**3.2 Factor analysis**

#### **3.2.1 Establishment of an index for multiple elements**

The contents of the eight elements measured in the kidney are shown in Table 1. The data were recalculated using factor analysis. The multiple variables used in the present study (contents of elements) were merged using factor analysis, a form of multivariate analysis. Thus, a higher factor score was thought to indicate more serious contamination by multiple elements. We obtained three significant factors, as shown in Table 2. No tendency for contamination was observed when the mean values of each category were compared. Thus, it is thought that the comparison using only mean values makes it difficult to understand the degree of contamination by multiple elements.


Table 1. The contents of the elements in kidneys from birds of each category. The results are represented as mean contents (μg/g dry wt.), and the standard error of the mean.

An Innovative Approach to Biological Monitoring Using Wildlife 163

We used previously published data to develop the indexes for Cd and Pb in our previous studies. However, we were unable to use a similar method in this study. Thus, the correlation between the factor score of factor 1 and that of factor 2 was investigated using various methods of classification. The correlation (Y=0.499X – 0.177, R2=0.656) was found between the factor score of factor 1 (Cr, Li and Ti) and that of factor 2 (Mo) in the results from common cormorants, as shown in Fig. 4. A similar correlation was obtained using the

Further, a correlation was also obtained when Figs 4 and 5 were summarized, as shown in Fig. 6. The regression line obtained was: Y=0.474X – 0.199, R2 =0.698. When the outliers among the data points were checked using the method of the 95% equal probability ellipse, three data points were identified as outliers. It was thought that correlation between two variables indicated normal equilibration in the target animals investigated using multiple elements. As mentioned above, we decided to use the regression line obtained in Fig. 6 and the equal probability ellipse as the multiple elements standard regression line, MSRL, and the multiple elements equal probability ellipse, MEPE,

Fig. 6. The relation between the factor score of factor 1 and that of factor 2 in the data used in Figs 4 and 5. Filled squares; birds collected in Chiba Prefecture (Fig. 5), empty squares; common cormorants (Fig. 4). Dotted line; 95% equal probability ellipse, solid line; regression

The factor scores of diving ducks and dabbling ducks were compared with the index obtained. As shown in Fig. 7, the factor score obtained from diving ducks was observed to fall within the MEPE, except for one data point. Similarly, two of nine data points were observed to fall outside the MEPE when the data from wild birds collected in Chiba

**3.2.2 Comparison of the degree of contamination of diving and dabbling ducks** 

results from wild birds captured in Chiba Prefecture (Fig. 5).

respectively.

line obtained from wild birds.

Prefecture were compared with the index (Fig. 8).


Table 2. The mean factor score and standard error of the mean for each category of birds.

Fig. 4. The relation between the factor score of factor 1 and that of factor 2 for the common cormorant. Filled squares; common cormorants collected in Shiga Prefecture; empty squares; common cormorants collected in Tokushima Prefecture.

Fig. 5. The relation between the factor score of factor 1 and that of factor 2 for wild birds collected in Chiba Prefecture. Filled squares; Anatidae, empty squares; other birds.

Anatidae 65 0.101±0.155 0.111 ±0.099 -0.026±0.099 Seabird 17 -0.261±0.078 -0.126±0.122 0.656±0.222 Cormorant 30 -0.133±0.103 -0.230±0.063 -0.259±0.019 Ardeidae 10 0.483±0.264 -0.067±0.159 -0.115±0.063 Others 5 -0.600±0.087 0.494±0.848 -0.112±0.206 Table 2. The mean factor score and standard error of the mean for each category of birds.

Fig. 4. The relation between the factor score of factor 1 and that of factor 2 for the common cormorant. Filled squares; common cormorants collected in Shiga Prefecture; empty squares;

Fig. 5. The relation between the factor score of factor 1 and that of factor 2 for wild birds collected in Chiba Prefecture. Filled squares; Anatidae, empty squares; other birds.

common cormorants collected in Tokushima Prefecture.

n **Factor 1 Factor 2 Factor 3 Cr, Li, Ti Mo Cu**

We used previously published data to develop the indexes for Cd and Pb in our previous studies. However, we were unable to use a similar method in this study. Thus, the correlation between the factor score of factor 1 and that of factor 2 was investigated using various methods of classification. The correlation (Y=0.499X – 0.177, R2=0.656) was found between the factor score of factor 1 (Cr, Li and Ti) and that of factor 2 (Mo) in the results from common cormorants, as shown in Fig. 4. A similar correlation was obtained using the results from wild birds captured in Chiba Prefecture (Fig. 5).

Further, a correlation was also obtained when Figs 4 and 5 were summarized, as shown in Fig. 6. The regression line obtained was: Y=0.474X – 0.199, R2 =0.698. When the outliers among the data points were checked using the method of the 95% equal probability ellipse, three data points were identified as outliers. It was thought that correlation between two variables indicated normal equilibration in the target animals investigated using multiple elements. As mentioned above, we decided to use the regression line obtained in Fig. 6 and the equal probability ellipse as the multiple elements standard regression line, MSRL, and the multiple elements equal probability ellipse, MEPE, respectively.

Fig. 6. The relation between the factor score of factor 1 and that of factor 2 in the data used in Figs 4 and 5. Filled squares; birds collected in Chiba Prefecture (Fig. 5), empty squares; common cormorants (Fig. 4). Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds.

#### **3.2.2 Comparison of the degree of contamination of diving and dabbling ducks**

The factor scores of diving ducks and dabbling ducks were compared with the index obtained. As shown in Fig. 7, the factor score obtained from diving ducks was observed to fall within the MEPE, except for one data point. Similarly, two of nine data points were observed to fall outside the MEPE when the data from wild birds collected in Chiba Prefecture were compared with the index (Fig. 8).

An Innovative Approach to Biological Monitoring Using Wildlife 165

be more serious in dabbling than in diving ducks. However, two of eight data points from wigeon collected in Chiba Prefecture were observed to fall outside the MEPE, as did seven of nine results from wigeon collected in Ibaraki Prefecture. As mentioned above, the area from

which the birds were collected was thought to influence the level of contamination.

Fig. 9. Comparison between the index and the results from dabbling ducks collected in Ibaraki Prefecture. Filled squares; wigeon, filled triangles; spotbill duck, empty squares; shoveler. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from

Fig. 10. Comparison between the index and the results from dabbling ducks collected in Akita Prefecture. Filled squares; mallard, empty squares; spotbill duck, filled circles; teal. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild

wild birds (empty circles) used in Fig. 6.

birds (empty circles) used in Fig. 6.

Fig. 7. The comparison between the index and the data obtained from seabirds. Filled squares: diving ducks collected in Chiba Prefecture, empty squares; diving ducks collected in Tokyo, asterisk; seabirds collected in Tokyo, filled triangles; diving ducks collected in Ishikawa Prefecture. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds (empty circles) used in Fig. 6.

Fig. 8. Comparison between the index and the data obtained from dabbling ducks collected in Chiba Prefecture. Filled squares; wigeon, filled triangles; teal. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds (empty circles) used in Fig. 6.

On the other hand, the data from dabbling ducks showed a marked tendency to deviate from the MEPE (Figs 9 and 10). Dabbling ducks inhabit inland water environments such as lakes and marshes. Thus, it is thought that the degree of contamination with multiple elements may

Fig. 7. The comparison between the index and the data obtained from seabirds. Filled squares: diving ducks collected in Chiba Prefecture, empty squares; diving ducks collected in Tokyo, asterisk; seabirds collected in Tokyo, filled triangles; diving ducks collected in Ishikawa Prefecture. Dotted line; 95% equal probability ellipse, solid line; regression line

Fig. 8. Comparison between the index and the data obtained from dabbling ducks collected in Chiba Prefecture. Filled squares; wigeon, filled triangles; teal. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds (empty circles) used

On the other hand, the data from dabbling ducks showed a marked tendency to deviate from the MEPE (Figs 9 and 10). Dabbling ducks inhabit inland water environments such as lakes and marshes. Thus, it is thought that the degree of contamination with multiple elements may

obtained from wild birds (empty circles) used in Fig. 6.

in Fig. 6.

be more serious in dabbling than in diving ducks. However, two of eight data points from wigeon collected in Chiba Prefecture were observed to fall outside the MEPE, as did seven of nine results from wigeon collected in Ibaraki Prefecture. As mentioned above, the area from which the birds were collected was thought to influence the level of contamination.

Fig. 9. Comparison between the index and the results from dabbling ducks collected in Ibaraki Prefecture. Filled squares; wigeon, filled triangles; spotbill duck, empty squares; shoveler. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds (empty circles) used in Fig. 6.

Fig. 10. Comparison between the index and the results from dabbling ducks collected in Akita Prefecture. Filled squares; mallard, empty squares; spotbill duck, filled circles; teal. Dotted line; 95% equal probability ellipse, solid line; regression line obtained from wild birds (empty circles) used in Fig. 6.

An Innovative Approach to Biological Monitoring Using Wildlife 167

Kenntner, N., Crettenand, Y., Fünfstück, H-J., Janovsky, M. & Tataruch, F. (2007). Lead

Mochizuki, M., Hondo, R., Kumon, K., Sasaki, R., Matsuba, H. & Ueda, F. (2002a). Cadmium

Mochizuki, M., Kitamura, T., Okutomi, Y., Yamamoto, H., Suzuki, T., Mori, M., Hondo, R.,

Mochizuki, M., Kudo, E., Kikuchi, M., Takano, T., Taniuchi, Y., Kitamura, T., Hondo, R. &

*Biological Trace Element Research,* Vol.142, No.1, pp.117-126, ISSN 0163-4984 Mochizuki, M., Mori, M., Akinaga, M., Yugami, K., Oya, C., Hondo, R. & Ueda, F. (2005).

Mochizuki, M., Mori, M., Hondo, R. & Ueda, F. (2008). A new index for evaluation of

Mochizuki, M., Mori, M., Hondo, R. & Ueda, F. (2009). A new index for heavy metals in

Mochizuki, M., Mori, M., Hondo, R. & Ueda, F. (2010a). A cadmium standard regression

Mochizuki, M., Mori, M., Hondo, R. & Ueda, F. (2011c). The biological monitoring of wild

Mochizuki, M., Mori, M., Kajigaya, H., Hayama, S., Ochiai, Y., Hondo, R. & Ueda, F. (2011d).

*Assessment,* Vol. 137, No.1-3, pp.35-49, ISSN 0167-6369

Publishers, ISBN, 978-1-60876-487-7, New York, USA

Vol. 19, Issue 6, pp. 535-546, ISSN 1054-853X

ISBN 978-960-474-125-0, Athens, Greece, September 28-30, 2009

No.1-3, pp. 211-223, ISSN 0163-4984

978-1-61761-672-3, New York, USA

No.3, pp. 664-668, ISSN 0090-3558

1054-853X

poisoning and heavy metal exposure of golden eagles (*Aquila chrysaetos*) from the European Alps. *Journal of Ornithology,* Vol. 148, No.2, pp.173-177, ISSN 0021-8375 Meador, J.P., Ernest, D., Hohn, A.A., Tilbury, K., Gorzelany, J., Worthy, G. & Stein, J.E.

(1999). Comparison of elements in bottlenose dolphins stranded on the beaches of Texas and Florida in the Gulf of Mexico over a one-year period. *Archives of Environmental Contamination and Toxicology,* Vol.36, No.1, pp. 87-98, ISSN 0090-4341

contamination in wild birds as an indicator of environmental pollution. *Environmental Monitoring and Assessment,* Vol.73, No.3, pp.229-235, ISSN 0167-6369 Mochizuki, M., Hondo, R., & Ueda, F. (2002b). Simultaneous analysis for multiple heavy

metals in contaminated biological samples. *Biological Trace Element Research*. Vol. 87,

Yumoto, N., Kajigaya, H. & Ueda, F. (2011a). Biological monitoring using new cadmium indexes: cadmium contamination in seabirds, In: *Advances in Medicine and Biology*. *Volume 33*, L.V. Berhardt, (Ed.), 87-96, Nova Science Publishers, Inc. ISBN

Ueda, F. (2011b). A basic study on the biological monitoring for vanadium-effects of vanadium on Vero cells and the evaluation of intracellular vanadium contents.

Thallium contamination in wild ducks in Japan. *Journal of Wildlife Diseases*, Vol.41,

cadmium pollution in birds and mammals. *Environmental Monitoring and* 

biological monitoring, *Proceedings of 5th international conference on energy, environment, ecosystem and sustainable development,* pp. 185-191, ISSN 1790-5095,

line: A possible new index for biological monitoring. In: *Impact, monitoring and management of environmental pollution,* Ahmed El Nemr (Ed.), 331-338, Nova Science

birds: Part II – The possibility of a new index for biological monitoring. *International Journal of Energy, Environment, and Economics*, Vol. 19, Issue 6, pp.525-534, ISSN

The biological monitoring of wild birds, PartⅠ: The cadmium content of organs from migratory birds. *International Journal of Energy, Environment, and Economics*.
