (opaque view caps) and by using a masking procedure during data post-processing, respectively. Even though the FOV of the LP 110's LAI sensor (Figure 1) is relatively narrow and cannot be manipulated compared to the Plant Canopy Analyzer, the sensitivity of this sensor is about tenfold higher. This higher LAI sensor sensitivity enables the user to take measurements using the LP 110 under conditions of low irradiance and also to take above-canopy (reference) readings on extremely narrow open plots, for instance, on narrow forest roads or lines. Furthermore, the above to below-canopy readings' ratio is higher, leading to increased accuracy of the measured transmittance and thus better LAIe estimation. On the other hand, it is necessary to increase the number of below-canopy readings per transect owing to the narrow FOV of the LP 110's LAI sensor.
There are some similarities between the LP 110 and the Plant Canopy Analyzer, for instance, in measuring conditions and in modifications of the LAI sensor zenith angle view (in directions of 0°, 16°, 32°, 48°, and 64° for the LP 110; and 7°, 23°, 38°, 53°, and 68° for the Plant Canopy Analyzer) to quantify the inclination angle of canopy elements. Similar to the Plant Canopy Analyzer, the LP 110 diminishes the effect of light reflectance and measures a real light absorption part of the light by foliage due to specific sensor wavelength characteristics. Other optical-based instruments such as SunScan, AccuPAR, TRAC39, or DEMON9,40 (for more details, see Table of Materials) measure under relatively wider light intervals regardless of the light reflectance. In dual sensor mode, it is possible to take automatic measurements with one sensor ordinarily placed in an open area to take above-canopy (reference) readings in time intervals ranging from 10-360 s and 5-3,600 s for the LP 110 and the Plant Canopy Analyzer, respectively, and there is the possibility to add GPS positions to individual measurements. For both the instruments, it is impossible to measure LAIe: i) during and immediately after rain conditions, as wet canopy elements, including stems enhance both light reflectance and transmittance values below the canopy; thus, actual LAIe is underestimated under such conditions; ii) during windy conditions when canopy elements are moving, and transmittance values vary greatly even though the sensor position is stable, and iii) during unstable synoptic situations when light conditions change rapidly. The last condition is not so limiting for the LP 110 due to the sensor's narrow FOV. Also, a distance of obstacles need to be considered. However, a suitable sensor orientation lessens the problem. For both devices, it is likewise possible to estimate LAIe during a sunny day, mainly close to sunrise or sunset. Except for midday when direct sun rays can enter the LAI sensor through the restrictor cap slot, taking LAIe measurements is feasible throughout the whole day; even if the LAI sensor is perpendicularly oriented toward the sun (relevant for the LP 110) or the back of the operator (relevant for the Plant Canopy Analyzer). However, some correction procedures presented by Leblanc and Chen41 must be applied. If the above-canopy readings vary by more than ±20% during a short time span (approximately 1-2 min), continuing to take LAIe measurements is useless due to the expected extremely high LAIe estimation error. That problem could be avoided with a precise synchronous estimation of above- and below-canopy readings in dual sensor mode employing two units with the same accurate time setup and calibration. The next critical step for estimating LAIe using the LP 110 is a selection of a suitable open area for above-canopy readings, especially for single sensor mode (the maximal time lag between above and below-canopy readings, i.e., forest stand and open plot, must be 15-20 min), where the size of the open area must respect the sensor FOV. Besides that, the LP 110 is similar to the Plant Canopy Analyzer, not suitable for accurately estimating LAIe in too dense (i.e., LAIe at stand level over 7.88)23, very low canopies grassland, or the transmittance below 1%.
All the obtained values of incident light and light transmittance below the canopy with a time entry are post-processed using specific software, providing many output parameters, especially with the Plant Canopy Analyzer. Contrariwise, the software for processing the data obtained from LP 110 needs to be improved to be more automatic and user-friendly, such as the software relevant to Plant Canopy Analyzer. Moreover, it is advisable to modify the restriction cup for the LP 110 by the producer to change or adjust the sensor FOV.
The authors have nothing to disclose.
The authors are indebted to the Journal of Forest Science editorial board for encouraging and authorizing us to use the representative results in this protocol from the article published there.
The research was financially supported by the Ministry of Agriculture of the Czech Republic, institutional support MZE-RO0118, National Agency of Agricultural Research (Project No. QK21020307), and the European Union's Horizon 2020 research and innovation program (grant agreement No. 952314).
The authors also kindly thank three anonymous reviewers for their constructive criticism, which improved the manuscript. In addition, thanks go to Dusan Bartos, Alena Hvezdova, and Tomas Petr for helping with field measurements and Photon Systems Instruments Ltd. company for their collaboration and providing device photos.
AccuPAR | METER Group, Inc., Pullman, WA, USA | AccuPaR LP-80 | https://www.metergroup.com/environment/products/accupar-lp-80-leaf-area-index/ |
DEMON | CSIRO, Canberra, Australia | DEMON | |
File Viewer | LI-COR Biosciences Inc., NE, USA | FV2200C Software | https://www.licor.com/env/products/leaf_area/LAI-2200C/software.html |
FluorPen | Photon System Instruments Ltd. (PSI), Czech Republic | FluorPen 1.1.2.3 Sofware | https://handheld.psi.cz/products/laipen/#download |
Hand-held GPS device | Garmin Ltd., Czech Republic | Garmin eTrex 32x Europe46 | https://www.garmin.cz/garmin-etrex-32x-europe46/80117 |
Hand-held device for leaf area index estimation(LP 110) | Photon System Instruments Ltd. (PSI) Czech Republic | LaiPen LP 110 | https://handheld.psi.cz/products/laipen/#info |
Plant Canopy Analyser | LI-COR Biosciences Inc., NE, USA | LAI-2000 PCA | LAI-2200 PCA or LAI-2200C as improved versions of LAI-2000 PCA can be used, see: https://www.licor.com/env/products/leaf_area/LAI-2200C/ |
Statistical software | Systat Software Inc., CA, USA | SigmaPlot 13.0 | https://systatsoftware.com/products/sigmaplot/sigmaplot-version-13/?gclid=Cj0KCQjwzYGGBhCTARIs AHdMTQzgfb42vv0mWmcbVcflNO UvrLl802Lrhkfh23Qie2mIZfw4O8kp 7p0aAsoiEALw_wcB |
Statistical software | StatSoft Inc., OK, USA | STATISTICA 10.0 | For LAI visualization, wafer-plots in STATISTICA 10.0 were employed. |
SunScan | Delta-T Devices, Ltd., Cambridge, UK | SS1 SunScan | https://www.delta-t.co.uk/product/sunscan |
TRAC | 3rd Wave Engineering, Ontarion Canada | Tracing Radiation and Architecture of Canopies | http://faculty.geog.utoronto.ca/Chen/Chen's%20homepage/res_trac.htm |
Tripod | Any | NA | Tripod with standard nut |
Water level | Any | NA |