The problem with panda pregnancies
Predicting a panda pregnancy is problematic.
For decades scientists have been searching for a sure-fire way of determining whether a giant panda is going to give birth.
The biggest problem is the bears' tendency to have a pseudo - or phantom - pregnancy.
That means indicators such as changes in behaviour and hormones are not definitive.
Even an ultrasound scan is not guaranteed to help as a panda foetus is tiny, difficult to detect and develops late.
At Edinburgh Zoo they are well aware of the problem as they try to assess whether Tian Tian, the UK's only female panda, is pregnant following an attempt in April at artificial insemination.Biological marker
"They have a complicated biology," explains Iain Valentine, director of the zoo's giant panda programme.
"You've got this pseudo-pregnancy factor so using hormones as a biomarker for pregnancy in pandas just doesn't work. You have to use other tests."
The search for this biomarker, or biological marker, has been going on for decades.
Recent advances in the US have given hope to panda breeders around the world, including the team in Scotland.
Researchers from Memphis Zoo in Tennessee have developed a test for an "acute phase protein" - a protein which is affected by inflammation.
End Quote Iain Valentine Edinburgh Zoo
The profile for a panda that will get pregnant and carry full term is different from one that will get pregnant and lose the pregnancy or one that is not pregnant at all”
An initial paper on the subject published in the journal PLOS ONE in 2011 suggested that a urinary protein called ceruloplasmin increased in pregnant pandas.
"The thought is that (the fertilised egg) causes an inflammation reaction within the uterus and then this causes this acute protein to appear and you then pick up the traces of this in the urine," said Mr Valentine.
Researchers have been refining the technique, he added, with a further paper due soon.
Helpfully ceruloplasmin also appears to predict whether or not the pregnancy will be successful.
"What seems to be the case is that the profile for a panda that will get pregnant and carry full term is different from one that will get pregnant and lose the pregnancy or one that is not pregnant at all," Mr Valentine added
In Tian Tian's case the test suggested that she is pregnant and is likely to carry to term.
Her urine is being analysed in Berlin: American research assessed in Germany to help the breeding of a Chinese panda which lives in Scotland.
It is "the proving of collaborative international science," according to Professor Chris West, chief executive of the Royal Zoological Society of Scotland, the charity which runs Edinburgh Zoo.
He believes such international co-operation is "hugely important" for the survival of the species.
"We aim to breed them as a genetic reservoir in case they go extinct in the wild," he says.
"All of the organisations that care for pandas including the Chinese share information and expertise very generously so we're not working on this alone."
Ceruloplasmin analysis sounds extremely promising but assessing a panda's hormone levels is not a waste of time.
Tian Tian's urinary progesterone showed a sharp jump on 15 July, a result which was confirmed this week by the Leibniz Institute for Zoo and Wildlife Research in Berlin.
Although this could suggest either a pregnancy or a pseudo-pregnancy, it does enable keepers to narrow down the window when the panda might give birth.
The gestation period is difficult to assess since giant pandas exhibit what is called "delayed implantation."
A study by Hemin Zhang and colleagues published by the journal Reproduction in 2009 describes the phenomenon thus: "...the embryo floats in the womb and arrests development until it attaches to the uterus months later."
There is an initial rise in progesterone when the animal breeds and research suggests that a secondary rise occurs around the time that the embryo is implanted in the wall of the uterus.
Scientists estimate that a panda will give birth between 40-55 days after the second rise in her progesterone.
In Tian Tian's case that means the possibility of a cub between 24 August and 10 September.
If she is not pregnant her progesterone levels will begin to drop sharply in mid-August. If she is, they will remain high and the hope of a cub will increase by the day.
"There is a sense of pressure. There is a sense of expectation," says Prof West, adding "we want to deliver."
So, presumably, does Tian Tian.