Manual and Remote Techniques to Measure Nonconventional Marine Turtle Morphometrics
Abstract
Sea turtle morphometrics yield critical insights into demography, life history, and conservation status; however, methods in obtaining manual parameters are often invasive and not generally standardized because of measurement variation across measurement modality. To the best of our knowledge, no comprehensive guidance on all established morphometrics exists, nor does any existing literature prescribe methods for remote measurements (such as using a stereo-video camera), which may vary slightly in their approach compared to physical measurements. Here we provide a morphometric guide that recommends manual and remote techniques to measure and standardize morphometric parameters of Chelonid sea turtles, with a primary focus on green (Chelonia mydas), loggerhead (Caretta caretta), and Kemp’s ridley (Lepidochelys kempii) sea turtles.
Turtle physiology and morphology provide invaluable insights into life history, growth rates, behavior, diet, habitat preference across life stages, and even conservation status of species (Joyce and Bell 2004; Wyneken et al. 2007). Although scientists have conducted research on sea turtles for hundreds of years and carapace length is a routine variable to measure, data on relationships with demographic rates, abundance, and morphometrics are, paradoxically, severely lacking (Bjorndal 1999; Georges and Fossette 2006; Hamann et al. 2010). Morphometric data collected sequentially on individuals help estimate turtle growth rates. For example, fluctuations or rapid shifts in growth rates can be indicative of indirect ambient environmental factors or responses, such as diet availability and water temperature (Eguchi et al. 2012). Sea turtle morphometric parameters are useful to examine their role in marine ecology (Bjorndal 1999), investigate changes in growth (especially for the post-hatchling and pelagic juvenile life stages, which are known as the “lost years”; Hamner 1988), discern beach nesting success rates (Bjorndal 1999; Bolten 1999), understand resource availability in relation to site carrying capacities (Bjorndal et al. 2000), or assess local versus global differences in size and length at intraspecific population levels (Siegfried et al. 2021a, 2021b). Discerning body lengths, such as straight carapace length (SCL), can help scientists better understand potential trends in shifting size structure of populations (Casale et al. 2009; Burgett et al. 2018). Additionally, morphometric data can provide insights into the energy production associated with fecundity selection (Bulté et al. 2008).
Interestingly, although SCL is frequently recorded in situ when sea turtles are captured for data collection, not much work has been placed on assessing the relationship between nonconventional body lengths and environmental factors (Georges and Fossette 2006). Georges and Fossette (2006) comment on the lack of environmental research regarding body parameters, suggesting that little work has been done with body length measurements in situ because researchers often assume the data will result in weak relationships with sea turtle ecology. Furthermore, field collection of sea turtle morphological data creates knowledge gaps because of the typically opportunistic or female-specific sampling, hindering comparison of environmental parameters to body lengths and introducing bias in data collection (Georges and Fossette 2006; Cuevas et al. 2020). Finally, existing morphological data might include older data sets that do not fully reflect the evolving relationship between sea turtle size and age, and thus may not consider how these factors are influenced by population density and fluctuating environmental conditions over time (Araujo et al. 2019; Avens et al. 2020; Ramirez et al. 2020). Currently, conducting manual sea turtle measurements can be physically taxing for both the researcher and the turtle, as handling typically occurs outside of the water and can result in a significant amount of stress (Jones et al. 2008; Wabnitz and Pauly 2008; Wildermann et al. 2019; Siegfried et al. 2021a).
Most conventional methods involving the collection of sea turtle morphological data require manual measurements to quantify straight or curved carapace length, width, or weight (Bolten 1999; Wyneken 2001). However, resources are limited in providing guidance for the proper techniques in measuring nonconventional manual morphometric parameters involving the head, tail, and flippers (see Bolten 1999; Wyneken 2001; Casale et al. 2017). Such parameters can reveal novel aspects of sea turtle behavior or morphology including length-weight relationships (van Dam and Diez 1998), drag and energetics (Prange 1976), flipper propulsion (Lindborg et al. 2019; van der Geest et al. 2023), digging and nesting behaviors (Lindborg et al. 2019), survival rates, and fecundity (Salmon et al. 2015). Furthermore, these parameters may offer insights into linking morphometrics and biomass to body condition and informing assessments of mortality risk (Berger 2012; Stewart et al. 2021).
More recently, scientists have turned toward remotely obtained morphological parameters using photogrammetry and advanced remote sensing technology as these approaches are noninvasive, minimize catchability issues, and are precise and accurate in obtaining remote morphological parameters (Harvey and Shortis 1998; Shortis et al. 1998; Harvey 2001; Harvey et al. 2002; Rohner et al. 2011; Jeffreys et al. 2013; Krause et al. 2017). Photogrammetry and remote sensing both yield morphological data on wild, free-ranging animals without physical interaction. Most marine megafaunal photogrammetric studies estimate body size and mass to further understand related health conditions within their environments (Bell et al. 1997; Allan et al. 2019; Christie et al. 2022). In fact, remote studies have been conducted on various marine animals, including Australian snubfin dolphins (Orcaella heinsohni), southern elephant seals (Mirounga leonina), fur seals (Arctocephalus pusillus doriferus), leopard seals (Hydrurga leptonyx), and others (Bell et al. 1997; Krause et al. 2017; Allan et al. 2019; Christie et al. 2022).
Here we provide guidance on using a stereo-video camera system (SVC) as a method of remote photogrammetry to collect both conventional and nonconventional morphometric data of sea turtles. Conventional sea turtle morphometric data, such as SCL, have been successfully measured using SVCs and estimated to have a mean percent bias of –0.6% to –4.5% across several sea turtle species (Siegfried et al. 2021a). However, the same stereo-morphometric approach has not been applied to additional sea turtle morphometric parameters. Although obtaining SCL measurements is imperative in delineating the ontogenesis of sea turtles, additional value lies in acquiring nonconventional morphometric parameters (whether individually or in combination with other morphometrics) to further assess sea turtle biomass estimates or other somatic-specific aspects of their life stages and/or ecology. For example, collecting sea turtle cranial morphometric measurements allows scientists to assess feeding and bite performance in relation to foraging and prey selection (Marshall et al. 2012). The fitness and body condition of chronically debilitated sea turtles can be examined and compared to healthy, wild individuals with a combination of parameters, including SCL, biomass, body depth, or even plastron concavity (Stacy et al. 2018). Finally, acquiring flipper length, head length, plastron measurements, or SCL has been assessed by Casale et al. (2017) to discern whether sea turtles exhibit intraspecific variability in individual body parts based on geographic location and allometry.
Body length measurements are clearly vital to address many scientific questions. Although some manual measurement parameters are fairly standardized (e.g., straight carapace length and width and flipper lengths), many lack standardized guidance on collection methodology. Furthermore, with recent technological innovations in photogrammetry, all morphometrics need to be standardized for remote measurements and made to be comparable to manual methods. This guide provides a framework to standardize 10 morphometric parameters of green (Chelonia mydas), loggerhead (Caretta caretta), and Kemp’s ridley (Lepidochelys kempii) sea turtles, manually and remotely, to further assess how morphology plays a role in their biology and conservation. Our goal is to encourage the addition of these methods into turtle biologists’ toolkits.
A PREFACE FOR REMOTE MEASUREMENTS
Here we propose a set of standardized techniques to collect 10 morphometrics manually and remotely. We conducted field tests for each of the techniques using stereo-video cameras, though these protocols would apply to any form of photogrammetry. For certain morphometrics, the methodology varies slightly compared to manual measurements, based on visibility of anatomical reference points. Summarized information regarding manual and remote techniques, optimal camera orientation, and special considerations for obtaining sea turtle morphometric parameters can be found in Table 1. Roberto (2023) evaluated the accuracy of the SVC-derived morphometrics presented, and the mean percent bias was –1.74% across all species. The main challenge for using photogrammetry when collecting morphometric data in the wild is that filming often occurs while animals are traversing among habitat structures or resting under or on hard substrate; therefore, it is often possible for certain anatomical features to be obscured in the right or left camera frames of the SVC, or the perspective of the animal may not be ideal for the desired morphometric, e.g., dorsal view when attempting to measure body depth. Considering that one of the primary benefits of using photogrammetry is that it is noninvasive, a scientist may prefer to leave animals undisturbed. Depending on the in-water study design and site location (e.g., dive/survey method, reef or structural layout, etc.), scientific research permits may not be required to collect the data as long as the animals are followed for short periods of time (e.g., < 5 min) and are not physically disturbed. In that case, provoking the animal to reorient body positioning would not be permitted. In some cases, it is possible to interpolate the positioning of certain anatomical landmarks and still ultimately obtain accurate morphometrics (Siegfried et al. 2021a).
Before using SVCs for photogrammetry, standard SVC calibration and synchronizing operations should be completed. We recommend the following procedures outlined by Harvey and Shortis (1998) and Shortis et al. (1998). It is possible to custom-build a stereo-camera system, but we used a swimmable SeaGIS system (SeaGIS Pty Ltd, Bacchus Marsh, Victoria, Australia). The camera system consists of 2 action cameras (GoPro Hero 5) angled inwardly at 4° to converge at ∼ 0.25 m distance from the cameras. The cameras, encased in an aluminum housing, were fixed at 0.8 m apart to optimize field of view and measurement precision (Shortis et al. 1998). The optimal range between the SVC system and turtle was between 1 and 5 m while underwater, depending on the turtle’s size class and spatial availability within each habitat (Siegfried et al. 2021a).
Following video collection, the SVC footage is uploaded to the EventMeasure SeaGIS v.5.72 (SeaGIS www.seagis.com.au) analysis software to remotely estimate each morphometric parameter. To ensure the highest possible accuracy in estimated length measurements, the average of 10 measurements from 10 separate video frames is taken to obtain each targeted body morphometric (Harvey 2001; Siegfried et al. 2021a). The software does not require the video frames to be in chronological or consecutive order to obtain successful morphometric parameters. Measurement vectors are defined by selecting 2 of the same 3D geospatial landmarks on the turtle’s body (e.g., the nuchal notch; Fig. 1) in both the left and right camera frames. Further, a second pair of 3D geospatial points on an additional region of the turtle’s body (e.g., the supracaudal scute; Fig. 1) is selected in both the left and right camera frames to generate a completed vector to estimate the targeted measurement length (e.g., SCL, Fig. 2; Harvey et al. 2002). The following descriptions are intended to further discuss the recommended criteria in obtaining manual and remote nonconventional sea turtle morphometric parameters using calipers and the remote Event Measure SeaGIS analysis software.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Manual and Remote Straight Carapace Length and Width. —
Figure 1 illustrates the several types of scutes present on a hard-shelled (Cheloniidae) sea turtle carapace for reference of carapace anatomy (Wyneken 2001). Standard manual SCL can be measured with a metal caliper from the center of the nuchal scute through the midline of the carapace to the tip of the supracaudal scute (Fig. 2, Bolten 1999; Wyneken 2001). Similarly, remote straight carapace length (rSCL) can be remotely measured by selecting the first 3D geospatial landmark on the center edge of the nuchal scute followed by selecting a second point on the tip of the supracaudal scute (Fig. 2, Siegfried et al. 2021a). For best results, point selection should occur at a dorsal view of the turtle’s carapace; however, it is possible to obtain rSCL from a lateral view where the turtle’s body is positioned parallel with the cameras if the position of the nuchal scute can be observed or interpolated.
Manual straight carapace width (SCW) can be measured by selecting the widest points of the opposing marginal scutes while ensuring that the calipers remain as perpendicular to the body as possible (Fig. 2, Bolten 1999; Wyneken 2001). Furthermore, point selection for remote straight carapace width (rSCW) should occur on the widest opposing marginal scutes and is thus considered a similar measurement technique to its manual counterpart, SCW. The optimal perspective to take the rSCW is when the turtle is positioned at a dorsal view from the cameras (Fig. 2).
Manual and Remote Head Length, Head Width, and Head Depth. —
Manual head length (HL) can be measured by placing the calipers on the posterior-most portion of the parietal facial scales (Fig. 3) through the midline of the head, extending to the anterior portion of the nares (Fig. 4). Distal supernumerary scales can be included as the starting point in the standardized length if they are present on the turtle; however, the presence of the supernumerary scales varies across individuals. Remote head length (rHL) can be similarly measured, with the sea turtle positioned dorsally to the cameras for optimal access of reference points (Fig. 4).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Manual head width (HW) can be measured by using a metal caliper to select the widest points of the individual’s head (Fig. 4; Bolten 1999); notably, the temporal facial scales serve as an excellent reference point to measure HW on Cheloniidae, regardless of species type (Fig. 3). Moreover, remote head width (rHW) can be similarly measured to its manual counterpart, but for a complete remote measurement, the head should be positioned at a dorsal view (Fig. 4). This is to ensure that point selection does not extend to the tympanic facial scales along the side of the face (Fig. 3), as this would produce an overestimate of the standardized width.
Manual head depth (HD) can be measured by using the calipers to select the highest portion located dorsally on the head and the region below the mandibles (Fig. 5). The calipers should be maintained flat and perpendicular to the head while not resting along the soft-tissue portion of the neck, as this point of reference is too far back to obtain an accurate depth measurement. Remote head depth (rHD) can be measured in a similar manner, preferably from a lateral view relative to the cameras (Fig. 5). The posterior region of the frontoparietal scale should serve as a standardized point of reference to not overshoot measurement estimation; however, the position of this reference point will vary across species.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Manual and Remote Front Flipper Lengths and Widths. —
Manual front flipper lengths (FFLs) can be measured with calipers starting from the central point of the ulna (Fig. 6) and following the midline to the posterior-most tip of the flipper (Fig. 7). The ulna can be located by briefly palpating to find the bone. Notably, the ulna is selected as the anterior-most point of reference on the flipper as it reduces variability in point selection; further, the humerus, which is proximal to the ulna, is far less rigid due to the front flipper’s range of motion. Therefore, the humerus is not included when obtaining FFL (Casale et al. 2017). For the best measurement results, the flipper should be maintained on a flat surface angled roughly 45° away from the body’s midline as the calipers are placed parallel to the front flipper. If there is an instance when a portion of the flipper is missing near the standardized reference point, best judgement should be used to measure the next optimal point. Reflecting its manual counterpart, remote front flipper lengths (rFFLs) should be measured with similar judgement and technique (Fig. 7). During remote measurements, palpating the flipper is not possible; the ulna can be visually inspected by locating the region of the flipper where the limb begins to curve, relative to the shoulder joint. Depending on visibility and lighting, it is possible for the ulna to visibly protrude in the flipper.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Selecting common measurement points for front flipper widths (FFWs) varies depending on the species of sea turtle being measured. For sea turtle species in the family Cheloniidae that have only 1 claw (e.g., C. mydas and L. kempii), FFW should be measured with calipers by selecting the widest points on the flipper proximal to the claw; moreover, for turtles that have 2 claws (e.g., C. caretta), the widest points should be selected near the claw that is most proximal to the body (Fig. 7). The length of the claw should not be included in any of the measurements, as some individuals do not have them due to injury. Furthermore, selecting the same measurement points for remote front flipper widths (rFFWs) will also vary across species and should be prudently measured. The optimal point selection for rFFL and rFFW occurs from a dorsal view to the cameras, when the turtle’s flipper is positioned flat at a 45° away from the body; however, it is possible to obtain these parameters when the flipper is positioned at other angles, except when parallel to the body, such as when tucked underneath the carapace while swimming.
Manual and Remote Rear Flipper Lengths and Widths. —
Manual rear flipper lengths (RFLs) can be measured with calipers from the anterior-most visible portion of the flipper from the carapace through the midline of the flipper to the posterior-most tip (Fig. 8). Ideally, the turtle should remain relaxed without the rear flipper curling underneath the carapace or extending past a neutral position when dry-docked. A “neutral” dry-docked position is defined as when the turtle’s rear flipper is relaxed as if it were gliding underwater without the flippers kicking, thrashing, or propelling. We advocate for this approach to better standardize rear flipper measurement points across all individuals. Remote rear flipper lengths (rRFLs) should be measured by using the same technique (Fig. 8).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
Further, the rear flipper width (RFW) can be measured with calipers to select the widest points of the flipper (Fig. 8). An ideal reference point is the portion of the claw most proximal to the carapace, regardless if the flipper has 1 or 2 claws. Ensure that the calipers remain perpendicular to the rear flipper during the measurement. The remote width of the rear flipper (rRFW) should be measured with the same technique (Fig. 8). An ideal reference landmark for rRFW is the portion of the claw most proximal to the carapace, regardless if the turtle possesses 1 or 2 claws. Optimal point selection for rRFL and rRFW should occur when the turtle is positioned at a dorsal angle from the cameras while resting or gliding in the water with relaxed rear flippers. Instances when the turtle is briskly swimming or curling their rear flippers are not ideal to obtain rRFL and rRFW.
Manual and Remote Body Depth. —
Manual body depth (BD) can be acquired by placing metal calipers at the apex of the turtle’s carapace and the corresponding point of plastron directly below. This morphometric parameter can vary across species but is typically selected at the pinnacle of the second vertebral scute (Fig. 1, for reference) along the carapace. The calipers should lay flat on the vertebral scute and plastron while remaining perpendicular to the turtle’s body when obtaining the measurement (Fig. 9). Juvenile to subadult turtles may be completely lifted from dry-docked locations (e.g., platforms or tubs) to obtain the BD morphometric. However, for adult turtles that are more difficult to handle due to size or space limitations, individuals may be gently tilted onto their side to acquire BD. Remote body depth (rBD) can be acquired in the same manner with the cameras positioned at a lateral view (Fig. 9). To ensure that the plastron is visible in the camera frames for optimal measurements, point selection should occur when the turtle is swimming parallel while the cameras are placed slightly below the turtle’s plane.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World’s Turtle and Tortoise Journal 23, 2; 10.2744/CCB-1624.1
DISCUSSION
This guide provides manual and remote morphometric recommendations for green, loggerhead, and Kemp’s ridley sea turtles of various size classes. The measurement accuracy for the photogrammetry-based approach (using SVCs) for all morphometrics was high for these species with the overall mean percent bias ranging from –6.60% to 5.00% across all species, and on average was –1.74% (Roberto 2023). The provided techniques may also be applied to freshwater turtles, tortoises, and all 6 species of the family Cheloniidae, with minor adaptations, given that morphology of keratinous scutes serve as geospatial landmarks for morphometric measurements. Further, this guide may also be useful for collecting morphometrics for leatherback sea turtles (Dermochelys coriacea) but may require some modifications for their unique external anatomy. While hatchlings or post-hatchlings were not included in the data set to assess accuracy, we anticipate that the same manual and remote guidelines can be followed with discernment.
Manual Morphometric Considerations. —
This guide serves to standardize sea turtle body morphometrics that are otherwise not typically explained in detail to reduce point selection variation. Sea turtle morphology varies across species and individually; therefore, when collecting data, best judgment should be used to achieve the most accurate manual measurements. To ensure consistency and minimize bias, measurement tool selection is important. For manual measurements, large metal tree calipers (e.g., Haglof Mantax metal calipers) that can measure up to 1.27 m may be preferable over a soft measuring tape to strictly obtain straight length measurements. However, we acknowledge potential budgetary or practical limitations for packing large calipers for remote field work. Notably, metal calipers best complement the morphometric parameters that are measured in photogrammetry software (e.g., EventMeasure), which does not account for curved body measurements. Improper caliper size selection can bias the accuracy and precision of sea turtle morphometric measurements, which can influence morphological discrepancies in statistical analysis. For example, for measuring the SCL of a juvenile turtle, it would be appropriate to measure the turtle using an 80-cm caliper rather than a 127-cm caliper, as it better reflects the general size class of the turtle (Bolten 1999). Researchers should also consider the unique morphology of individuals, such as the presence of epibionts (e.g., barnacles, leeches, algae, etc.) or attached marine debris that may obstruct the typical landmark selection points on the turtle. If epibionts are unable to be removed prior to obtaining measurements, then the researcher should use their best judgment to select the optimal positioning of the geospatial landmark. Finally, we did not include in this guide recommended morphometric techniques for obtaining tail length as we did not have a sufficient sample size in adult turtles to remotely validate that the lengths are true to their manual counterparts (Roberto 2023); however, standardized tail length manual measurement techniques can be followed as described in Bolten (1999) and Wyneken (2001). For remote measurement of tail length, the body position and overlap of the carapace often prevented view of the tail, especially in juvenile individuals.
Remote Morphometric Considerations. —
The position of the turtle relative to the cameras can pose challenges in obtaining remote standardized morphometrics and may potentially introduce measurement bias. If the turtle is not within the proper field of view, remote morphometrics may be over or underestimated. Overall, we recommend positioning the SVC dorsal to the turtle’s body when obtaining remote measurements; however, this differs for rHD and rBD parameters, as the cameras should maintain a lateral view for best results. In practice, and depending on objectives for morphological data collection, we recommend swimming around the turtle from lateral to dorsal views. Siegfried et al. (2021a) found that rSCL estimations in EventMeasure were relatively robust regardless of the turtle’s orientation to the camera. Similar to obtaining manual morphometric parameters, some human error can occur while obtaining remote morphometrics; however, allowing for a dorsal view of the turtle’s body minimizes error and best reflects realistic sea turtle SVC encounters that occur in the wild within relatively shallow habitats (Siegfried et al. 2021b) and, thus, is the most optimal point of view in this guide.
CONCLUSION
Although sea turtles have long garnered the curiosity of scientists and the public alike, their morphometrics are understudied. Our goal with this guide and the remote techniques described is to standardize morphometric parameters across 3 sea turtle species. Standardizing manual and remote morphometric reference points and techniques are crucial as photogrammetry becomes increasingly normalized in contemporary sea turtle research. We encourage scientists to use this guide to reduce the potential for measurement discrepancies across sea turtle species and increase accuracy and sampling efficiency in situ. Studying morphometrics can give an enhanced view of growth dynamics, relationships with environmental factors, carrying capacities, health condition, and population status, which informs researchers on how to move forward in conservation management (Casale et al. 2018). Applying standardized photogrammetric techniques can be the first step in linking such relationships to sea turtle morphometrics nonintrusively and further uncover the unknowns of their ontogeny. Thus, this guide will support and promote the use of photogrammetry to fill in knowledge gaps without the need for physical capture in situ and for the sea turtle scientific community to broadly adopt noninvasive measurement techniques in the near future.
Sea turtle carapace illustrating the nuchal notch, supracaudal, and marginal scutes for measurement landmarks.
Suggested carapace scute landmark selection to measure (A) manual straight carapace length (SCL) and straight carapace width (SCW) and (B) remote straight carapace length (rSCL) and remote straight carapace width (rSCW) on a sea turtle.
Sea turtle facial scales located on the head of a hard-shelled (Cheloniidae) turtle.
Suggested facial scale landmark selection to measure (A) manual head length (HL) and head width (HW) and (B) remote head length (rHL) and remote head width (rHW).
Suggested facial scale landmark selection to measure (A) manual head depth (HD) and (B) remote head depth (rHD).
Dorsal view of the skeletal anatomy of the front flipper illustrating the ulna and humerus for measurement landmark selection.
Suggested landmark selection to measure (A) manual front flipper length (FFL) and front flipper width (FFW) and (B) remote front flipper length (rFFL) and remote front flipper width (rFFW).
Suggested landmark selection to measure (A) manual rear flipper length (RFL) and rear flipper width (RFW) and (B) remote rear flipper length (rRFL) and remote rear flipper width (rRFW).
Suggested landmark selection to measure (A) body depth (BD) and (B) remote body depth (rBD).
Contributor Notes
Handling Editor: Jeffrey A. Seminoff
