Results and Discussion

3.1. Spectral Separation

The wetland plant spectra displayed a range of JM distance values. Table 2 displays the JM distance values for all the plant spectra in a matrix against the invasive species [23].

According to the JM values, Heteranthera dubia (water star-grass) and Lythrum salicaria are relatively easy to separate with the highest JM value of 1.2645. Heteranthera dubia has the highest separability from most of the invasives compared to the other species collected (Table 2). The invasive with the lowest separation value from Heteranthera dubia is Elodea canadensis (another submergent species, Canadian waterweed) at 0.8599, still a moderate separation value.

Heteranthera dubia and Elodea canadensis are both perennial forbs with somewhat similar morphology and physiology, growing at the water-surface with a mat-like foliage texture. Both species had substantial amounts of water present in their spectra under field-canopy conditions resulting in distinguishable signatures compared to the other emergent- and more upland- aquatic invasive species in the study area. The VIS, NIR, and SWIR reflectance for Heteranthera dubia and Elodea canadensis spectra never surpassed ten percent reflectance factor because of the high amounts of water absorbing energy in the FOV.

Table 2. JM distance matrix for invasive species. Vallisneria americana-18 refers to a collection of spectra made with Vallisneria americana located approximately 18 inches below the water surface (averaged spectra n=32). (Low, medium, high detail evenly divided categories)

Table 2. JM distance matrix for invasive species. Vallisneria americana-18 refers to a collection of spectra made with Vallisneria americana located approximately 18 inches below the water surface (averaged spectra n=32). (Low, medium, high detail evenly divided categories)

Wetland species

Elodea

Lythrum

Nymphaea

Phragmites

canadensis

salicaria

odorata

australis

Eleocharis rostellata

0.9258

0.4213

0.1124

0.3245

Elodea canadensis

-

1.066

0.9693

1.0357

Filmacutee

0.9468

0.3406

0.0969

0.2435

Heteranthera dubia

0.8599

1.2645

1.228

1.2528

Iris versicolor

0.6335

0.7902

0.5909

0.7293

Leersia oryzoides

1.1229

0.2365

0.5282

0.3381

Lemna minor

1.1257

0.2525

0.5408

0.353

Lythrum salicaria

1.066

-

0.3251

0.1093

Mowed field grass

1.0351

0.1132

0.2187

0.0051

Myriophyllum verticillatum Nymphaea odorata Nuphar lutea

0.3723

n ncriQ

0.9407 0.3251 0.7317

0.7983

0.8968 0.2225 0.6647

0.9693 0.6966

0.5134

Phragmites australis

1.0429

0.1093

0.2225

-

Ponterderia cordata

1.0093

0.1818

0.1585

0.078

Polygonum pensylvanicum

1.0429

0.0763

0.2712

0.0643

Potamogeton spirillus

0.9264

0.418

0.1094

0.321

Sagittaria latifolia

1.0421

0.079

0.2551

0.0388

Salix eriocephala

0.9616

0.3141

0.0593

0.2127

Scirpus validus

0.7527

0.6696

0.435

0.599

Sparganium androcladum

0.6413

1.215

1.1649

1.199

Typha latifolia

0.7354

0.6946

0.4629

0.6234

Vallisneria americana

0.079

1.1444

0.9486

1.0194

Vallisneria americana-18

0.8

1.0515

1.2098

1.2371

Table 2. Continued

Wetland species

Polygonum

Pontederia

Sagittaria

Typha

pensylvanicum

cordata

latifolia

latifolia

Eleocharis rostellata

0.3663

0.2579

0.3535

0.7354

Elodea canadensis

1.0429

1.0093

1.0421

1.062

Filmacutee

0.2753

0.168

0.2672

0.3305

Heteranthera dubia

1.2552

1.2424

1.255

1.248

Iris versicolor

0.7487

0.68

0.7445

0.675

Leersia oryzoides

0.2949

0.4004

0.307

0.302

Lemna minor

0.309

0.4139

0.3218

0.4126

Lythrum salicaria

0.0763

0.1818

0.079

0.6946

Mowed field grass

0.0693

0.0761

0.0438

0.1546

Myriophyllum verticillatum

1.2182

1.202

1.218

0.8988

Nymphaea odorata

0.2712

0.1585

0.2551

0.4629

Nuphar lutea

0.6865

0.6108

0.6816

0.6578

Phragmites australis

0.0643

0.078

0.0388

0.6234

Ponterderia cordata

0.1179

-

0.1044

0.1224

Polygonum pensylvanicum

0.1179

0.0275

0.6474

Potamogeton spirillus

0.3625

0.254

0.3499

0.318

Sagittaria latifolia

0.0275

0.1044

-

0.0988

Salix eriocephala

0.2519

0.1389

0.2404

0.6418

Scirpus validus

0.6207

0.5378

0.6149

0.5982

Sparganium androcladum

1.2023

1.1848

1.2021

1.188

Typha latifolia

0.6474

0.5671

0.6418 -

Vallisneria americana

1.0276

0.9919

1.0264

1.0023

Vallisneria americana-18

1.2398

1.2256

1.2396

1.2217

Separability level medium high

Not surprisingly, Nymphaea odorata (water lily) had the lowest separation distance from all other species, including the invasives. This species was one of three floating species (Lemna minor and Nuphar lutea) found in this study, and typically displayed low to moderate densities. Thus, water played a large role in determining the "green" and NIR reflectance at this study site. Sagittaria latifolia (arrowhead) was the one invasive that displayed the lowest separability value against the other invasive species. Sagittaria latifolia had a relatively high separability measure from submergent invasive species such as Elodea canadensis at 1.0421, but very low separation scores from the other emergent/terrestrial invasives with an average of 0.1006 indicating difficulty in differentiating this species. Sagittaria latifolia had very small separation distances from Polygonum pensylvanicum (Pennsylvania smartweed) and these two are likely to cross-classify. These two plant species have relatively similar plant architectures and inhabit very similar niches in this ecosystem. The similarity of these two species with respect to the JM index lends support to the claim that plant canopy structure plays a large role in species separability. In contrast, Sagittaria latifolia also had very low separation values from Phragmites australis (common reed), an aggressive, very densely growing erect stalk with coarse texture that extends upwards of 2m, and Nymphaea odorata, a floating-leaf forb with large, thick circular leaves (25cm) with a significant spectra contribution from water, which suggests that plant canopy structure does not play a singularly strong role in differentiating spectra.

3.2. Continuum Removal

In theory a normalization process based on continuum removal can remove albedo, and/or background signal, from a spectral signature. In this study the modified continuum removal generally decreased separation distance as measured by the JM metric (Table 3). The differentiation of Lythrum salicaria from Sagittaria latifolia, Potamogeton spirillus (spiral pondweed), and Polygonum pensylvanicum increased slightly, although these species had very high JM separation values before the continuum removal was applied. The separability of Potamogeton spirillus from five of the invasives also increased slightly. Potamogeton spirillus can be submerged or float on the water surface with long (20cm), simple leaves. Therefore, an increase in species separability from Lythrum salicaria, Nymphaea odorata, Phragmites australis, Pontederia cordata (pickerelweed), and Sagittaria latifolia was contrary to expected results based on plant architecture, while emphasizing the role of leaf water content influencing separation. The decrease in separability for Elodea canadensis versus the non-submergent species further suggests that background signal and canopy architecture were indeed disregarded via continuum removal.

The continuum removal results suggest that applying this processing technique for all wetland plant species is not useful. While continuum removal might be effective in identifying absorption feature characteristics or particular wavelengths associated with biophysical attributes, applying the technique to aid in separating plant species (or classifying image data) can be disadvantageous. Clearly, for the wetland ecosystem studied here, continuum removal decreased abilities to separate the invasives species. These results further suggest that background and canopy architecture contributes to improving the separation of wetland plant species. The results here further advocate emphasizing life form when attempting to map wetland invasive species.

The background signal, or local environment, is what often creates conditions that support hydrophytic plants. The background signal includes variations in soil moisture or water content along with understory debris and the plant residue from previous growing seasons. These background factors provide useful biophysical information that is well-known to influence spectral reflectance. Therefore, when continuum removal techniques are applied, the loss of these background signals is detrimental to spectral separation of species in many cases. In other ecosystems or applications, such as geological and mineral identification, background signal may not be useful; in wetland ecosystems they are critical.

3.3. Spectra Features from Derivatives

The second-derivative approach identified wavelengths in this investigation similar to those identified in Becker et al. [8] using a dataset from a different sampling location. Figure 2 summarizes results of the derivative approach. With little substantial inflection (little change in slope resulting in a derivative near zero) in the NIR down slope or SWIR regions, no unique wavelengths were selected from these domains with the 2nd derivative approach; nearly all wavelengths had relatively equivalent strength in these wavelengths domains. The second SWIR plateau also tended to be noisy which derivative analysis is well-known to be

Table 3. Change in JM distance values with modified continuum removal applied. Negative values indicate increase in separation values. Vallisneria americana-18 refers to a collection of spectra made with Vallisneria americana located approximately 18 inches below the water surface (averaged spectra n=32).

Table 3. Change in JM distance values with modified continuum removal applied. Negative values indicate increase in separation values. Vallisneria americana-18 refers to a collection of spectra made with Vallisneria americana located approximately 18 inches below the water surface (averaged spectra n=32).

Wetland species

Elodea

Lythrum

Nymphaea

Phragmites

canadensis

salicaria

odorata

australis

Eleocharis rostellata

0.5615

0.3275

0.0344

0.1982

Elodea canadensis

0.7279

0.6421

0.7038

filmacutee

0.5976

0.2086

-0.0111

0.0843

Heteranthera dubia

0.8599

1 .2645

1.228

1.2528

Iris versicolor

0.3549

0.6905

0.5135

0.6105

Leersia oryzoides

0.8061

0.2127

0.5114

0.2968

Lemna minor

0.8418

0.2326

0.5038

0.3382

Lythrum saicaria

0.7279

0.3017

0.0765

Mowed field grass

0.739

0.0626

0.1648

-0.0329

Myriophyllum verticillatum

0.0678

0.8029

0.6837

0.7336

Nymphaea odorata

0.6421

0.3017

0.1707

Nuphar lutea

0.4064

0.6282

0.433

0.5388

Phragmites australis

0.7134

0.0401

0.1707

Pontederia cordata

0.7275

0.1201

0.1103

0.0087

Polygonum pensylvanicum

0.797

-0.0207

0.1822

-0.0273

Potamogeton spirillus

0.5914

-0.0948

-0.4067

-0.1708

Sagittaria latifolia

0.7686

-0.0081

0.164

-0.0274

Salix eriocephala

0.6238

0.2438

-0.0147

0.1607

Scirpus validus

0.4855

0.602

0.3898

0.5023

Sparganium androcladum

0.5946

0.9468

1.1649

1.199

Typha latifolia

0.3958

0.6254

0.4134

0.5229

Vallisneria americana

-0.0509

0.8959

0.7165

0.7411

Vallisneria americana-18

0.6076

0.6523

0.8297

0.8302

Table 3. Continued.

Wetland species

Polygonum

Pontederia

Sagittaria

Typha

pensylvanicum

cordata

latifolia

latifolia

Eleocharis rostellata

0.2157

0.1488

0.1879

0.4562

Elodea canadensis

0.797

0.7275

0.7686

0.3958

filmacutee

0.1221

0.0495

0.0859

0.2067

Heteranthera dubia

1.2552

1 .255

1.2567

Iris versicolor

0.6689

0.6248

0.6279

0.5645

Leersia oryzoides

0.2194

0.3624

0.2326

0.2826

Lemna minor

0.2221

0.355

0.2523

0.3565

Lythrum saicaria

-0.0207

0.1201

-0.008

0.6254

Mowed field grass

0.0157

0.0362

0.0063

0.1058

Myriophyllum verticillatum

1.0843

1 .0943

1.2034

0.8955

Nymphaea odorata

0.1822

0.1103

0.1641

0.4134

Nuphar lutea

0.5914

0.5437

0.5517

0.6245

Phragmites australis

-0.0273

0.0087

-0.0273

0.5229

Pontederia cordata

0.1179

0.0401

0.4901

Polygonum pensylvanicum

0.0748

-0.0202

0.5285

Potamogeton spirillus

0.3194

-0.232

-0.0926

-0.0264

Sagittaria latifolia

0.0078

0.0401

-0.0065

Salix eriocephala

0.2054

0.088

0.2232

0.2036

Scirpus validus

0.5166

0.4748

0.4916

0.5062

Sparganium androcladum

0.7865

1 .1 033

1.2021

1.1462

Typha latifolia

0.5285

0.4901

0.5072

Vallisneria americana

0.8655

0.8042

0.8221

0.8634

Vallisneria americana-18

0.9187

0.8852

0.8752

0.6947

highly susceptible [20]. Therefore we focus the discussion on the visible to NIR portion of the spectrum up to where the first sensor transfer occurs in the ASD instrument; another problematic region for derivative techniques.

Continuum removed spectra had relatively no influence on wavelength selection via 2nd derivatives compared to unprocessed spectra. Remember that continuum removal techniques isolate features and does not change the wavelength location in which features exist. Figure 2A displays normalized (into percent) frequency of occurrence for invasives, noninvasives, and all species with their continuum removed. No unique grouping-specific wavelengths or continuum removed wavelengths were selected. The three species that receive much attention from managers for possessing overly aggressive behavior in this study region- Phragmites australis, Typha latifolia, and Lythrum salicaria- had similar wavelengths selected.

(O 400450500550600650 700 750 800850900950 1000 ¡¡= _

(O 400450500550600650 700 750 800850900950 1000 ¡¡= _

b hh

IL Ñ\

/' I4 * ', ' «» » / V* ' \f

400 450 500 550 600 650 700 750 800 850 900 950 1000

400 450 500 550 600 650 700 750 800 850 900 950 1000

Figure 2. Frequency of large-magnitude 2nd derivative occurrences normalized into percent and offset with major axis representing units of five. Figure 2A details continuum removed 2n derivative frequency for all species (---), grouped 2nd derivative frequency for all invasives (-▲-), grouped 2nd derivatives frequency for all noninvasives (-o-), and grouped "Big 3" (Phragmites australis, Typha latifolia, Lythrum salicaria) frequency (-□-). Figure 2B displays plant canopy architecture groupings for floating (-□-), shrub (-▲-), and erect orientation (—). Note the consistency of the wavelength location of the occurrences across all groupings.

Figure 2B illustrates derivative results by plant canopy architecture groups. The groupings by morphological orientation (erect, shrub, floating) had similar wavelength locations selected. These results indicate that background signal does not largely determine the wavelength location where inflection points of utility occur. The relative wavelength locations are generally consistent across canopy types and invasive status according to the 2nd derivative approach. Clearly general domain windows were identified with the NIR region representing many large-magnitude occurrences. These findings are consistence with wavelengths selected on other wetland species and other groupings [8]. The 2nd derivative analysis shows that the wavelength location of absorption and reflectance features are consistent across species and that they are largely not category (e.g., invasive or architecture) dependant.

3.4. Shape Filter

While the absorption/reflectance features from average species spectra provide useful information, in reality the reflectance for individual wetland plant spectra display considerable variation. A limitation of this study is the lack of spectra collected under varied light conditions, atmospheric conditions, and time of year. However, this investigation is the first step or pilot study to investigate if the potential exists for the techniques presented to enhance species separation in Great Lakes coastal wetlands. Figure 3 illustrates the reflectance variability for Scirpus validus (softstem bulrush), Phragmites australis, Lythrum salicaria, and Typha latifolia (broad leaved cattail). Recall that the shape filter method [23] is intended to identify species of interest, such as invasives, using the uniqueness of the absorption features and reflectance variability. In essence, the more unique an absorption feature of a given species is, the easier that species can be distinguished.

Figure 3. Reflectance factor (n=32) variability, or shape-space, (maximum o, minimum ■, range -A-) illustrated for Scirpus validus, Phragmites australis, Lythrum salicaria, and Typha australis at 50nm intervals. Pre-processing removed wavelength regions severely affected by atmospheric absorption in the spectral ranges of 1350-1480, 1775-2000, and >2400nm.

The shape space varies by wavelength domain and by species. For example, Scirpus validus had relatively less variation compared to Phragmites australis, Lythrum salicaria, and Typha latifolia. The variation that does exist in Scirpus validus occurred primarily in the upper near-infrared shoulder, first SWIR plateau, and second SWIR plateau. This is likely due to the erect, small diameter structured growth and the fact that Scirpus validus tends to occur as a transitional plant between standing water and higher substrate on a microtopographic scale. Thus only minute differences were detected in leaf water content and plant biomass volume compared to variation in soil moisture and understory debris, again emphasizing the utility in background signal for identification purposes. Phragmites australis and Typha latifolia have larger reflectance variability in the NIR down slope (976-1190nm) and the upper near-infrared shoulder (1191-1450nm); however, both these regions have high separation abilities when the shape filter was applied (Figure 4.)

Elodea canadensis

Lythrum salicaria

Elodea canadensis

Nymphaea odorata

Nymphaea odorata

Lythrum salicaria

Phragmites australis

Phragmites australis

Pontederia cordata

Sagittaria latifolia

vVY.

Typha latifolia

Polygonum pensylvanicum

Figure 4. Number of plants species separated (displayed at 50nm intervals) by shape filtering for the wetland invasive plant species. Pre-processing removed wavelength regions severely affected by atmospheric absorption in the spectral ranges of 1350-1480, 1775-2000, and >2400nm.

The wavelengths identified as most useful for separating invasive species by the shape filter vary by species (Figure 4). This is critical as classification and processing techniques and/or choice of wavelengths might require evaluation based on the species of interest. Further, the biophysical properties influencing reflectance become valuable as background and plant canopy architecture vary by species thus potentially improving identification. For example, using the shape filter approach Polygonum pensylvanicum has the most separation around 605nm. Compared to the other invasives this wavelength has low separation value. Nymphaea odorata is most separable from the other species in the visible and chlorophyll domain (350-675nm) likely due to the species canopy being large (~24cm), round leaves that float on the water surface. Using the shape filter technique, Lythrum salicaria had between five (minimum at 2000nm) and 20 (maximum at 700nm) species distinguished.

The results from applying the shape filtering technique confirm that information provided by increased spectral data does increase abilities to distinguish plants of interest. The wavelength domains of utility vary by species therefore data reduction and wavelength selection methods need to consider evaluating species of interest and their individual absorption/reflectance features. The concept of spectral libraries and classification techniques based on shape filtering is promising for distinguishing invasive species. In the wetland ecosystem where this study was conducted, even very similar spectra were able to be filtered.

0 0

Post a comment