Regression Model Development and Updating
W×Se in soils was estimated from the regression equation described in Figure 1 (-0.747 * "log T-Se" -1.1733 ) for paired soil total Se and W×Se. Estimated W×Se for 54 soil samples in 2014 (when W×Se was not measured) was from 0.014 to 0.64 µg/g (geometric mean [GM] = 0.077 µg/g), with W×Se about 6.8% of total Se .
Paired samples were also used to describe the relation between detritus and W×Se concentrations (r2 = 0.51, n = 103, P < 0.001). Detritus Se concentrations were estimated using the following regression equation:
Detritus [Se] = (10(1.773+0.29102 × (Log W×Se))
The log-log relation between paired plant and WxSe (r2 = 0.398, n = 90, P < 0.001; Figure 2A) or total soil Se (Figure 2B) concentrations (r2 = 0.399, n = 90, P < 0.001) were examined and there was no difference between the two calculations (P = 0.528). The following models describe WxSe and total soil Se to plant Se:
Plant [Se] = (10(0.335 × Log W×Se + 0.808)
Plant [Se] = (10(0.280 × Log TSe + 0.405)
We used the geometric means of Se by trisection (Ohlendorf and Santolo 1994) for samples from 1988 to 2006 to relate total soil Se, WxSe, and plant Se to Se in herbivorous invertebrates. The plant to herbivorous invertebrate regression (r2 = 0.335, n = 84, P < 0.001) was not as strong as (i.e., more variable than) the relation between total soil Se (r2 = 0.397, n = 81, P < 0.001) or WxSe and herbivorous invertebrate Se (r2 = 0.492, n = 81, P < 0.001):
Herbivorous Invertebrate [Se] = (10(0.227 × Log TsoilSe + 0.780)
Herbivorous Invertebrate [Se] = (10(0.288 × Log W×Se + 1.12)
Carnivorous invertebrate Se then was related to herbivorous invertebrate Se (r2 = 0.335, n = 80, P < 0.001) by the regression equation:
Carnivorous Invertebrate [Se] = (10(0.549 × Log HerbInv + 0.610)
Previous studies using kestrels fed seleno-L-methionine mixed into a commercial meat-based diet (Yamamoto et al. 1998) described the relation between diet and blood Se (r2 = 0.91, n = 157, P < 0.001). However, because birds at Kesterson (and presumably other contaminated sites) feed on diet items with a range of Se concentrations, a correction factor was used for calculating blood and egg Se from the assumed diet. In the study by Yamamoto et al. (1998), one group of American kestrels were fed a prepared diet (i.e., 9 µg/g as SeMet) and another group were fed animals collected at Kesterson with a similar mean concentration (8 µg Se/g), and blood samples were taken during the time they were being fed these diets on Days 14, 35 and 56; the difference in blood Se ranged from 45 to 88 percent. The average difference of 0.76 was used as the initial correction factor for diet to blood Se:
Kestrel blood [Se] = (10((0.345 + 0.552 × (Log kestrel diet [Se]) × 0.76))
We used the kestrel blood Se equation for estimating barn owl blood Se but adjusted the correction factor to 50% based on measured barn owl blood Se from previous years:
Barn owl blood [Se] = (10((0.345 × (Log barn owl diet [Se] + 0.552) × 0.50))
Paired samples of Se in starling (Sturnus vulgaris) diet and their blood Se (Santolo 2007) were used to describe the relation between passerine diet and blood Se concentrations (r2 = 0.47, n = 9, P = 0.042). In addition, we assumed that Se uptake from plant and invertebrate diets was less than complete and used an initial correction factor of 0.70. Passerine blood Se concentrations were estimated using the following regression equation:
Passerine blood [Se] = ((10((0.524 × (Log Diet [Se] + 0.447)) × 0.70))
The passerine blood Se equation was used to calculate shrike blood Se but because their dietary items include vertebrates, we used the maximum correction factor from the kestrel study of 88% as the initial assumption, so shrike blood Se was estimated using the following regression equation:
Shrike blood [Se] = ((10((0.524 + 0.447 × (Log Diet [Se])) × 0.88))
We did not have paired diet:blood samples for killdeer so we used the kestrel blood Se equation for killdeer blood and assumed that, because they are invertivores, the correction factor would be similar to kestrels and shrikes and used 0.85 as the initial correction factor:
Killdeer blood [Se] = (10((0.345 + 0.59 × (Log killdeer diet [Se])× 0.85))
Santolo and Yamamoto (1999) estimated concentrations of egg Se for free-living birds using a blood-to-egg Se concentration relation established in captive American kestrel studies (Santolo et al. 1999) and assumed the relation to be similar for other predatory birds. However, the relation was developed from birds fed a constant elevated concentration of SeMet in their diet, which over-estimated egg Se in wild birds because free-living birds would be exposed to various concentrations and forms of Se in their diets. For example, at Kesterson invertebrate Se ranged from 0.6 to 48 µg Se/g in 1994 and 1995 (Santolo and Yamamoto 1999), and small mammal Se concentrations ranged from 2.4 to 37 µg Se/g in 1999 (Santolo 2009); we used the antilog of parent blood Se and 76% as the initial correction factor:
Egg Se = 10^(log parent blood Se)× 0.76)
Paired samples of Se in starling blood and egg Se (Santolo 2007) were used to describe the relation between passerine blood and egg Se concentrations (r2 = 0.74, n = 5, P = 0.002). Passerine egg Se concentrations were estimated using the following regression equation:
Passerine Egg [Se] = (10((0.42 + 0.28 × (Log Blood [Se])))
We used the passerine egg equation for loggerhead shrike eggs and, because of the exclusive animal diet of this species, we assumed greater Se uptake from the diet and applied a correction factor of 0.90 and used the following equation:
Loggerhead Shrike Egg [Se] = ((10((0.42 + 0.28 × (Log Blood [Se]))) × 0.9)
Blood samples of female adult California gulls (Larus californicus) and eggs (r2 = 0.45, n = 12, P = 0.012) from the Great Salt Lake, Utah (Conover and Vest 2009) were used to describe the relation between killdeer egg and blood Se concentrations. Killdeer egg Se concentrations were estimated using the following regression equation:
Killdeer Egg [Se] = (10(1.72 + 2.32 × (Log Killdeer Blood [Se]))× 0.001)
Trophic Transfer Factors
TTFs were used when there were not individually paired samples of consumers and their diet or no strong relationship was found. For some, a TTF was used in the initial model, but as more samples were collected, it became clear that there was a significant relation and a regression equation was developed and added to the model.
Because Se concentrations in carnivorous invertebrates were significantly higher than concentrations in herbivorous invertebrates (non-carnivores; P < 0.001), terrestrial invertebrates were separated into categories of herbivorous insects (1,046 samples; 39% were herbivorous beetles [Coleoptera], 12% were crickets, and 48% were grasshoppers [Orthoptera]) and carnivorous invertebrates (435 samples of which 6% were mantids [Orthoptera], 10% were scarab beetles [Scarabaeidae], and 84% were spiders [Arachnida]). We assumed that terrestrial non‑carnivorous insects were eating only plants. The TTF was based on 667 spatially and temporally paired plant and insect samples collected from 1989 to 1994, as follows:
Herbivorous insect [Se] = (Plant [Se]) × 2.8
We assumed that carnivorous invertebrates were eating only insects. The TTF was based on 218 spatially and temporally paired spider and insect samples collected from 1989 to 1994, described as follows:
Carnivorous invertebrate [Se] = (Herbivorous insect [Se]) × 1.5
We assumed that small reptiles (lizards and small snakes [i.e., ≤ 15 cm]) that are eaten by American kestrels and loggerhead shrikes because of their small size would likely be eating only arthropods and that there is a higher abundance of herbivorous insects than carnivorous invertebrates in most years. We chose dietary WFs to be 70% herbivorous insects and 30% carnivorous invertebrates. Few lizards had been observed at Kesterson since the 1990s and they were not caught in pitfall traps or found in kestrel pellets during sampling at Kesterson. Common kingsnakes (Lampropeltis getulus), common garter snakes (Thamnophis sirtalis), and gopher snakes (Pituophis catenifer catenifer) had been observed during monitoring at Kesterson, and six gopher snakes were captured and analyzed for whole body Se concentrations in 1989 and three in 1990 (Santolo, unpublished data). Thus, a TTF was calculated based on these spatially and temporally paired gopher snake and insect samples:
Small reptile [Se] = (Herbivorous insect [Se] × 0.7 + Carnivorous invertebrate [Se] × 0.3) × 1.43
Diets and WFs for voles were determined from stomach content analysis of 55 California voles captured at Kesterson (Santolo, unpublished data). Only trace amounts of invertebrates were observed in diets and they were likely consumed incidentally during feeding. The TTF was based on spatially and temporally paired plant, invertebrate, and vole samples collected from 1989 to 1994, as follows.
California vole (whole-body) [Se] = (Plant [Se] × 0.991 + Herbivorous insect [Se] × 0.009) × 2.009
We assumed that predatory birds at Kesterson did not discriminate between deer mice, western harvest mice, and house mice and fed on them based on availability. Therefore, we combined these species (mice) for predator diets. Deer mice were also modeled separately. Diets and WFs for mice were determined from stomach content analysis of 158 deer mice, 47 western harvest mice, and 84 house mice captured at Kesterson (Santolo, unpublished data). The TTF was based on 390 spatially and temporally paired plant, invertebrate, and mouse samples collected from 1989 to 1994, as follows.
Mice (whole-body) [Se] = (Plant [Se] × 0.794 + Mushroom [Se] × 0.004 + Herbivorous insect [Se] × 0.101 + Carnivorous invertebrate [Se] × 0.101) × 1.485
Diets and WFs for American kestrels at Kesterson were determined from examination of 31 pellets collected from kestrel nest boxes at Kesterson (Santolo and Yamamoto 2009). The fraction of onsite and offsite foraging is the estimated area and time that is spent foraging on Kesterson and offsite food items and is expressed as a percent in the model.
American kestrel diet [Se] = ((Herbivorous insect [Se] × 0.45 + Carnivorous invertebrate [Se] × 0.30 + Reptile [Se] × 0.01 + Passerine [Se] × 0.09 + Mice [Se] × 0.16)/5) × (% onsite foraging) + (% offsite foraging × 1.9))
Diets and WFs for shrikes were determined from the literature (Yosef and Lohrer 1995) and site observations. Because most shrike blood was collected during the fall and winter, the diet was adjusted to be dominated by vertebrate species that would be available as prey during that time:
Loggerhead shrike diet [Se] = ((Herbivorous insect [Se] × 0.10 + Carnivorous invertebrate [Se] × 0.14 + Reptile [Se] × 0.10 + Passerine [Se] × 0.11 + Mice [Se] × 0.55)/5) × (% onsite foraging) + (% offsite foraging × 1.9))
Diets and WFs for barn owls were determined from the literature (Marti et al. 2020) and site observations. Because most barn owl blood was collected during the fall and winter, the diet was adjusted to be dominated by vertebrate species that would be available as prey during that time:
Barn owl diet [Se] = ((Passerine [Se] × 0.018 + Mice [Se] × 0.032 + Vole [Se] × 0.95)/3) × (% onsite foraging) + (% offsite foraging × 1.9))
Diets and WFs for passerines were determined from stomach content analysis of nine western meadowlarks collected at Kesterson. Passerines were assumed to forage only within Kesterson. The TTF was based on 22 spatially and temporally paired plant and invertebrate samples collected from 1989 to 1994:
Passerine diet [Se] = (Soil [Se] × 0.03 + Plant [Se] × 0.12 + Herbivorous insect [Se] × 0.58 + Carnivorous invertebrate [Se] × 0.33) × 1.024
Diets and WFs for killdeer were determined from the literature (Jackson and Jackson 2000) and site observations. Except for 1990, aquatic invertebrates were not collected from Kesterson during sampling after 1988 when aquatic habitats were removed for mitigation (Ohlendorf and Santolo 1994; Ohlendorf et al. 2020). Carnivorous invertebrates were assumed to have similar concentrations of Se and were used in diet calculations for killdeer. Also, because killdeer were found almost exclusively using the perimeter of Kesterson (mostly along the drainage canal on the east side), we assumed that their diet included offsite invertebrates as well:
Killdeer diet [Se] = ((Carnivorous invertebrate [Se]) × (60% onsite foraging)) + (40% offsite foraging × 1.9))
Model Testing, Verification, and Calibration
For the model testing and verification, we used the measured W×Se for 1995, 1996, 1998, and 2001, which included a range of dry to wet years. The model was run, and means were compared to measured results for the year, so the model was tested and updated with additional sample results each year through 2001 biological monitoring. The results from sampling conducted in 2004 and 2006 were used to validate and further calibrate the terrestrial Se bioaccumulation model. The results for each of the years are based on the model calibrated using the previous year’s results.
The model performed well when compared to the results for 1995; Se concentrations modeled in the various plants and animals ranged from 99% to 160% of the measured sample results. Measured and predicted results were significantly different only for voles and kestrel blood. The model predicted Se concentrations that were significantly higher than those measured for voles (P = 0.002) and kestrel blood Se (P = 0.025).
Plants and invertebrates were not sampled in 1996 but WxSe was measured and used in the model. Before running the model with the 1996 WxSe results we made a minor change based on the 1995 results by reducing the percentage of onsite feeding of kestrels from 58% to 50% and increasing the correction factor from 0.76 to 0.88. The 1996 predictions ranged from 60% to 135% of measured sample results. Three predicted results were significantly different from measured results. Se in whole-body deer mice (P = 0.027) and blood Se in passerines (P = 0.006) were significantly higher than measured concentrations and predicted shrike blood Se was significantly lower (P < 0.001) than measured in 1996.
Using adjustments made from 1996 sampling for 1998 comparisons, the model predictions ranged from 65% to 120% of measured Se for plants and animals. The predicted Se concentrations were significantly lower than measured Se for carnivorous invertebrates (P <0.001) and killdeer eggs (P = 0.002). The model reasonably predicted concentrations in all other media and similarly predicted concentrations in eggs of kestrels, barn owls, shrikes, and killdeer.
For 2001 comparisons, Se concentrations predicted for mice (P <0.001), deer mice (P = 0.013), shrike blood (P = 0.008), and killdeer eggs (P = 0.016) were all significantly higher than measured concentrations, and passerine and barn owl blood predicted Se concentrations were significantly lower than the measured Se concentrations.
Results from limited monitoring conducted in 2004 and 2006 were added to the model to be used for comparison to the 2014 monitoring results (Table 1). For 2014, WxSe was not analyzed and no invertebrate samples were collected for analysis. The adjusted model predictions ranged from 62% to 206% of 2014 measured Se for various media. Kestrel blood samples were all from birds using nest boxes on Kesterson, so the predicted onsite foraging was increased from 50% to 75%. Also, based on the results from 2001, the killdeer correction factor was changed from 0.85 to 0.75 prior to running the model. Mouse and deer mouse predicted Se concentrations were significantly higher than measured Se (P < 0.001 for both). Only kestrel blood Se prediction was significantly lower than measured Se. The model reasonably predicted concentrations in all other media and similarly predicted concentrations in eggs of kestrels, barn owls, and shrikes.