Hematology and Plasma Chemistry Reference

Intervals for Cultured Tilapia (Oreochromis Hybrid)

Terry C. Hrubec, DVM, PhD; Jenifer L. Cardinale, DVM; Stephen A. Smith, DVM, PhD

Abstract: Tilapia are a commonly aquacultured fish yet little is known about their normal physiology and
response to disease. In this study we determined the results of complete hematologic (n = 40) and plasma bio-
chemical profiles (n = 63) in production tilapia (Oreochromis hybrids). The fish were raised in recirculating
systems with a high stocking density (120 g/L), and were in the middle of a 15-month production cycle. Blood
was analyzed using standard techniques, and reference intervals were determined using nonparametric meth-
ods. Non-production tilapia (n = 15) from low-density tanks (4 g/L) also were sampled; the clinical chemistry
results were compared to reference intervals from the fish raised in high-density tanks. Differences were noted
in plasma protein, calcium and phosphorus concentrations, such that reference intervals for high-den-sity
production tilapia were not applicable to fish raised under different environmental and management con-
ditions. (Vet Clin Pathol 2000;29:7-12)

Key Words: Clinical chemistry, fish, hematology, reference intervals, tilapia

◆
Tilapia are the second most commonly cultured fish in the Aquaculture Center.The fish, in the middle of their pro-
world,1,2 and are a food staple in many parts of Africa, Asia duction cycle (mean weight 240 g, mean length 22 cm),
and South America. Tilapia consumption has increased in were fed a commercial tilapia feed (Southern States,
the United States, where it is the fourth most commonly Richmond, VA, USA) at 4% of body weight per day. Tanks
cultured food fish. Aquaculture of tilapia, as with other received a 12% fresh water exchange per day. Water quality
species of finfish, is adversely affect-ed by production was monitored daily and was considered acceptable for
related disorders and infectious dis-eases.1 Unfortunately, high-density production systems (Table 1). Each of the 8
there are few diagnostic tools available to veterinarians and tanks was sampled once, with 10 fish per tank bled for
fish health professionals to evaluate disease in fish. Many biochemical analysis and 10 fish bled for hematologic
of the clinical tools used to evaluate mammalian health are analysis. Fish with any gross abnormal-ity were not
not developed for use in fishes. As the aquaculture industry included in the study. Careful netting and handling was
expands, there is an increasing need for improved implemented to minimize stress. Fish were rapidly netted,
diagnostic methods. Hematology and clinical chemistry anesthetized in aerated buffered tricaine methanesulfonate
analysis, although not used regularly in fish medicine, can (MS-222, Sigma Chemical Co. St. Louis, MO, USA) and
provide sub-stantial diagnostic information once reference bled with a needle and syringe from the caudal vessels. The
values are established. In this study, we determined blood was placed in tubes containing either lithium heparin
reference inter-vals for hematologic and plasma chemistry for chemistry analysis, or EDTA for hematologic analysis.
analytes in cultured tilapia. We also evaluated clinical Any hemo-lyzed, clotted or insufficient volume samples
chemistry results from a small group of tilapia raised under were dis-carded. After sampling, fish were placed in a
differ-ent culture conditions. To our knowledge, this is the separate tank of fresh water for recovery.
first study to determine complete hematologic and clinical
chemistry results for tilapia, and to report the values as Blood in heparinized tubes was centrifuged imme-
reference intervals suitable for diagnostic use. diately at 14,000 g for 5 minutes. The plasma was col-
lected and frozen at –10°C until all fish were sampled.
Plasma samples were analyzed using an automated dry
Materials and Methods chemistry system (Kodak Ektachem 700, Eastman Kodak
Co., Rochester, NY, USA) for total protein, albu-min,
Production tilapia (Oreochromis nilotica O. mossam-bicus creatinine, ammonia, total bilirubin, cholesterol, sodium,
O. aureus hybrids) were maintained indoors in 8 high- chloride, potassium, calcium, magnesium and phosphorus
density (120 g/L) tanks (10,220 L) at Virginia Tech’s concentrations, and alkaline phosphatase

From the Department of Biomedical Sciences and Pathobiology, Virginia-Maryland Regional College of Veterinary Medicine, Virginia
Polytechnic Institute and State University, Blacksburg, VA 24061. Address correspondence to Dr. Hrubec ([email protected]).

Vol. 29 / No. 1 / 2000 Veterinary Clinical Pathology Page 7

 Tilapia Hematology and Plasma Chemistry Values

Table 1. Mean water quality values for high-density (120 Table 2. Plasma chemistry reference intervals for hybrid
g/L) and low-density (4.3 g/L) tilapia production systems. tilapia (n = 63) raised in high-density production systems.

Parameter High Density Low Density Analyte Reference Interval Median

Temperature (°C) 29.9 26.7 Total Protein (g/dL) 2.9-6.6 3.9
pH 7.4 7.5 Albumin (g/dL) 1.3-2.6 1.8

NH3 non-ionized (mg/L) 0.020 0.004 Globulins (g/dL) 1.6-4.2 2.1

NO2-N (mg/L) 0.36 0.01 Creatinine (mg/dL) 0.1-0.5 0.2

70 3 Ammonia (µg/dL) 111-414 249

NO3-N (mg/L)
Alkalinity (mg/L)* 105.0 34.2 Total bilirubin (mg/dL) 0.0-0.3 0.2
† 281.0 51.3 ALP (U/L) 15-39 22
Hardness (mg/L)
AST (U/L) 9-102 26
Dissolved oxygen (mg/L) 9.4 ND
Sodium (mEq/L) 139-160 151
Turbidity (NTU) 9 ND
Potassium (mEq/L) 3.5-5.4 4.3
*Alkalinity is a measure of the buffering capacity of the water.
† Chloride (mEq/L) 128-142 136
Hardness is the sum of the calcium and magnesium ions
in the water. ND = not determined Calcium (mg/dL) 13.6-69.4 31.0
Magnesium (mg/dL) 1.9-3.5 2.5
(ALP) and aspartate aminotransferase (AST) activities. Phosphorus (mg/dL) 5.5-22.1 9.1
Globulins were calculated from the difference between total
Glucose (mg/dL) 30-69 46
protein and albumin values.
Blood from the EDTA tubes was drawn into micro- Cholesterol (mg/dL) 110-318 189
hematocrit tubes and the PCV was determined after
ALP = alkaline phosphatase; AST = aspartate aminotransferase
centrifugation for 5 minutes at 10,000 g. Plasma pro-tein
was determined with a clinical refractometer using plasma
from the microhematocrit tube. The total RBC count was thrombocyte clumping (< 4 cells clumped) were used for
determined manually with a Neubauer hemacytometer differential counts. If thrombocyte clumping was more
using Natt-Herrick’s solution as a dilu-ent stain.3 Slight severe, the samples were discarded. Leukocyte size was
thrombocyte clumping prevented accurate enumeration of a measured in 10 WBC of each type, in 10 different fish, to
combined WBC + thrombo-cyte count on the give a range of diameters for each cell type. Hemoglobin
hemacytometer. Blood smears were made within 45 was determined, using the cyanomethemoglobin method
minutes of sample collection, stained with Wright-Giemsa, (Sigma). Prior to reading the absorbance, hemo-globin test
and used to determine the WBC + thrombocyte count, samples were centrifuged to remove dis-persed nuclear
differential WBC counts, and cell size estimates. material. The RBC, MCV, MCH, and MCHC were
Percentages of RBC and WBC + throm-bocytes were calculated by standard formulas.
determined by counting 1,500 cells. The WBC + Reference intervals were determined following the
thrombocyte percentage was multiplied by the RBC count guidelines proposed by the National Committee for Clinical
from the hemacytometer to determine the WBC +
Laboratory Standards (NCCLS).7 As suggested in these
thrombocyte absolute count. For the differential count,
WBC and thrombocytes were counted until 200 WBC were guidelines, outliers were determined using the 1/3 ratio
enumerated on blood smears, and the per-centages of each difference/range ratio. Two hematologic values and one
WBC type and of thrombocytes were multiplied by the total biochemical value were identified as outliers and were
WBC + thrombocyte count to obtain absolute differential deleted. Remaining values were then ranked, and the high
cell counts. Thrombocyte numbers were subtracted from the and low 2.5% were discarded. The range of the remaining
WBC + thrombocyte count to obtain a total WBC count. values provided the reference interval.
This method of man-ually determining total WBC and Plasma chemistry values from 15 non-production
tilapia (mean weight 350 g) were compared to reference
differential counts has been recommended for avian4 and
intervals determined for the production fish. Non-pro-
fish5 blood because nucleated RBC prevent accurate duction fish conditions were characterized by lower
enumeration using automated analysis.6 Only smears with stocking density (4.3 g/L), lower feeding rate of the
mild

Page 8 Veterinary Clinical Pathology Vol. 29 / No. 1 / 2000

 Hrubec, Cardinale, Smith

Table 3. Plasma chemistry results for hybrid tilapia (n = 15) Table 4. Hematology reference intervals for hybrid tilapia
raised in low-density systems compared to reference intervals (n = 40) raised in high-density production systems.
from tilapia raised in high-density systems (see Table 2).

Analyte Reference Interval Median

Analyte Range* Median Compared to
PCV (%) 27-37 33
Ref. Intervals (n)
Low Within High Hemoglobin (g/dL) 7.0-9.8 8.2
Total protein (g/dL) 2.3-3.6 2.9 7 8 0 MCV (fL) 115-183 135.7
Albumin (g/dL) 1.0-1.6 1.2 9 6 0 MCH (pg) 28.3-42.3 34.9
Globulins (g/dL) 1.3-2.1 1.6 6 9 0 MCHC (g/dL) 22-29 25.7
Creatinine (mg/dL) 0.2-1.1 0 0 14 1 Plasma protein (g/dL) 4.8-7.8 6.1
Total bilirubin (mg/dL) 0.0-0.1 0 0 15 0 6 1.91-2.83 2.31
RBC (X 10 /µL)
ALP (U/L) 16-38 26 0 15 0 WBC (/µL) 21,559-154,690 75,659
AST (U/L) 5-124 18 1 13 1
Small lymphocytes (/µL) 6,776-136,390 61,164
Sodium (mEq/L) 140-156 150 0 15 0
Large lymphocytes (/µL) 2,852-30,833 10,720
Potassium (mEq/L) 3.2-4.3 3.9 2 13 0
Neutrophils (/µL) 557-9,873 1,805
Chloride (mEq/L) 136-147 141 0 10 5
Monocytes (/µL) 400-4,286 1,520
Calcium (mg/dL) 10.5-19.0 11.8 14 1 0
Eosinophils (/µL) 35-1,645 334
Magnesium (mg/dL) 2.3-2.8 2.5 0 15 0
TLC (/µL) 35-4,336 972
Phosphorus (mg/dL) 3.5-7.2 4.6 13 2 0
Thrombocytes (/µL) 25,068-85,216 52,762
Glucose (mg/dL) 39-96 52 0 13 2
TLC = thrombocyte-like cell
Cholesterol (mg/dL) 64-299 156 2 13 0

*Minimum-maximum values
ALP = alkaline phosphatase; AST = aspartate aminotransferase and thrombocytes. The RBC were nucleated, oblong cells
measuring 7.0 12.9 µm in size. Nuclei stained purple and
cytoplasm stained reddish-gray. Immature RBC, or
same diet (1% of body weight per day), greater fresh water polychromatophilic RBC, had a blue-grey tinge to the usual
exchange rate (35% per day) and, consequently, more eosinophilic cytoplasm; these cells were included in RBC
optimal water quality (Table 1). Fish were sam-pled and the counts.
blood was processed as described for fish raised in high-
density tanks. Cell morphology

Results Six types of WBC were distinguished and counted: small

lymphocytes, large lymphocytes, neutrophils, monocytes,
Analyses eosinophils and thrombocyte-like cells (TLC) (Figures 1-6).
Small lymphocytes ranged from 4.6 to 5.0 µm in diameter,
Reference intervals for plasma chemistry analytes were and had dark purple, round to oval, condensed nuclei with
summarized (Table 2). Plasma chemistry values from non- clumped chromatin. Small lymphocyte cytoplasm was deep
production tilapia were summarized and com-pared to blue and often consist-ed of only a thin rim encircling the
reference intervals from production tilapia (Table 3). nucleus. Large lym-phocytes measured 5.7 to 6.4 µm in
Protein concentrations in tilapia raised in high-density diameter, and had round nuclei that were larger and had a
systems were slightly higher, while calci-um and more open chromatin pattern than small lymphocytes.
phosphorus levels were much higher com-pared to fish in Large lym-phocyte cytoplasm was more deeply basophilic
low-density systems. and abundant than that of small lymphocytes. Both small
Reference intervals for hematology analytes were and large lymphocytes had high N:C ratios. Frequently,
summarized (Table 4). As in other species of fish, there lymphocytes had cytoplasmic pseudopods and azurophilic
were three types of circulating blood cells: RBC, WBC cytoplasmic granules.

Vol. 29 / No. 1 / 2000 Veterinary Clinical Pathology Page 9

 Tilapia Hematology and Plasma Chemistry Values

Figure 1. Small lymphocyte (SL), large lymphocyte (LL), Figure 4. Eosinophil (E) with an eccentric nucleus and moderate num-
and a thrombocyte (T). Thrombocytes can be distinguished ber of eosinophilic granules. A small lymphocyte (SL) is also present.
from lym-phocytes by the highly condensed nucleus and Photograph taken with a didymium filter. Wright-Giemsa. Bar = 10 µm.
grey cytoplasm. Wright-Giemsa. Bar = 10 µm.

Figure 5. Eosinophil (E) with a large number of granules. A

throm-bocyte (T) with a slightly indented nucleus and one
Figure 2. Monocyte (M) with an indented nucleus and vacuoles.
with a seg-mented nucleus (insert) are present. Photograph
Wright-Giemsa. Bar = 10 µm. taken with a didymium filter. Wright-Giemsa. Bar = 10 µm.

Figure 3. Neutrophil (N) with an irregular cytoplasmic Figure 6. Thrombocyte-like-cell (TLC), thrombocyte (T) and a small
border. A small lymphocyte (SL) and a thrombocyte (T) lymphocyte (SL). The TLC can be distinguished by a more open
are also present. The insert depicts a neutrophil with a nucleus and more abundant grey cytoplasm than is seen in either
bean shaped nucleus. Wright-Giemsa. Bar = 10 µm. the thrombocyte or lymphocyte. Wright-Giemsa. Bar = 10 µm.

Page 10 Veterinary Clinical Pathology Vol. 29 / No. 1 / 2000

 Hrubec, Cardinale, Smith

Monocytes measured 9.4 to 10.7 µm in diameter and defined.

had round or indented, eccentric nuclei with an open Leukocyte numbers reported for one parent species, O.
chromatin pattern (Figure 2). Monocyte cytoplasm was mossambicus,9 were similar to those for Oreochromis
deeply basophilic and often contained clear, punc-tate, hybrids reported herein. Additionally, WBC types and
cytoplasmic vacuoles. Neutrophils were 9.6 to 10.8 µm in morphology in this study were similar to those described
diameter, and had round, oval or reniform nuclei with an for O. mossambicus.11 In O. mossam-bicus, three
open chromatin pattern (Figure 3). Neutrophil cytoplasm granulocyte types were distinguished by ultrastructural
was stippled, gray to lightly basophilic, and had occasional observation. Two granulocytes repre-sented the neutrophils
vacuoles and basophilic, intracytoplas-mic inclusion bodies. and eosinophils seen by light microscopy, while the third
Cytoplasmic borders were irregu-lar. Eosinophils were 5.7 granulocyte was unidenti-fied by light microscopy. This
to 6.7 µm in diameter, and had round, light purple, often third cell type may be the TLC described here. The TLC is
eccentric nuclei with open chromatin (Figures 4, 5). a distinct cell type that has been observed and described in
Eosinophil cytoplasm was lightly basophilic and contained other fish species.12 The cells were first described in hybrid
eosinophilic granules ranging from small and few to large striped bass, in which they superficially resembled
and numerous. The eosinophil granules occasionally thrombocytes; they were termed TLC to distinguish them
obscured the nucleus. Thrombocyte-like cells were clearly from other unidentified cells. The function and origin of
identified in WBC differentials as a distinct cell type TLC were not determined. In tilapia, TLC appeared similar
(Figure 6). The TLC were larger (5.6 to 7.2 µm diameter) to a small neutrophil with lighter cytoplasm. Because TLC
and had a less con-densed nuclear chromatin pattern than were identified in all individual tilapia, as well as in other
thrombocytes. The cytoplasm of TLC was pale gray, fish species, the cell probably represents a matu-rational
slightly stippled and more abundant than that of stage of one of the WBC, although the cell’s lin-eage and
thrombocytes. Throm-bocytes (counted separately from function remain unknown.
WBC) were 4.7 to 5.5 µm in diameter, and had very dark,
dense, purple nuclei that were round, polygonal or Distinguishing lymphocytes from thrombocytes can be
segmented (Figures 1, 3, 5, 6). Thrombocyte cytoplasm was difficult; however, in our laboratory we do a number of
clear to slightly basophilic, and occasionally vacuolated. procedures to help ensure cells are identi-fied correctly,
including routine repetitions of differen-tial counts, and
In a few fish, large cells with extremely blue cyto- having more than one person repeat the differential counts.
plasm were rarely noted. These cells were considered While these measures do not always ensure the cells are
undifferentiated blast cells; due to their immaturity, we accurately identified, they do pro-vide consistency. The
were unable to morphologically determine the cell line. The high WBC and lymphocyte counts were consistent with
blast cells were not included in cell counts. values obtained in our laboratory for other fish species
raised in high-density recircula-tion systems, including
hybrid striped bass, yellow perch and tilapia (unpublished
Discussion
observations).12 Leukocyte counts in these same species
Although tilapia are the second most frequently cul-tured maintained under low production densities are lower by
fish in the world, there are surprisingly few reports of 40% to 50%.12 Fish in high-density production systems are
normal blood values. In addition, published values are exposed constantly to high bacterial loads in the water and
severely limited by low fish numbers or by the few analytes less than optimal water quality, factors which may influence
measured.8-10 Compared with previ-ously reported values, WBC counts.
our results were similar for most analytes. 8-10 Terao and
Ogawa8 reported much higher mean concentrations of When reference intervals for production tilapia were
cholesterol (567 mg/dL), glu-cose (408 mg/dL) and compared to clinical chemistry values obtained from tilapia
creatinine (4.3 mg/dL). Hussein et al 10 obtained slightly raised in low-density tanks, there were notable differences
lower values for PCV (20%), hemo-globin concentration in protein, calcium and phosphorus concentrations. Protein
(6.0 g/dL) and RBC count (1.31 x106/µL) than those in our concentrations in tilapia raised in high-density systems
were slightly higher, while cal-cium and phosphorus levels
study. Haniffa and Vijayarani9 reported even lower PCV
were much higher com-pared to fish in low-density
(10%), hemoglobin (3.7 g/dL) and RBC (0.90 x10 6/µL) systems. Similar differ-ences in blood values were observed
values.These differences may be due to the fact that the in hybrid striped bass raised in high-density production
production fish in our study were hybrids of parental
species used in previous stud-ies.The earlier studies are of systems,13 although the differences in calcium and
limited use for comparison and diagnostic purposes because phosphorus values were greater in these tilapia.The reason
too few fish were sampled and the fish populations were for high-er calcium and phosphorus values is unknown, but
not well is

Vol. 29 / No. 1 / 2000 Veterinary Clinical Pathology Page 11

 Tilapia Hematology and Plasma Chemistry Values

probably an effect of water quality or stocking density, ulations of fish within a single species, since culture
since the two groups of tilapia were from the same stock conditions and environmental variables can markedly affect
source and were fed the same diets. Water hardness was blood values.12,13,16,17 As demonstrated in our com-parison
much higher in production fish tanks compared to the low- of blood values from tilapia in high and low-density
density tanks, and may have influenced blood calcium systems, reference intervals developed in high-density
levels.Total protein, albumin, and globulin con-centrations production tilapia are not applicable to fish raised under
in fish from high-density, recirculating sys-tems may be different environmental and management conditions.
influenced by the characteristic high organic load and
bacterial count, which could have induced a generalized As the aquaculture industry expands, tools to mon-itor
immune response. the health status of fish using standardized non-lethal and
Calculation of reference intervals as ± 2 SD from the inexpensive methods will be needed. Evaluation of
mean is valid only when blood values follow a nor-mal hematologic and blood chemistry ana-lytes will enhance the
distribution. It is incorrect to assume that biological culture of fish by facilitating early detection of infectious
parameters are distributed normally, therefore non- disease and identification of sub-lethal conditions affecting
parametric methods are more accurate for determining production performance. This, in turn, will contribute to
reference intervals.14 Techniques to properly determine more specific, timely and effective disease treatments in the
reference intervals have been established by the NCCLS, 7 future. ◊
and suggestions for their use and interpreta-tion have been
discussed.15 Unfortunately, reference values are not used on Acknowledgements
a routine basis in fish medicine, and the number of studies The authors thank Butch Kukanich for technical assistance, and
the staff and students of the Virginia Tech Aquaculture Center,
in which reference intervals have been determined for fish
particularly Brian Brazil, for the culture and mainte-nance of the
species is limited. The majority of blood values determined tilapia used in this study. This project was fund-ed in part by the
for fishes have been reported as mean ± SD. As for Office of Research and Graduate Studies at the Virginia-Maryland
mammals, refer-ence intervals should be determined for Regional College of Veterinary Medicine.
different pop-

References
1. Centers for Epidemiology and Animal Health. Overview of 10. Hussein SY, El-Nasser MA, Ahmed SM. Comparative studies on
Aquaculture in the United States. Fort Collins, Colo: US Dept the effect of the herbicide atrazine on freshwater fish Oreochromis
Agriculture, APHIS; 1995. niloticus and Chrysichthyes auratus at Assiut, Egypt. Bull Environ
Contam Toxicol 1996;57:503-510.
2. Status of the world aquaculture 1995. Aquaculture Magazine
Buyer’s Guide ’96. 1996:6-27. 11. Doggett TA, Harris JE. Ultrastructure of the peripheral leuco-cytes
of Oreochromis mossambicus. J Fish Biol 1989;33:747-756.
3. Natt MP, Herrick CA. A new blood diluent for counting ery-
throcytes and leukocytes of the chicken. Poult Sci 1952;31:735- 12. Hrubec TC, Smith SA, Robertson JL, et al. Comparison of
738. hematologic reference intervals between culture system and type
of hybrid striped bass. Am J Vet Res 1996;57:618-623.
4. Zinkl JG. Avian hematology. In: Jain NC, ed. Schalm's Veterinary
Hematology. 2nd ed. Philadelphia, Pa: Lea and Febiger; 13. Hrubec TC, Smith SA, Robertson JL, et al. Blood biochemical
1986:256-260. reference intervals for sunshine bass (Morone chrysops Morone
saxatilis) in three culture systems. Am J Vet Res 1996; 57:624-627.
5. Stoskopf MK. Clinical pathology. In: Stoskopf MK, ed. Fish
Medicine. Philadelphia, Pa: WB Saunders; 1993:113-131. 14. Reed AH, Henry RJ, Mason WB. Influence of statistical method
used on the resulting estimate of normal range. Clin Chem
6. Huffman PA, Arkoosh MR, Casillas E. Characteristics of
1971;17:275-284.
peripheral blood cells from rainbow trout evaluated by parti-cle
counter image analysis and hemocytometric techniques. J Aquatic 15. Lumsden JH. “Normal” or reference values: questions and
Animal Health 1997;9:239-248. comments. Vet Clin Pathol 1998;27:102-106.
7. National Committee for Clinical Laboratory Standards. How to 16. Hrubec TC, Robertson JL, Smith SA. Effects of temperature on
Define, Determine, and Utilize Reference Intervals in the Clinical hematologic and serum biochemical profiles of hybrid striped bass
Laboratory; Proposed Guidelines. Villanova, Pa: NCCLS; 1992. (Morone chrysops X Morone saxatilis). Am J Vet Res 1997;58:126-
Document C28-P. 130.
8. Terao T, Ogawa T. On the biochemical components in the blood of 17. Hrubec TC, Robertson JL, Smith SA. Effects of ammonia and
the cultured cichlid fish Tilapia nilotica. Sci Rep Hokkaido Fish nitrate concentration on hematologic and serum biochemical
Hatchery 1984;39:83-88. profiles of hybrid striped bass (Morone chrysops X Morone sax-
atilis). Am J Vet Res 1997;58:131-136.
9. Haniffa MA, Vijayarani SM. Hematological effects of textile mill
effluent on freshwater fish Oreochromis mossambicus (Trewaves).
Ind J Exp Biol 1989;27:476-478.

Page 12 Veterinary Clinical Pathology Vol. 29 / No. 1 / 2000