A novel eDNA assay for detection of the enigmatic eastern pine snake (Pituophis melanoleucus)

Snakes are experiencing worldwide population declines. Consequently, researchers require effective and reliable methods for detecting the presence of threatened snake species in critical habitats. Snakes, however, can be challenging to reliably detect during surveys due to their secretive habits and cryptic lifestyle. Environmental DNA (eDNA) assays are increasingly being used to aid ecological studies for the detection and discernment of target species populations. eastern pine snake (Pituophis melanoleucus) are secretive, semi-fossorial snakes found in southeastern United States and are considered a species of conservation concern in all states where they occur. At present, no species-specific primers exist for the detection of the eastern pine snake from eDNA. We developed a quantitative PCR (qPCR) and end-point PCR assay using novel eastern pine snake-specific primers targeting the NADH dehydrogenase subunit 4 (ND4) gene region. The developed primers and the assay conditions were validated for specificity, sensitivity, and application to eDNA for the detection of eastern pine snake from field samples. Our novel assay was determined to be specific and sensitive for the detection of eastern pine snake when tested against DNA from tissues or shed skins of heterospecific species and when applied to eDNA from the soil samples obtained from field sites. The developed primer set and eDNA assay provide an efficient, noninvasive tool to complement traditional survey techniques (i.e., camera traps and visual surveys) for detecting the presence of eastern pine snake from soil samples.

Determining the presence and distribution of threatened or endangered species across the landscape is essential for the development of effective conservation and management plans (Shaffer et al., 1998; Pollock, 2006; Rose et al., 2024). Detection of species of conservation concern, however, can be challenging because rare taxa often inherently occur at low population densities (Durso et al., 2011). In particular, secretive species, such as fossorial taxa that burrow or spend the majority of their time below ground, present additional challenges for the detection and assessment of population status (Jordaan et al., 2023). Common survey techniques such as active or passive monitoring may rarely detect or fail to detect the presence of individuals in a given survey area, even when the species is truly present (Sewell et al., 2012; Anderson et al., 2020; Findlay et al., 2020). Environmental DNA (eDNA) is increasingly being used as an effective tool for detecting the presence of elusive and secretive species, including several species of reptiles (e.g., Hunter et al., 2015; Katz et al., 2021; Matthias et al., 2021; Santamaria et al., 2024).

A major advantage of the use of eDNA survey methods is that it allows noninvasive detection of target species through the presence of small amounts of genetic material left in the environment (e.g., saliva, feces, skin, and blood), without direct observation of individuals (reviewed in Piggott and Taylor, 2003). eDNA has been frequently used in aquatic environments to detect fish, amphibians, reptiles, and mammals (e.g., Jerde et al., 2011; Pilliod et al., 2014; Ushio et al., 2017; Adams et al., 2019). However, it has been less commonly used for the detection of terrestrial vertebrates where eDNA is present in the soil (Adams et al., 2019; Leempoel et al., 2020). In contrast to aquatic environments, where the eDNA shed by animals can easily disperse in the water column, the eDNA in soil is likely to be much more localized to areas where an animal has been in direct contact with the substrate, therefore potentially limiting detection (reviewed in Harrison et al., 2019; Kirse et al., 2021). In addition, organic compounds such as the humic and tannic acids contained within soil samples can interfere with the isolation of eDNA and inhibit the amplification of target genes (Uchii et al., 2019; Wnuk et al., 2020). Despite these challenges, eDNA can be effective in detecting the presence of rare and elusive species in terrestrial environments. For example, eDNA was used to detect the presence of several rare South African species of fossorial golden moles (Chrysochloridae) in areas where the populations were previously not known to occur (Mynhardt et al., 2024). Similarly, eDNA from soil also was used for the detection of the endangered Louisiana pine snake (Pituophis ruthveni) (Katz et al., 2021).

Snakes, like many reptile taxa, are experiencing worldwide population declines (Reading et al., 2010; Gibbons et al., 2000; Farooq et al., 2024). Researchers, therefore, require reliable methods for the detection of species of concern in remaining habitat areas. Snakes, however, can be notoriously difficult to locate during field surveys due to their secretive nature and cryptic lifestyle (Willson et al., 2011, Willson et al., 2024). Fossorial species, in particular, can be especially difficult to detect using common survey techniques such as camera traps, box traps, or coverboards because they often spend relatively little time on the surface of the ground (Jordaan et al., 2023). In these situations, eDNA, used alone or in combination with traditional survey methods, can be effective for determining the presence of rare and secretive snake species in a variety of ecological settings (Matthias et al., 2021; Samuels et al., 2025).

Eastern pine snakes (Pituophis melanoleucus, Colubridae) are large (800–1800 g), secretive, semi-fossorial snakes typically associated with open-canopy, sandy woodland habitats in the Eastern United States (Gerald et al., 2006; Burger and Zappalorti, 2011). The population status of eastern pine snakes is poorly resolved. However, it is likely that eastern pine snake populations are in decline in many areas of their geographic range due to habitat loss and suppression of natural seasonal fire regimes (Burger and Zappalorti, 2016; Burger et al., 2017). In South Carolina, eastern pine snakes are listed as species of highest conservation concern, with patchy, discontinuous populations generally associated with the remaining areas of sandy, longleaf pine savannas that are maintained through regular prescribed burns (Grisnik et al., 2025). Because of their secretive lifestyle and relatively low detectability using traditional survey methods, eDNA molecular detection may enhance the ability to identify the presence of eastern pine snakes in predicted habitat areas. To date, however, species-specific primers are not available for the detection of eastern pine snakes from eDNA.

The purpose of this study was to develop a novel, species-specific primer set for the detection of eastern pine snakes using eDNA. We generated primers targeting the eastern pine snake NADH dehydrogenase subunit 4 (ND4) gene region and tested these primers for specificity, sensitivity, and ability to detect eastern pine snakes from field environmental samples.

DNA extraction from tissues and shed skin

DNA was extracted from frozen tissues using the DNeasy® Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol, with the exception of performing all spins at 12°C. The quantity and the quality of extracted DNA were determined using a NanoDrop™ One Microvolume UV–Vis Spectrophotometer (Thermo Fisher™, Florence, SC, USA). DNA quantity was reassessed for accuracy using the Qubit 2.0 Fluorometer with the Qubit™ dsDNA Broad Range Assay Kit (Invitrogen™, Waltham, MA, USA). DNA was extracted from shed snake skins following Fetzner (1999), with modifications. In brief, an approximately 1-in. piece of skin was cut from each snake skin, placed in a 1.5-ml tube, and washed in 10× phosphate-buffered saline (PBS) twice on a shaking platform at 50 rpm for 5 min each wash. The PBS-washed skin was then rinsed in nuclease-free sterile dH 2 O for 5 min at 50 rpm on a shaking platform. Subsequently, shed skin samples were placed in 300 μl of cell lysis buffer (10 mM Tris, pH 8.0, 100 mM EDTA, pH 8.0, 2% SDS, and 0.4 mg/ml proteinase K) and left at 56°C overnight to enable digestion of keratin and adequate cell lysis. The supernatant from the cell lysis was transferred to a new 1.5-ml tube and centrifuged at 8,000 rpm for 3 min. Following centrifugation, the supernatant was transferred to a new 1.5-ml tube, 600 μl of isopropanol was added, and the tube gently mixed by inversion, then placed at −80°C for 1 h to precipitate the DNA. The DNA was pelleted by centrifugation at 16,000 rpm for 20 min. The pellet was washed in 70% EtOH and centrifuged for 1 min at 16,000 rpm. This EtOH wash was repeated one time. The resulting DNA pellet was air-dried overnight, followed by rehydration in 20–100 μl Tris-EDTA (TE, pH 8.0), with the volume of TE added dependent on the size of the DNA pellet. All centrifugation steps occurred at 12°C to prevent degradation of the DNA. The snake and eastern glass lizard (Ophisaurus ventralis) specimens used for tissue extraction were acquired as dead specimens resulting from roadkill, stored at −20°C in Ziploc freezer bags, and identified by morphology. The identity of eastern pine snake shed skins was confirmed by Sanger sequencing using the PitmelF primer (this study) after PCR amplification.

Assay design and validation

We generated novel primers targeting the mitochondrial ND4 gene of P. melanoleucus from tissues, shed skins, and eDNA samples. An alignment of approximately 700 bp of ND4 nucleotide sequence from 56 snake isolates representing 13 snake species or subspecies was constructed using BIOVIA Discovery Studio Visualizer (v. 2024) (Supplementary Figure 1). All accessions were downloaded from GenBank (Supplementary Table 1) and manually aligned. The snake species chosen for the alignment included, but were not limited to, interspecific snake species commonly found in eastern pine snake habitat (S.L. Parker, unpublished data). Nucleotide regions with species informative single nucleotide polymorphisms (SNPs) were assessed for potential as P. melanoleucus -specific primers. Primers were tested for specificity in silico using the BLAST algorithm (Camacho et al., 2009). Versatility of eDNA assays is essential for widespread implementation of the assay in multiple labs. Therefore, the application of our P. melanoleucus -specific primers was tested in vitro using two different polymerase mixes: 1× SsoAdvanced™ Universal SYBR® Green SuperMix (Bio-Rad, Hercules, CA, USA) for the SYBR Green-based quantitative PCR (qPCR) assay or 1× Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) for the non-quantitative traditional end-point-based PCR (PCR) assay.

For in vitro assessment of the specificity of the devised primer set, DNA from a total of 13 snakes—four eastern pine snakes, one mud snake (Farancia abacura), one scarlet snake (Cemophora coccinea), one eastern kingsnake (Lampropeltis getula), one corn snake (Pantherophis guttatus), one copperhead (Agkistrodon contortrix), one eastern coachwhip (Masticophis flagellum), one bull snake (Pituophis catenifer), one canebrake rattlesnake (Crotalus horridus), and one lizard, the eastern glass lizard (Ophisaurus ventralis), as the outgroup—was extracted as described previously, diluted to 500 pg/μl, and then used in qPCR/PCR to assess the specificity of the PitmelF/PitRev primers. DNA from the eastern pine snakes was obtained from individuals used in a previous radiotelemetry study (Grisnik et al., 2025), while DNA from all other species was obtained from freshly obtained shed skins (bull snake) or from frozen roadkill specimens obtained from coastal South Carolina. In brief, a total of 1 ng (2.14 × 10 9 copies) of DNA (2 μl of 500 pg/μl dilutions) extracted from shed skins or tissues of the aforementioned snake and lizard species was added in a 10-μl qPCR or PCR reaction as described below. All reactions were run in triplicate for technical replication.

Sensitivity

The sensitivity of the devised P. melanoleucus -specific primer pair for the amplification of eastern pine snake DNA from eDNA samples was assessed using both qPCR and PCR. The minimum concentration (in nanograms per microliter) of eastern pine snake DNA required in soil to be detected (limit of detection, LoD) and quantified (limit of quantification, LoQ) in the assay using the devised primer pair was assessed using soil samples spiked with eastern pine snake DNA. Negative soil (soil collected from a site within the study area that lacked suitable habitat for eastern pine snakes) was spiked with eastern pine snake DNA at a 1:2 serial dilution spanning from 800 to 12.5 ng/μl for the generation of the standard curve. Spiked soil samples were processed for eDNA extraction using the DNeasy® PowerSoil® Pro Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions with the exception of adding a 10 min. 65 °C heating step following the addition of solution CD1 and vortexing briefly. The quantity and the quality of the extracted eDNA from spiked soil samples were determined using a NanoDrop™ One Microvolume UV–Vis Spectrophotometer (Thermo Scientific™, Florence, SC, USA). DNA quantity was reassessed for accuracy using the Qubit 2.0 Fluorometer with the Qubit™ dsDNA Broad Range Assay Kit (Invitrogen™, Waltham, MA, USA). A total of 400 pg of spiked and processed eDNA for each dilution of the standard curve was used in a 10 μl amplification reaction along with 1× of SsoAdvanced™ Universal SYBR® Green SuperMix (Bio-Rad, Hercules, CA, USA) or 1× Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) for qPCR or PCR, respectively. All reactions were run in triplicate for technical replication. Technical replicates with a >2 C q (quantification cycle) difference from the average of the two remaining replicates were removed to account for bias from pipetting errors. At least two out of three technical replicates for each standard were used for the generation of the standard curve. R 2 values were determined in Microsoft Excel by regression analysis using the average C q for each standard. The LoD of the reaction was determined as the minimum quantity of DNA that enabled amplification in at least two out of three replicates, with at least nine out of 10 subsequent replicates also positive for detection. The LoQ was defined as the quantity of DNA amplified in at least two out of three replicates, with at least nine of the 10 subsequent replicates also positive for amplification and a C q value equal to or less than 36 cycles to enable precise quantification. Only qPCR reactions using the SsoAdvanced™ Universal SYBR® Green SuperMix (Bio-Rad, Hercules, CA, USA) were used to generate the standard curve and determine the LoQ of the assay.

A 1:10 dilution series of eastern pine snake DNA extracted from shed skins spanning from 100 to 0.01 pg/μl was used to ascertain the LoD and LoQ for the P. melanoleucus -specific primers in the qPCR/PCR when a purer source of DNA was utilized in the PCR reaction. A minimum of five technical replicates for each dilution were used to generate the standard curve. Five technical replicates were used for each standard due to the low concentrations being tested and the need for increased precision. Only qPCR reactions using the SsoAdvanced™ Universal SYBR® Green SuperMix (Bio-Rad, Hercules, CA, USA) were used to generate the standard curve and determine the LoQ of the assay. The LoD of the reaction was determined as the minimum quantity of DNA that enabled amplification in at least two out of three replicates, with at least nine of the 10 subsequent replicates also positive for detection. The LoQ was defined as the quantity of DNA amplified in at least two out of three replicates, with at least nine of the 10 subsequent replicates also positive for amplification and a C q value equal to or less than 36 cycles to enable precise quantification.

Repeatability

The lowest quantity of DNA utilized in our standard curve determined to produce an amplifiable product in at least two out of three PCR reactions was deemed the LoD and was repeated in 10 additional qPCR/PCR reactions for confirmation of primer sensitivity. In the case of eDNA, 10 independent spikes of P. melanoleucus DNA in negative soil were used to check the repeatability of the determined LoD. The extracted eDNA from each independent spike was diluted to 200 pg/μl, with 2 μl of this dilution (a total of 400 pg of eDNA) used in 10 μl qPCR/PCR reactions following the reaction conditions and cycling described below to confirm primer sensitivity.

eDNA field survey

Approximately 30–50 g soil was collected from 27 sites within a small preserve on the Coastal Plain of South Carolina (the location of the study site is concealed to protect the snakes and their habitats) using sterile instruments and changing gloves at each collection site. The soil was placed in Ziploc freezer bags in a field cooler for transportation and then stored at 4°C until DNA extraction. Soil samples were collected during the summer months of June through August 2023. The collection sites were chosen from within areas where eastern pine snakes were known to be present based on a 2021–2022 eastern pine snake radiotelemetry study (Grisnik et al., 2025) and represent several hundred hectares of suitable habitat. All collection sites were from areas that fit the predicted habitat suitable for eastern pine snakes during the summer months and included soil collected from entrances of root cavities, stump holes, and similar refugia. DNA was extracted from environmental samples using the DNeasy® PowerSoil® Pro Kit (Qiagen, Hilden, Germany) as described previously. Three independent eDNA extractions were performed from each bag of soil to optimize the chances of finding pine snake DNA. The bags of soil were shaken for better distribution of DNA molecules that may have otherwise settled to the bottom of the bag directly prior to eDNA extraction. Although three independent eDNA extractions were generated for each bag of soil collected, only a single eDNA extraction was examined for the presence/absence of eastern pine snake DNA in bags 22–27 due to limited reagents. An environmental negative control (negative soil) was collected from a site that did not have suitable habitat for pine snakes and, therefore, was highly unlikely to contain pine snake DNA. The quantity and the quality of the extracted eDNA were determined using a NanoDrop™ One Microvolume UV–Vis Spectrophotometer (Thermo Fisher™, Florence, SC, USA). eDNA quantity was reassessed for accuracy using the Qubit 2.0 Fluorometer with the Qubit™ dsDNA Broad Range Assay Kit (Invitrogen™, Waltham, MA, USA). All eDNA was diluted to 200 pg/μl, with 2 μl of this dilution (a total of 400 pg) used in the qPCR reactions. All eDNA samples were assessed with three technical replicates in qPCR reactions. Assays were scored as positive if at least two out of three technical replicates produced a product of the expected size as indicated by the melting curve or gel electrophoresis. Assays resulting in a positive product for one out of three technical replicates were reassessed with another set of three technical replicates. If at least one of the three replicates were positive in the second set, the sample was deemed positive; otherwise, the sample was deemed as putatively positive. Samples that did not result in positive product for any of the three technical replicates were noted as negative for pine snake. Positive products from eDNA amplification were cleaned of residual PCR components using 30% PEG 8000/30 mM MgCl 2 solution (Carnaggio and Barthet, 2025) and sent to Eurofins Genomics (Louisville, KY, USA) for Sanger sequencing to confirm species identity.

Quantitative PCR/PCR

All qPCR/PCR reactions using the P. melanoleucus -specific primer pair included 0.6 μM of each primer and 1× final concentration of the polymerase master mix, either the SsoAdvanced™ Universal SYBR® Green SuperMix (Bio-Rad, Hercules, CA, USA) for qPCR or the Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) for PCR. The cycling conditions when using the SsoAdvanced™ Universal SYBR® Green SuperMix were as follows: 98°C for 3 min initial denaturation, followed by 45 cycles of 98°C for 15 s, 61.6°C for 15 s, and a melting curve using default settings (an increase from 65°C to 95°C, with a ramp rate of 0.5°C/cycle/5 s) in the CFX96 Touch Real-Time System (Bio-Rad, Hercules, CA, USA). The cycling conditions when using the Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) were as follows: 98°C for 30 s initial denaturation, followed by 50 cycles of 98°C for 10 s, 61.6°C for 30 s, 72°C for 10 s, and a final extension of 72°C for 10 min. The reactions were run in a C1000 Thermal Cycler (Bio-Rad, Hercules, CA, USA) for end-point PCR. All qPCR/PCR reactions were run with a minimum of three technical replicates.

Internal Transcribed Spacer (ITS) amplification

Verification of amplifiable DNA in eDNA, tissue, and shed skin samples was accomplished using the ITS3_Kyo2 (5′-GATGAAGAACGYAGYRAA-3′) and ITS4_Kyo3 (5′-CTBTTVCCKCTTCACTCG-3′) primers (Toju et al., 2012), which target the internal transcribed spacer region 2 (ITS2) of fungi as an internal positive control for all DNA extractions. The ITS PCR reactions included 1.0 μM of each primer and 1× final concentration of the Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA). The cycling conditions utilized a 2 min 98°C initial denaturation stage, followed by 35 cycles of 98°C for 10 s, 47°C for 30 s, 72°C for 30 s, and a final extension of 72°C for 10 min.

Primer design and specificity

Alignment of 714 nucleotides of the ND4 gene from 56 reptile accessions resulted in the identification of two potentially P. melanoleucus -specific primer sequences: PitmelF (5′-CATAATCCTCACACGGGGGT-3′) and PitRev (5′-GGATAGGAATATGTGTAGG-3′) (Supplementary Figure 1). Each primer sequence contained SNPs restricted to the P. melanoleucus species or the Pituophis genus, respectively (Supplementary Figure 1), enabling specificity to the target species eastern pine snake (P. melanoleucus) in the PCR amplification. PitmelF was determined as 100% conserved among the P. melanoleucus accessions, with the exception of GenBank accession: ND4DQ902312.1 (Supplementary Figure 1), which has high sequence similarity to P. catenifer and suggestive of a mislabeled accession. The PitRev reverse primer was 100% conserved among all Pituophis, but not restricted to melanoleucus (Supplementary Figure 1).

Specificity of the primers to P. melanoleucus was first tested using the BLAST algorithm (Camancho et al., 2009) using the “Highly Similar Sequences” megablast selection for comparison to known sequences in GenBank. Although each primer was shown to potentially bind to regions in non-target species including some snake species, the only common species with homology to both primers was P. melanoleucus (Supplementary Table 2). In vitro specificity tests using DNA extracted from frozen heterospecific snake and other reptile shed skins or tissues demonstrated that the PitmelF/PitRev primer pair was specific to only the eastern pine snake when used with the prescribed master mix and cycling conditions in qPCR or PCR, respectively (Figures 1A–C).

Assay validation (LoD and LoQ)

Assays using the PitmelF/PitRev primer set successfully amplified the eastern pine snake DNA from the spiked eDNA samples, with a single PCR product of 216 bp corresponding to a melting temperature of 81°C evident after amplification (Figures 2A, B). The qPCR assay validation using spiked soil resulted in a standard curve with a R 2 = 0.947 (Figure 2B). Soil spiked with 12.5 ng/μl of eastern pine snake DNA (5.36 × 10 10 copies/μl of target region) resulted in a positive qPCR product (~216 bp) in two out of three replicates, the lowest point of our standard curve. As such, we determined 12.5 ng/μl as the LoD and LoQ of the qPCR assay. An additional 10 independent spikes of 12.5 ng/μl of P. melanoleucus DNA in negative soil were used to assess the reproducibility of detection and quantification at the determined LoD/LoQ. Pine snake DNA was detected from all 10 spikes (all three replicates showing a positive product for each spike, with the exception of one 12.5 ng/μl spike that had two out of three replicates positive), with an average C q of 34.5 ± 0.47 standard error (SE) (Figure 2D). Similar results were obtained when using traditional end-point PCR with a LoD of 12.5 ng/μl, although a regression line was not generated from end-point PCR products (Figures 2C, D). A negative soil control (soil without spiked DNA added to it) was processed alongside the spiked soil samples to ensure amplified product from spiked soil was not due to contaminating pine snake cells in background soil (Figures 2C, D).

The application of extracted DNA from shed skins or tissues directly into the qPCR reaction resulted in a standard curve with a R 2 = 0.9635 using a minimum of five technical replicates for each point on the curve (Figure 3A). Five technical replicates were used in the generation of the standard curve to ensure accuracy when testing with low concentrations. The LoD of the PitmelF/PitRev amplification reaction was determined as 0.1 pg/μl of input DNA equivalent to an input of 4.29 × 10 5 copies/μl of the target region, with an average C q = 39.096 ± 0.26, while the LoQ was determined as 1 pg/μl equivalent to 4.29 × 10 6 copies/μl of the target region, with an average C q = 33.98 ± 0.26. DNA input of 0.01 pg/μl or less in qPCR never resulted in an amplified product in any qPCR trial. The same LoD (0.1 pg/μl) was determined when using Phusion HF Master Mix for amplification (Figure 3B). A total of 10 additional technical replicates were assayed at 1 and 0.1 pg/μl (the defined LoQ and LoD of our qPCR assay, respectively) for confirmation of the LoQ and LoD. An input of 1 pg/μl of DNA resulted in amplification in all 10 technical replicates in qPCR, with an average C q of 36.46 ± 1.19 SE. Input at the determined LoD of 0.1 pg/μl resulted in the amplification of nine out of 10 replicates, with an average C q of 37.75 ± 0.62 SE. Similar results were obtained when using Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) for amplification, with 10/10 replicates positive when using 1 pg template DNA, but only four out of 10 replicates positive when using 0.1 pg template DNA in the reaction (Figure 3C). Thus, 1 pg of template DNA was considered the LoD when using end-point PCR. Although the addition of 0.01 pg to the reaction when using Phusion HF Master Mix (Thermo Scientific™, Florence, SC, USA) produced faint product in three out of three technical replicates, no product was observed in any of the ten subsequent technical replicates (Figure 3C).

Both the qPCR standard curves generated in the current study resulted in R 2 values in the acceptable range (R 2 greater than or equal to ~0.95 to 1.00) (Bustin et al., 2009). However, the efficiency of amplification was substantially less for the eDNA standard curve generated from the spiked soil samples (Figure 2B), with an efficiency of only 67.7% compared with that of the qPCR reaction when eastern pine snake DNA was added directly to the reaction, with an efficiency of 97.33% (Figure 3A). The difference in efficiency between the two standard curves suggests that the PitmelF/PitRev primers were adequate for amplification, but most likely PCR inhibitors from soil such as humic, fulvic, and tannic acids derived from decomposing plant matter (Uchii et al., 2019) remained in the extracted eDNA, preventing optimum amplification conditions when eDNA was utilized in the reaction. Solvents such as dimethyl sulfoxide (DMSO), which are known to reduce hairpin structures in primers and enhance PCR efficiency (Jensen et al., 2010; Varadharajan and Parani, 2021), were tested for impact on the qPCR amplification of eastern pine snake-spiked eDNA standards, but were determined ineffective. The addition of either 1% or 3% DMSO to the qPCR master mix resulted in R 2 values of 0.8212 (1% DMSO) or 0.75 (3% DMSO), with no improvement in the qPCR efficiency (67.8% and 73.86%, respectively) when using eastern pine snake-spiked eDNA for the generation of the standard curve, thus was not implemented in further assays.

Field tests

A total of 69 soil samples representing 27 collection sites were processed for eDNA and tested for the presence/absence of eastern pine snake DNA. All the extracted eDNA field samples and the negative soil control were determined to be of suitable quantity and quality for use in qPCR/PCR assays (Supplementary Table 3). Out of the 69 soil samples, one sample (A24) generated a positive signal in two out of three technical replicates (Figures 4A, B), with an average C q = 42.08 corresponding to 0.39 ng/μl of template DNA (1.67 × 10 9 copies of target region/μl), a concentration below our deduced LoQ and LoD for eDNA. The amplified product from the A24 sample was PEG-cleaned and sent off for Sanger sequencing. The sequenced product resulted in a 98.8% identity match to existing GenBank accessions of P. melanoleucus (e.g., accessions: EU272829 and AF141114) (Supplementary Figure 2). Two additional samples, A38 and A48, produced a positive signal in one out of three technical replicates (Figure 5; Table 1). Positive amplification in both samples had corresponding C q values (42.13 and 41.53, respectively) well below that of our LoQ and LoD, without any further positive signal in additional trials. Low amplification levels precluded the ability to sequence products from the A38 and A48 eDNA samples. Thus, A24 was considered as positive detection for eastern pine snake, while A38 and A48 were considered as putatively positive. Amplification using the ITS primers ITS3_Kyo2 and ITS4_Kyo3 (Toju et al., 2012) demonstrated the presence of amplifiable DNA with an ITS product of ~400 bp in all eDNA samples, with the exception of B15 (Figure 4C, Supplementary Figure 3).

Detecting the presence of elusive reptiles such as snakes can be challenging, especially for secretive fossorial species that spend the majority of their time underground (Jordaan et al., 2023). We successfully used the ND4 gene region to develop a qPCR/PCR assay using a novel set of primers, PitmelF/PitRev, specific to P. melanoleucus even when tested against a closely related member of the same genus (P. catenifer) that differs by only two nucleotides in primer sequence (Supplementary Figure 1). The primers and the assay developed in the current study enable specific detection of P. melanoleucus from shed skin, tissues, and eDNA field samples (Figures 1A, B, 4B) without the requirement of additional probes to enhance specificity.

The ability to detect fossorial species using eDNA has aided the detection of threatened and endangered vertebrate species (e.g., Katz et al., 2021; Matthias et al., 2021; Brammell et al., 2023; Mynhardt et al., 2024). An essential criterion for the development of eDNA assays for conservation is the ability to detect only the target species from complex media without additional amplification of non-target species through the development of species-specific primers and optimized eDNA assay conditions. A species-specific eDNA assay requires primers with both enhanced specificity and adequate sensitivity to detect the target species from complex analytes such as soil, air, or water. In the absence of high specificity of primers to the target species only, DNA from non-target species may also be amplified from complex sample media, making detection of the target species difficult or impossible, even if additional molecular tools such as probes are used (e.g., Wilcox et al., 2013). For example, an eDNA assay generated for the Louisiana pine snake (P. ruthveni) resulted in moderate success, with two out of five known snake locations (determined by radiotelemetry) showing positive signal for P. ruthveni (Katz et al., 2021). The assay developed by Katz et al. (2021) for the Louisiana pine snake, however, lacked specificity even with the addition of probes in the qPCR reaction. Closely related species such as the bull snake (P. catenifer) and the eastern pine snake have been shown to either amplify or likely to amplify using the primer/probe combinations selected by Katz et al. (2021) for targeting Louisiana pine snake. Furthermore, in silico analyses suggested that several other non-target species may potentially be amplified with the same primers used for the detection of the Louisiana pine snake (Katz et al., 2021). Due to this limitation, the assay described in Katz et al. (2021) is likely to be most effective in areas where closely related taxa are unlikely to co-occur. Similarly, in a study examining primer specificity in mixed populations of trout, the C t values resulting from deduced species-specific primer/probe combinations were altered by the presence of non-target species in the sample (Wilcox et al., 2013). These alterations in the C t values suggest that the ability of primers to bind to target species was reduced when other species were present, demonstrating the importance of specificity and optimization of the primers and probes of eDNA assays for the accurate detection and quantification of target species.

The non-target species identified in the Louisiana pine snake assay that may potentially amplify using existing primer combinations based on in silico analysis included the eastern king snake, eastern coachwhip, and eastern racer, all of which coexist with the eastern pine snake (Powell et al., 2016; Chen et al., 2017; Katz et al., 2021; Krohn et al., 2024), rendering the primer combinations generated by Katz et al. (2021) limited for eastern pine snake detection without further optimization and testing. We tested the devised primers (PitmelF/PitRev) under optimized assay conditions from the current study in vitro against several sympatric species that reside in the same habitat as the eastern pine snake and demonstrated primer specificity to only the eastern pine snake when used under the prescribed eDNA assay conditions (Figures 1A, B). These results support the utility of our assay to detect eastern pine snake in all potentially suitable environments when using either qPCR or end-point PCR assays.

Sensitivity is also an important consideration in eDNA assay development. Sensitivity to the target species refers to the lowest quantity of target DNA required for detection using the prescribed assay method with a 95% detection rate (Klymus et al., 2020). Many eDNA assays report specificity levels at one copy of target DNA per microliter or less based on standard curves generated using synthetic oligomers such as gBlocks ™ (e.g., Katz et al., 2021; Samuels et al., 2025). Although this extreme sensitivity suggests strong detection of the target species, it may also result in false positives due to laboratory contaminants, wind, rain, or other mechanisms that may disperse DNA fragments unrelated to the actual presence of the target organism itself (Serrao et al., 2018; reviewed in Ke et al., 2025). Furthermore, tests assessing the utility of gBlocks ™ for endogenous spiked controls in aquatic environmental samples resulted in no amplification, even though standard curves demonstrated repeatable amplification in technical replicates down to 7.24 copies/μl (Zeinstra et al., 2025). The use of genomic DNA in standard curves may result in sensitivity at higher copy number per microliter, as observed in the current study, but also presents a more realistic threshold of detection for the organism itself. Katz et al. (2021) stated a sensitivity based on a standard curve using gBlocks ™ of 1 copy/μl target DNA. However, only 20% (two out of five) telemetry-confirmed sites resulted in positive amplification for Louisiana pine snake using the devised primer/probe combinations. Similarly, an eDNA assay for sharp-tailed snake was determined to have a sensitivity of 0.2 copies/sample; however, the percent detection rates using eDNA either from coverboard swabs or soil samples from sites where there were confirmed visual sightings of sharp-tailed snakes were 57% and 41%, respectively (Matthias et al., 2021). In both cases, false negatives were evident even with a highly sensitive assay as determined using gBlocks ™. Moreover, in Matthias et al. (2021), positive detection from all technical replicates was only achieved when the DNA copy number was in the thousands, supporting that the sensitivity determined by gBlocks ™ in the standard curve may not be realistic to the true detection limit of actual genomic DNA in the environment. In the current study, we validated the LoD/LoQ of the eDNA assay by applying eastern pine snake DNA to 10 endogenous spiked controls. All 10 endogenous spiked controls using genomic DNA from P. melanoleucus and processed using the DNeasy® PowerSoil® Pro Kit resulted in positive amplification (Figure 2D), supporting that the deduced LoD using genomic DNA was reproducible in soil.

Several factors may impact the detection and amplification efficiency of target DNA from soil even when an assay is optimized with high sensitivity. Humic, fulvic, and tannic acids derived from decomposing plant matter are some of the principal compounds from soil known to inhibit PCR even when using optimized polymerases for eDNA work (Uchii et al., 2019). Additional factors that may limit the detection of a target species from soil include the amount of time the target species spent in contact with the particular area of soil being sampled and the amount of time passed between the presence of target species (shed cells) in the soil and the collection of soil for eDNA processing (Samuels et al., 2025). For example, eDNA from the eastern indigo snake was present and detectable in soil up to 200 h after the snake was removed from the collection area (Samuels et al., 2025). However, the ability to detect eDNA significantly decreased over time and was dependent on how long the snake was present in the enclosure where the soil was sampled (Samuels et al., 2025). Furthermore, the type of medium used for the collection (e.g., sand versus soil) and the humidity, pH, salinity, temperature, and exposure to UV light may influence the rate of DNA decay and decrease the ability to detect the target species DNA (Strickler et al., 2015; Sirois and Buckley, 2019; Ali and Rasheed, 2024; Samuels et al., 2025). The soils where samples were collected for determination of eastern pine snake presence were rich in organic matter with an acidic pH (Laliberte et al., 2007) and collected during the summer, which is typically associated with relatively high humidity. The soil samples used in our study therefore most likely contained several endogenous organic inhibitors of PCR amplification, as well as increased DNA decay due to the humid and acidic conditions. Regardless of these factors, we demonstrated effective, specific, and repeatable amplification of P. melanoleucus DNA from spiked soil samples and field eDNA samples, supporting the efficacy of the PitmelF/PitRev assay.

One limitation to the effectiveness of eDNA assays in detecting the presence of target species in soil, as opposed to water samples, is the relatively small quantity of soil that can be tested at a time using commercial kits. Kits such as the DNeasy® PowerSoil® Pro limit the amount of soil processed per a reaction to 250 mg. This relatively small quantity of soil means that there may be a low probability of detection of the target species from a given location due to sampling error. In the current study, 30–50 g bags of soil were collected from each collection site, with three aliquots of 250 mg of soil sub-sampled from each bag to increase the likelihood of detection of P. melanoleucus. Accordingly, a caveat to our experimental design was that we were unable to determine whether the samples that failed to produce a positive signal for eastern pine snake were false negatives resulting from inadequate sub-sampling from the collection bags. Additional sub-sampling may reduce false negatives, but increase the total costs of the assay, which may be prohibitively expensive for certain budgets (Buxton et al., 2021). Alternative techniques for acquiring eDNA and enhancing the detection of target species include swabbing coverboards (e.g., Ratsch et al., 2020; Matthias et al., 2021; López-González et al., 2025) and filtration followed by sedimentation of larger quantities of soil (Lentendu et al., 2023; Miller et al., 2024; Samuels et al., 2025). Filtration followed by sedimentation has been shown to be effective for enriching protist DNA relative to fungal and plant DNA using 20 g of starting soil (Lentendu et al., 2023). Similar filtration techniques were used to detect invasive snakes in Florida (Miller et al., 2024) and the eastern indigo snake (Samuels et al, 2025) using eDNA from terrestrial habitats. Such techniques hold promise for improved detection rates; however, they also require additional tools such as specialized filters and vacuum pumps, which may limit their use in some research settings. One additional method that is only recently being applied to the detection of vertebrates from soil eDNA samples is the use of digital PCR (Đokić et al., 2025). Digital PCR offers high sensitivity, fewer amplification issues due to inhibitors, and a reduction in the number of false negatives compared with qPCR in eDNA assays (e.g., Ongkulee et al., 2025). Digital PCR, however, requires specialized machinery, has increased costs, and is a fairly new technology, requiring further validation before broad application to eDNA.

Next-generation sequencing (NGS) technologies such as those used with metabarcoding are an alternative method for the detection of rare and elusive species from eDNA (e.g., Tsukamoto et al., 2021; McCulloch et al., 2025). Although qPCR may be more sensitive for the detection of a single species than metabarcoding (McColl-Gausden et al., 2023), there are several advantages of the metabarcoding approach. For example, it allows the detection of multiple species that may be present, enabling researchers to capture potentially important information on community structure or competitive interactions (Tsukamoto et al., 2021). At present, there are few DNA barcoding primers specific to reptiles, but none have been used for detection using eDNA, nor are any specifically designed for snakes (Nagy et al., 2012). The development of snake-specific metabarcoding primers will enhance the ability to efficiently detect individual species and multiple snake species using common habitats.