Is government being honest about how rhinos are dying?

The numbers regarding rhino mortality in South Africa, and specifically SA National Parks’ (SANParks’) Kruger National Park, are not adding up.

It was a pleased department of environmental affairs (DEA) that claimed another drop in the number of poached carcasses discovered in South Africa for the second year in a row, now down to 1 054 for 2016.

But here’s where it starts being complicated:

“We have found that the deaths of rhino in the KNP, as a result of the drought or drought condition, has risen,” the DEA said in its statement.

This number is not recorded and the DEA referred The Citizen to SANParks for comment.

“Those numbers are for the whole of 2016,” said SANParks’ most senior spokesperson, William Mabasa.

“The numbers issued by the ministry are strictly for poached animals,” Mabasa said.

But he then contradicted himself by saying: “It depends on when the animal died and when was the census done.”

Mabasa confirmed that when rhino were poached, the details were given to the department as soon as they became known, but Mabasa could not specify whethe the poached number of rhino was included in the yearly census, or not.

He referred The Citizen to the DEA, which had earlier referred us to SANParks.

The seeming obfuscation sheds no further light on how many rhino died from drought, how many actually died from a poacher’s bullet, how many were translocated, and how many, approximately, are actually in South Africa.

Approximately, because a true number is almost impossible to obtain given the dense thicket the animal will almost always try and hide in when it hears a helicopter overhead.

According to the DEA statement, a rhino survey in September 2016 using the scientifically accepted block-count method recorded that a total of 6 649 to 7 830 white rhino lived in the Kruger National Park.

This was lower than the 8 365 to 9 337 that lived in the park during 2015.

“It must be noted that the natural deaths of white rhino increased due to the unprecedented drought conditions,” said the DEA.

In an admittedly oversimplified calculation, 8 365 minus 6 649 equals 1716 plus 662 animals poached means that about 2 378 rhino disappeared from the books last year in the Kruger National Park alone.

However, says an Outraged SA Citizens Against Poaching (Oscap) source, where are the carcasses?

Using averages with a margin of error (see box below), the Oscap source said if the actual decrease was only 847 rhino lost, then the reported poaching and relocation would account for nearly 100% of these losses.

“This would be a very positive outcome, and would make sense, since the growth rate, if poaching was ignored, would likely still be positive (more are born than die naturally) despite the drought. The drought is likely to decrease nativity, lengthen gestation, etc, but is not likely to send the population into decline on its own,” said the source.

“After all, as many have said, we did not see (non-poached) carcasses of rhino, or even slightly emaciated rhino, but we know they are drought-resilient because of very effective digestive systems that can process even the least palatable of grass.

“On the other hand, if the decrease is closer to 2 377 rhinos lost, we are looking at devastating losses due to natural causes or a possible scandal of rhino disappearing ‘off the books’ by artificial means.

“In this case, 1 533 had to have died from ‘natural’ causes, but as already explained, this makes very little sense. Some rhino might have migrated into adjacent reserves on the western border in search of better grazing, and thus evaded the count.

“In conclusion, the number of rhino poached is either the only major negative contributor to the KNP rhino population or rhinos are being ‘made to disappear’ under the smokescreen of the drought,” said the source.

According to the DEA, 106 rhino were translocated during 2016 to private rhino strongholds, following suitability assessments conducted by SANParks late last year.

“Overall, our translocations have been successful and no translocated rhino were poached,” DEA said.

Somewhere in the middle, the truth lies.

Rhino statistical maths (eish, numbers)

Let the ‘average’ population from 2015 be ‘A’ and that from 2016 be ‘B’, and the error margin on those values be ‘a’ and ‘b’, respectively. Hence, eg, A = 8851, and a=486, so that A+a=9337 and A-a=8365. This gives the range reported by the minister.

Then, define C = A – B. This should give the 1612 difference. The mathematically correct way of calculating the corresponding error margin on C, let’s call it ‘c’, is:

c = square root of [(a*a)+(b*b)], which, if a=486 and b=591, gives c=765.

This is how I calculated the other day that the number of rhino less from 2015 to 2016 lies between (C-c)=1612-765=847 and (C+c)=1612+765=2377.

Unfortunately, there is no way, from a purely statistical viewpoint, to give preference to any value within the reported range. That said, be very careful assuming that the average value, eg, the 1612 value, is the best representative value. It isn’t necessarily.

Just so, neither are the upper or lower limits. All we know is that the value lies somewhere in that range, and we can narrow it down based only on physical, real-world clues.